Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session

Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session

Session at a glanceSummary, keypoints, and speakers overview

Summary

The India AI Impact Summit gathered government officials, CEOs of major technology firms, and industry leaders to discuss how artificial intelligence can drive India’s economic growth and global leadership [1-3].


Sundar Pichai emphasized that India is poised to become a global AI leader thanks to its talent, deep-tech expertise and startup ecosystem, and announced Google’s full-stack commitment including a $15 billion Vizag AI Hub and partnerships across agriculture, healthcare and skilling [4-10][11-13].


Dario Amodei of Anthropic highlighted the need to both amplify AI’s benefits and manage its risks, proposing collaborative tracking of AI’s economic impact and sharing data to mitigate disruptions [21-35].


Demis Hassabis described AI as a transformative force comparable to ten times the Industrial Revolution, citing AlphaFold’s breakthrough in protein folding as an early example of AI accelerating science and medicine [39-49].


Alexander Wang of Meta outlined plans to empower tens of millions of Indian small businesses through WhatsApp-based services and to extend governance tools to citizens, reinforcing the company’s commitment to AI-driven digital inclusion [54-68].


Arthur Mensch warned that unchecked market concentration could hinder equitable AI benefits and advocated open-source models and multilingual support to ensure broad participation [72-78].


Sam Altman called for the democratization of AI, stressing that sovereign approaches and iterative deployment are essential for a large democracy like India to lead responsibly [82-99].


Brad Smith of Microsoft stressed the historic synergy between Indian and U.S. IT sectors and urged protection of digital sovereignty while enabling cross-border technology collaboration [149-162].


Industry representatives announced specific AI-related investments: Philips on healthcare AI and data access [165-177]; FedEx on logistics analytics and a ₹10 000-crore hub in Navi Mumbai [125-139]; Fujitsu on high-performance computing and data sovereignty [141-146]; Micron on memory manufacturing for AI workloads [190-205]; Qualcomm on a 2-nm chip design in India [210-221]; Cisco on skilling and infrastructure support [230-238]; Cloudflare on an accessibility framework and free credits for startups [248-267]; and Palo Alto on security, governance and AI services for citizens [270-276].


Mukesh Ambani reiterated the “Manav” vision, pledging a 10 lakh-crore AI investment and emphasizing AI’s role in job creation and affordable intelligence for all citizens [278-303].


Natarajan Chandrasekaran highlighted India’s ambition and skill as drivers of AI leadership and committed Tata Group to scaling AI hardware, data centres and global business outreach [306-326].


Sunil Mittal praised India’s rapid 5G rollout and frugal innovation, while Rishi Sunak called for responsible, transparent AI development that lifts the floor for health and education worldwide [329-347][385-406].


Prime Minister Narendra Modi concluded by reaffirming India’s commitment to inclusive, sovereign AI development and urging continued partnership among global stakeholders to realize the nation’s transformative vision [509-528].


Keypoints


Major discussion points


India’s ambition to become a global AI super-power – Speakers repeatedly highlighted India’s talent, startup ecosystem and government vision as the foundation for leading the next AI revolution. Sundar Pichai called India “poised to be a global AI leader” and said Google will bring a “full-stack commitment” [8-9]; Demis Hassabis described India as a “real powerhouse in the AI revolution” and a critical part of a new “golden era of scientific discovery” [41-44]; Mukesh Ambani noted the rapid digital transformation and pledged massive AI investment, saying “India is ready… to take advantage of the AI revolution” [278-292]; Rishi Sunak echoed the ranking of India as a leading AI super-power [387-389].


Large-scale investment and infrastructure commitments across the AI stack – Multiple firms announced hardware, cloud, semiconductor and talent initiatives. Google’s Vizag AI Hub ($15 bn) and end-to-end partnerships were cited [10-11]; Microsoft’s skilling and cross-border collaboration was emphasized [152-155]; Micron detailed its 500,000 sq ft clean-room and memory production to fuel AI [190-204]; General Catalyst pledged a $5 bn fund for Indian AI startups [452-455]; Vinod Khosla and Lightspeed reiterated multi-billion-dollar investment plans and support for AI-driven services [458-465][488-492].


Democratizing AI and driving social impact – Speakers stressed that AI must be accessible to citizens, small businesses, farmers, students and patients. Sam Altman urged “AI has to be democratized” and placed responsibility on India to lead [85-99]; Matthew Prince proposed frameworks to ensure 500,000 AI companies, business models for creators and tools for the poorest [250-259]; Nandan Nilekani gave a concrete example of AI for dairy farmers that went from idea to live service in a month [354-366]; Khosla suggested free AI tutors, doctors and agronomists integrated with Aadhaar [462-470].


Governance, ethics, data sovereignty and security concerns – Several leaders warned that rapid AI diffusion must be paired with robust oversight. Dario Amodei called for shared economic data to mitigate disruptions [31-35]; Arthur Mensch warned about “excessive concentration of power” and advocated open-source models [76-78]; Takahito Tokita highlighted data sovereignty, human dignity and ethical safeguards [141-146]; Nikesh Arora stressed the need for governance, accountability and “kill-switches” to prevent rogue AI [270-276].


Overall purpose / goal


The round-table was convened to align India’s government vision with the global AI ecosystem, secure commitments for infrastructure, talent development and financing, and establish a collaborative framework for responsible, inclusive and economically transformative AI deployment in India and beyond.


Overall tone


The discussion was overwhelmingly upbeat and forward-looking, with participants praising India’s progress and expressing enthusiasm for partnership. As the session progressed, the tone shifted to incorporate cautionary notes on ethics, data sovereignty and security, reflecting a balanced view that combines optimism about AI’s potential with a call for responsible governance.


Speakers

Full session reportComprehensive analysis and detailed insights

The round-table opened with Minister Ashwini Vaishnaw welcoming Prime Minister Narendra Modi, senior government officials and a roster of roughly 28 CEOs, urging participants to keep their remarks concise as the discussion moved forward [1-3].


Sundar Pichai thanked Prime Minister Modi and declared that India is “poised to be a global AI leader” because of its world-class talent, deep-tech expertise and vibrant start-up ecosystem [8]. He outlined Google’s “full-stack commitment” that will span TPUs, cloud infrastructure, research collaborations and end-to-end partnerships across agriculture, healthcare and language access [9-13]. The cornerstone of this commitment is the Vizag AI Hub - a $15 billion investment - which he presented as Google’s initial step in India [10-11].


Dario Amodei of Anthropic stressed that India’s central role must include both amplifying AI’s benefits and managing its risks [22-23]. He called for collaborative tracking of AI’s economic impact, proposing that companies share usage statistics and economic data so that “we can accentuate the good parts and mitigate any of the disruptions” [31-35]. He also highlighted Anthropic’s partnership with the XSTEP Foundation and its Agritech initiatives [36-38].


Demis Hassabis of DeepMind described AI as a transformative force “ten times the Industrial Revolution, ten times the speed, over a decade” [47-49]. He cited AlphaFold’s solution of the 50-year protein-folding problem as an early illustration of AI-driven scientific acceleration [43-44] and argued that India’s talent pool and research capacity are critical to unlocking a new “golden era of scientific discovery” [41-46].


Alexander Wang of Meta highlighted the company’s plan to empower “tens of millions of Indian small businesses” through WhatsApp-based services and to extend digital-governance tools directly to citizens, noting successes such as 100 million subway tickets sold via WhatsApp in Andhra Pradesh [60-66]. He also showcased Meta Ray-Ban glasses and the Be My Eyes capability as product demos for inclusive AI [67-70]. He reaffirmed Meta’s commitment to continue partnering with the Indian government on AI-enabled inclusion [58-59].


Arthur Mensch of Mistral AI warned that unchecked market concentration could lead to “excessive price and extractive economy” [76-77]. He advocated open-source AI as a “common good” that would allow broad participation, multilingual support and cultural nuance, and highlighted recent work on Indian-language audio models [73-78][79-80].


Sam Altman of OpenAI framed AI democratisation as a national imperative, stating that “no company or person or country is equipped to help society navigate that change” [89-90] and urging sovereign, iterative deployment that puts tools in the hands of the Indian populace while mitigating disruption [91-99].


Brad Smith of Microsoft underscored the historic synergy between the Indian and U.S. IT sectors, pledging continued investment, partnership and skilling in India [152-155]. He called for protection of digital sovereignty and cross-border trust, suggesting that India and the United States could serve as a model for responsible AI collaboration [158-162].


Sector-specific announcements followed. Philips detailed its $1.7 billion AI-driven software investment, its work with the Indian Ministry of Health on data-level support for the Digital Act, and plans to aid primary-care workers with AI tools [165-176]. FedEx described a ₹10 000 crore logistics hub in Navi Mumbai, noting AI’s role in cutting freight costs from 15 % to 8 % and the generation of two petabytes of data daily [125-138]. Fujitsu highlighted its capability to build super-computers and quantum computers, stressing data sovereignty, human dignity and ethical AI as essential foundations [141-146]. Micron announced a 500 000 sq ft clean-room and memory-assembly line that will account for 10 % of its global output, positioning memory as “the fuel of AI” [190-204]. Qualcomm celebrated the design of India’s first 2-nm chip, signalling a “tremendous” opportunity for the country to become a global semiconductor player [210-221]. Cisco reported training 800 000 Indians in cybersecurity, AI and networking, and pledged to provide the underlying infrastructure needed for AI democratisation [230-238]. Cloudflare proposed a framework to nurture 500 000 AI companies, provide business models for creators and small businesses, and offered free credits and AI-for-Bharat services across Indian languages [248-267]. Palo Alto Networks announced an AI-security competence centre in Bangalore with 1 500 staff, focusing on governance, accountability and “kill-switches” for autonomous agents [270-276].


Mukesh Ambani, representing Reliance-Jio, reiterated the Prime Minister’s ‘Manav’ (human-centric) vision as an AI manifesto, noting that India has moved from 138th to world-leader in digital connectivity and pledging ₹10 lakh crore (≈ $120 bn) of AI investment over seven years to make intelligence “democratic and affordable” for every citizen [278-303].


Natarajan Chandrasekaran of the Tata Group highlighted India’s “ambition and skill” as twin drivers of AI leadership and committed Tata to scaling AI hardware, building data centres, and extending AI services globally, thereby supporting both national transformation and international outreach [317-320][322-326].


Sunil Mittal of Airtel praised the rapid 5G rollout-“from 138th to number one” in connectivity-and framed AI delivery as a continuation of India’s tradition of frugal innovation, promising to keep the “muscles” of data-centres and fibre affordable for a billion users [329-340].


Former UK Prime Minister Rishi Sunak (actually the current UK Prime Minister) recalled his 2019 Bletchley Park AI summit, cited Stanford’s AI Index ranking India as a leading AI super-power, and warned that AI companies will spend 20 × the resources of the Manhattan Project[385-406][387-389]. He noted that England is just ahead of India in ICC test rankings [390-391] and emphasized that the summit’s ranking comes from Stanford’s AI index [392-393]. Sunak stated that the “first duty … is to ensure safety” and the “second responsibility … is to benefit everyone, especially health and education” [394-395][396-398], urging participants to “raise the ceiling and lift the floor” for all [399-400]. He also highlighted the “special and distinctive thing about these summits” – their public-private nature – as a catalyst for responsible AI development [401-404].


Nandan Nilekani illustrated rapid AI diffusion with the Amul dairy example: an idea presented on 8 January was live by 11 February, delivering AI-driven cattle health advice to millions of farmers, especially women, while keeping data sovereign within Amul [354-366].


Vinod Khosla advocated integrating free AI tutors, doctors and agronomists into the Aadhaar ecosystem, likening them to UPI as a public utility that would ensure universal access and democratic permission for AI [462-470]. He also noted early investments in sovereign AI models such as Sarvam and highlighted the need for AI-driven services that benefit the entire population [458-465].


Hemant Taneja of General Catalyst announced a $5 billion, five-year fund for Indian AI start-ups, describing AI as an “abundance” engine that can empower the country’s young workforce and multiply productivity without causing social disruption [452-455].


Additional commitments included Adobe’s decision to make Photoshop, Acrobat and Firefly free for students and to launch a content-authenticity initiative [115-123]; G42’s plan to build a “token factory” and an “agent factory” in partnership with Indian stakeholders [433-440]; Vertiv’s pledge to expand manufacturing and services in India [422-425]; and the Blue Raman subsea cable project, a joint Google-SPACL effort to enhance AI-ready connectivity between Italy and Mumbai [416-418].


Prime Minister Narendra Modi concluded by reaffirming India’s commitment to an inclusive, sovereign AI ecosystem, urging continued partnership among global stakeholders, and emphasizing that the nation’s future prosperity will be built on “faith, collaboration and the democratic permission of the people” [509-528].


Consensus and points of agreement emerged around three core themes: (1) AI’s transformative economic impact comparable to a ten-fold Industrial Revolution (Hassabis, Altman, Arora, Sunak) [47-49][85-89][270-272][386-389]; (2) the need to democratise AI through open-source models, massive ecosystem growth and integration with public platforms such as Aadhaar, WhatsApp and free-tool initiatives (Altman, Mensch, Khosla, Wang, Narayen) [85-89][73-78][462-470][60-66][115-123]; and (3) large-scale skilling as essential for realising AI’s potential (Pichai, Sweet, Patel, Smith) [9-14][102-108][232-238][152-155].


Key disagreements centred on (a) open-source versus proprietary full-stack approaches (Mensch vs. Pichai, Hassabis and other corporate speakers) [73-78][7-10]; (b) whether AI services should be delivered as Aadhaar-linked public utilities (Khosla) or via private-sector platforms such as WhatsApp and Adobe’s free tools (Wang, Narayen) [462-470][60-66][115-119]; and (c) the scale and focus of investment, with Reliance pledging ₹10 lakh crore for social sectors, General Catalyst committing $5 billion to start-ups, and Khosla highlighting earlier sovereign-AI investments [278-303][452-455][458-465].


Take-aways from the summit include: (i) AI is viewed as a multi-digit GDP driver that will outpace the Industrial Revolution in speed and scale; (ii) India’s talent, start-up ecosystem and recent digital infrastructure position it as a global AI super-power; (iii) there is broad consensus on democratising AI through free tools, open-source models and public-utility integration; (iv) private-sector commitments span the full AI stack-from chips and memory (Micron, Qualcomm) to connectivity (Blue Raman) and sector-specific applications (healthcare, logistics, creative economy); (v) partnerships between government and industry are deemed essential for skilling, research, regulation and responsible deployment; (vi) governance, ethics, security and data sovereignty are non-negotiable pillars; (vii) concrete use-cases such as AI-enabled dairy health, logistics optimisation and AI-driven medical devices are already being piloted; and (viii) large-scale talent development programmes are already underway.


Unresolved issues include the precise design of regulatory frameworks for AI governance and autonomous agents, mechanisms for equitable value distribution, standards for “kill-switches” and cross-border security, and the operationalisation of Aadhaar-linked AI services. Addressing these gaps will require coordinated policy action, continued public-private dialogue and transparent data-sharing mechanisms as advocated by Anthropic and Cloudflare.


In sum, the India AI Impact Summit demonstrated a high degree of alignment among government, multinational corporations and investors on the strategic importance of AI, while also surfacing nuanced debates over openness, public-utility delivery and investment magnitude. The commitments announced-ranging from a $15 billion Vizag hub to a $5 billion venture fund-lay the groundwork for India to pursue an inclusive, sovereign and globally influential AI future, provided that the agreed-upon governance and ethical safeguards are implemented in a timely and collaborative manner.


Session transcriptComplete transcript of the session
Ashwini Vaishnaw

Honorable Prime Minister Sir and the leaders from the industry, captains of industry thank you all for joining us in this India AI Impact Summit. In this round table we will, I request that just like we economize with coding, please do economize with the comments, that would be great. This is a longish list about 28 CEOs here, so I’ll start with the alphabet, Mr. Sundar, so the floor is yours Did I do a cold call to you?

Sundar Pichai

You did cold call me, but I’ll be brief. First of all, thank you to Prime Minister Modi. What an extraordinary opportunity to have the AI Summit here I think we are coming here with a clear message. I believe India is going to have an extraordinary trajectory with AI, and Google wants to be a partner. It’s very clear to me India is poised to be a global AI leader, the world -class talent, deep tech expertise, and a vibrant startup ecosystem. And we will bring a full -stack commitment to India, all the way from TPUs to infrastructure investments to research and models. And the Vizag project, the AI Hub, which is a $15 billion investment, is our start.

And we are partnering. The Prime Minister had challenged us to partner across agriculture, healthcare, drive language access, and we are doing all of that. And going forward, we’ll work on skilling, working with the government. We are partnering with Indian companies and startups end -to -end to build AI -powered services. And we want to work together with you to establish partnerships. Good frameworks for AI companies. governance. Prime Minister, your vision for digital India was an inspiration, and we can see the change that’s happened in the country, and I think AI can accelerate that vision, and we are excited to work with you to build a future where everyone benefits with this transformation. Thank you.

Ashwini Vaishnaw

That was really good, really good. We’ll go to Mr. Dario from Anthropic.

Dario Amodei

Thank you again, Mr. Prime Minister. It’s clear, as I said in my remarks earlier today, that, you know, India really has a central role to play, especially in accentuating the benefits, but, you know, also in addressing the risks. You know, we come to this with unique perspective as an enterprise company. You know, we view… India not just as a source of consumers, but, you know, a place where we can work along with companies to help them do what they do better and help to augment that with AI, whether that’s, you know, business processes or distribution or the software that they’ve built or their specific understanding of the Indian market. You know, we are generally seeking to infuse AI into what these companies do and, you know, partner with them and grow together along with them and, you know, through them bring AI technology to kind of all the places in the Indian market where consumers are already being served.

We’re also, I think, you know, very interested and encourage everyone else to do things that bring, you know, unique social benefits or work with the XSTEP Foundation and Agritech and some of the other organizations. Some of the other areas. We’ve only started, you know, doing this work for a few months, but it’s already starting to bear fruit, and there’s, you know, there’s an enormous amount of benefit to be had. And then finally, you know, it seems as if it is already starting, and, you know, I encourage more of it. We want to help however we can tracking the economic impacts of AI across India, both the ways in which it accelerates growth and the shifts.

And, you know, we’re willing to help however we can. We keep our own kind of economic statistics that we publish of how people use our models. We encourage other companies to kind of do the same, and I think it’s complementary to the view that the government has. And so if we all share the different kinds of data that we have and try and understand the economic transformation as it happens, we can accentuate, you know, the good parts and mitigate any of the disruptions.

Ashwini Vaishnaw

Great. Thank you, Dario. Yours and Sam’s picture is today already viral. Thank you. We have next Nobel laureate, Sir Demis Hassabis from DeepMind, part of Google.

Demis Hassabis

Thank you, and thank you, Prime Minister Modi, for organizing this amazing summit. I’ve been so impressed with what I’ve seen in India, the energy and enthusiasm around the technology and what it can unlock, especially from the youth of the country. I think India is poised to be a real powerhouse in the AI revolution. For myself, I think we’re at the amazing threshold moment with the technology, which I always dreamed of, which is to use AI as the ultimate tool for accelerating science. And our program, AlphaFold, that solved the 50 -year grand challenge of protein folding, I think is just the first example of what I hope will be many examples over the next decade of how we can advance science and medicine with AI.

That’s our passion. I think Sundar mentioned all the investments we’re making into the industry and the India ecosystem. We see it as a key thing for us in terms of research and also the use of our products. I think we’re poised that this next step to the impact that AI is going to have is sort of hard to estimate. The way I sometimes talk about it is it’s going to be sort of 10 times the industrial revolution, the impact of the industrial revolution, but maybe at 10 times the speed happening over a decade instead of a century. So if you think about that, that’s kind of 100 times the impact of the industrial revolution, which of course was already enormous.

But I think if we get this right and these next steps, and I think India has a really critical part to play in this on the global stage, then we could unlock a new golden era of scientific discovery and acceleration to benefit everywhere around the world. Thank you.

Ashwini Vaishnaw

Thank you, Sir Hassabis. Mr. Alexander Wang from Meta. The floor is yours.

Alexander Wang

Congratulations, Prime Minister Modi, on such an incredible summit. It was so incredible to see all of the who’s who, as you mentioned, in AI, coming here along with so many of the great business leaders here and so many of the great… world leaders. We’re very excited. It’s very clear that India will be a very major player in AI and really, as Sir Dema said, one of the AI powerhouses of the future. We see it as one of the most entrepreneurial and vibrant countries in the world, and we’re so excited to see all the growth here. And at Meta, we’re excited for continued partnership as we look into this next era. We were so grateful to have you at our booth to see the Meta Ray -Ban glasses and the Be My Eyes capabilities, and so we were so grateful to have you.

We’re also so excited to support small businesses on top of WhatsApp and Facebook. We have tens of millions of small businesses in the country of India, and we’re so excited to empower them with AI and with more digital tools to enable them to grow their businesses and continue succeeding in the country. And we’re also excited to partner with the government of India to bring governance to citizens through WhatsApp as well. You know, we’re seeing this already in the state of AP. You know, we’re bringing lots of governance tools to the citizens directly through WhatsApp. I think last year over 100 million subway tickets were sold on WhatsApp. And so we’re so excited to continue partnering. I think the vision and ambition in the country is so clear in terms of digitization, modernization, and AI.

And we’re excited to continue partnering closely. Thank you.

Ashwini Vaishnaw

Great. Thank you, Alexander. We have Dr. Arthur Mensch from Mistral AI.

Arthur Mensch

Thank you, Mr. Prime Minister, for having us here. I would say I’ll be brief. I’d say the AI is going to change the economy profoundly. As an enterprise AI company, we serve the enterprise transformations. And as a consequence of that, we work with a lot of integrators in India already. we need to see that AI is going to drive multiple digits of the global GDP in the coming years and what that means is that we need to worry about two things first that everybody needs to participate in it and get access in it and in that your leadership in bringing the technology to people is actually very commendable. The other thing that is I think important is that everybody needs to take a share of the value creation that is going to occur and in that respect a risk that we may all suffer from is excessive concentration of power and excessive market concentration this may lead to excessive price and extractive economy we should beware of this concentration and what the way to actually face that is to we are betting on open source technology which is a way to create in common goods for everybody to be able to take a hold of it and to modify it and to deploy it without external control we should worry about cultural nuances and languages this is something that we have heavily invested on as initially European company and that’s something that we have been doing for Indian languages recently in audio So open source can take us to make sure that everybody gets a chance to take a share of all of the wealth that AI is going to create.

And so we are grateful for your leadership in that domain and would be very happy to be collaborating. Thank you.

Ashwini Vaishnaw

Great. Thank you. From OpenAI, Sam.

Sam Altman

Thank you, Prime Minister, for having us. As my colleagues have said, India will no doubt be a powerhouse in AI in many ways. The investment across the full stack from infrastructure to models to applications is quite something. There’s one area in particular that I hope India will lead. I think AI has to be democratized. As everyone has said, this will be a seismic shift. This will be maybe, as Demis said, ten times the size of the Industrial Revolution, ten times faster. No company or person or country is equipped to help society navigate that change. This has to be democratized. We have to put these. These tools in the hands of lots of people. Countries need their own sovereign approaches.

This will be a huge disruption to many ways of life. This will be wonderful in many other ways. There will be major things to mitigate. But this is a case where no one knows exactly what’s going to happen, and we have to figure it out iteratively. We have to deploy it step by step, integrate it in the world, and put this in the hands of people and countries and figure out the path forward together. And as the largest democracy in the world, I hope and I believe India will lead in this effort to democratize AI.

Ashwini Vaishnaw

Thank you, Sam. And as we move from the model layer to the services layer now, Exchange Chair, Ms. Julie Sweet, the floor is yours.

Julie Sweet

Thank you, and thank you very much, Prime Minister Modi, for bringing us together. I do believe that this is India’s moment because of the investments that you have made, not only in the digital infrastructure of the country, but where we see it, which is in your people. We have over 350 ,000 people here, and we are growing ourselves. And we are bringing other companies here through the Global Capability Centers as we advise our clients on their talent strategy. And top of that strategy is coming to India to access the incredible talent here. And so at a moment when the next decades will be shaped by AI, we have one of the largest AI workforces in the world here in AI.

We’re investing to train everyone. And your support of that is why we can all be positioned to lead globally because of the talent. So thank you for your investments. We look forward to continuing to partner both in bringing more companies here and in continuing to invest in our people so that we can be a part of the vision that you have so clearly laid out. Thank you.

Ashwini Vaishnaw

Thank you, June. Now we go to Adobe. Thank you, Shantanu. basically provides creative services. Thank you.

Shantanu Narayen

Namaste, Pradhan Mantriji. I first wanted to start off by saying what you talked about this morning about Manav or the human vision really resonated in terms of accountability, accessibility, and inclusion particularly. The last time we met, you had specifically asked us to focus on creativity at the WAVES conference. And so in terms of an update, I just wanted to let you know we’re deeply committed. And yesterday, in conjunction with your ministry, we announced that all of our AI products, Photoshop, Acrobat, and Firefly will be available free for students so that students can get the right skills to create the creative economy that you are interested in ensuring that India is at the forefront of. In addition to that, you said that we really need to focus on content and content provenance.

And so the watermarking that you had talked about and making sure that we take care of that is also we have now announced a content authenticity initiative. And so we’re very pleased. We’re clear that there’s a massive future for AI, and at Adobe we’re really proud to be part and to work in conjunction with your government to make that happen. Thank you.

Ashwini Vaishnaw

Thank you, Shantanu. We go to Raj from FedEx, one of the big users of AI.

Raj Subramaniam

Thank you very much, Prime Minister Modi Ji. Very nice to see you again, and thank you for putting this Global AI Summit here together. We at FedEx move the world’s high -value supply chains of the world every day, $2 trillion of commerce, and we do this with 700 planes and 200 ,000 motorized vehicles. But our mission is to make supply chains smarter for everyone. And when we move this, we also generate two petabytes of data. And AI, that’s where AI comes in, AI superpowers our data to make supply chains smarter for everyone. We have been working very heavily in India. As I mentioned to you when I met you, the cost of logistics in India from 15 % to bring it down to 8%.

And with PM Gatishakti plan as well as the national logistics policy, so we are very much involved in that. We are also investing very much in India. I was just in Navi, Mumbai yesterday. We announced the groundbreaking of our hub in Navi, Mumbai with Patnavisji and also with the Adani Group. So we are very excited to be investing in India. We are investing about 10 ,000 crores in India in this time frame and doubling in the next three years. Thank you again. Thank you very much for having us here today.

Ashwini Vaishnaw

Thank you Raj, we now move to Tokita -san from Fujitsu

Takahito Tokita

Thank you for organizing a great meeting thank you very much Fujitsu is the only one technology company in Japan that can develop supercomputers and quantum computers and we are discussing and collaborating with the Japanese government for an AI -led society this is because in order to process a huge amount of data generated from people’s lives and all economic activities a computing platform with high performance and high power saving performance is necessary I believe that there is a lot to learn from India in the transition and transformation of an AI -led society in Japan in addition there are many issues that we should work on together Thank you very much Thank you very much those initiatives will definitely be a good lesson for many other countries.

One of the challenges of an AI -driven society is data sovereignty, which requires safe and reliable data space, as well as protecting human dignity, which should also think about the dignity of data. AI should function for people and must not have human dignity. For this reason, I think it’s important not only to promote the evolution of AI, but also to ensure the safety and reliability of data at the same time. Technology evolves day by day, but I think it’s also important to continue to have the ethics that we have in our society, that the technology should have, not just following the benefits. which is also working on research and implementation of ethics that AI should have.

Thank you very much.

Ashwini Vaishnaw

Thank you, Tokita -san. Brad from Microsoft, who has been one of the leaders in AI thinking.

Brad Smith

Well, thank you, and thank you, Mr. Prime Minister, for bringing the world together. Thank you for your vision this morning. It’s certainly one that we at Microsoft embrace. As you know, we are investing here. We’re building here. We’re partnering. We’re skilling. We are devoted to India’s present and future, and we’re very bullish on the prospects for AI leadership. There’s really just one comment that I would add to the others that so many people, I think, have offered so well already. I have long believed that the Indian and American IT sectors, often do their best work when they work together. We’ve long strived to be a voice for the Indian IT sector’s needs in Washington, D .C., and we partner together around the world.

Despite the enormous success of both of our countries, we still live in a world where 70 % of the world’s GDP is elsewhere. And so, so much of our success inevitably comes not just from what we do in our own two countries, but our ability to take what we offer and take it with trust to the rest of the world. As we’re working through what is obviously, let’s just say, an interesting time in international relations, as countries need to protect their digital sovereignty, as we thrash out trade issues, I hope with your help and leadership, we can all find ways to protect all the things that need to be protected. while ensuring that technology and technology services can cross borders.

If India and the United States can be a role model for the rest of the world, then that will help us all. It will help us create jobs in India and the United States and do what we do best, bring what we create to the rest of the world as well. Thank you.

Ashwini Vaishnaw

Thank you, Brad. And now we go to Mr. Roy Jakobs, Philips, who is using AI in so many medical equipment.

Roy Jakobs

Thank you so much. And compliments for the address this morning. I think the vision you’re laying out is building on a strong tradition in which India is leading in informatics and software and now wants to charge ahead also in AI. And when you said it’s all about what we can design in India and deliver to the world, that’s something that Philips actually is already doing. Thank you for having me today. if you look to our innovation capacity half of what we deploy, 1 .7 billion of investment is in software and AI of which majority is done out of India and actually we believe we can do more also for India and we engaged successfully today with your Ministry of Health to see how we can on a data level support you on your new digital act the Eisman digital act where we can make sure that these data turn into better outcomes for the Indian public secondly we connected how we can help your primary care centers and the ASHA workers with technology they can use in an easy manner to actually for example address pregnancy challenges but also cardiac challenges in communities and therefore actually start to address these challenges early on before they become bigger and more expensive to deal with a bigger impact on the population later on we also discussed how we can help in transparent and regulated AI, because you clearly say it’s about we need to trust AI.

AI needs to support people and therefore needs to be able to do that in a trustworthy manner, which needs to be regulated, but it’s not easy as it goes so fast. So what is it that we can actually share from our learnings and how we can regulate ourselves, but also how we can regulate together so that actually brings the biggest benefit for patients, as we know that it also has side effects that you need to regulate. Last but not least, we would like to offer also to really work of unlocking big part of the data sets that you have. You have a huge amount of data in India, but they’re not that easy accessible for research and development purposes, but also even for use within country.

I think that’s another step that would unlock the full power of AI. And as a brand that is 97 years in India, seen as an Indian household brand, we believe we have a responsibility to support you, and we would really like to step up. We would like to step up with you in the field of AI for the benefit of the Indian population. Thank you so much.

Ashwini Vaishnaw

Thank you and now we go to the automobile sector Mr. Michael Johnson. I think he’s not there engine Mr. Sébastien Fabre.

Sébastien Fabre

Prime Minister thank you very much for hosting this event and for inviting us for those who don’t know us we are AI we do artificial intelligence for intelligence defense and industry. An AI that must be trusted by their users given the criticality of the application and thank you very much for attending our demo this morning just wanted to come back on one point sovereignty to me sovereignty is not isolation sovereignty is about open architecture modular architecture being able to deploy AI on sovereign infrastructure leveraging sovereign data And given the nature of our business, we had to actually embark that in our design from the start. So I just wanted to say that we are very aligned with the values you are projecting.

And at Safran, we’ve been in India for 65 years. Make it in India is in our DNA. We will double our presence by 2030. And we are very committed to support the deployment of a sovereign AI for India. Thank you.

Ashwini Vaishnaw

And as demand for memory goes up, demand for Mr. Sanjay Mehrotra goes up. Micron.

Sanjay Mehrotra

Thank you, Ashwini. Namaste, Prime Minister Modi. Really inspired by your vision, the mana vision for AI this morning. And memory and storage is very much a key enabler, essential element of AI. Simply put, if AI is the engine of the digital economy, then memory is the fuel of AI. And I’m very proud that Micron is here in India, February 28th, in your presence. We’ll be honored to have you there as we have the grand opening of the largest semiconductor single -story clean room, 500 ,000 square feet of clean room, size of 10 cricket fields. This is about steel that is used here is about three and a half times of the steel in Eiffel Tower, size of 100 Olympic pools of concrete into this.

All of this coming to fruition here with production already started there of assembly and test, advanced packaging of memory here in India. 2 ,000 team members already working on this, going toward 5 ,000 as we continue to ramp up. In just one short year of the 5 ,000, full production in that factory. 10 % of Micron’s global production will be assembled and tested here in India, and this percentage will continue to increase from here on. If you just look at what that means, it’s about multiple hundreds of millions of chips that will be assembled and packaged here using advanced semiconductor technology. This would not have been possible without the support of your government, always listening in terms of what is needed to become world -class and acting quickly.

Today, last year’s union budget codified the advanced pricing agreement. This year’s budget has strengthened it further with a two -year timeline to conclude APA negotiations. This is important for us, not just for Micron, but for the entire semiconductor ecosystem, because this provides certainty, ease of doing business, and we really appreciate all the initiatives that the government has in this regard. We look forward. to sharing your vision and working with very many of the global leaders here to meet the requirements of memory, to address the full potential of AI, and really build the future of intelligence for all together here. Thank you.

Ashwini Vaishnaw

Thank you, Sanjay. And anybody who wants to have ATMP services in India, there are eight plants which are coming online in the coming couple of years. So please do consider for your fabless companies, wherever you get them fab, have them ATMPed in India. And very soon the fab will also come. Great. We move to Cristiano from Qualcomm.

Cristiano Amon

happening in electronics and semiconductors. I think the initiatives to make a hub of electronics manufacturing as well as a semiconductor supply chain are very, very important. And as you think about the transition, we have seen that this technology is going to create a transition, is going to change many of what is assumed, you know, to be the major players. New players are going to come, and it creates an opportunity also for India to have a global role. We’re incredibly excited about this. I think the opportunity is really tremendous. We’re very thankful and privileged to be part of this. And we just, you know, we have a significant R &D that we do here in India.

And just this week, you know, the first two -nanometer chip in India has been designed by our team. I thank Mr. Vesnav to be part. And I thank Mr. Vesnav to be part of the launching. And I think we just need to keep going. This is going to be an incredible future. Thank you so much for the opportunity for me here, and thank you for the partnership.

Ashwini Vaishnaw

Thank you, Cristiano. And gentlemen, after AMD, Intel, Renaissance, I think Qualcomm is next to do the two nanometer chip design in India end to end. That is the vision our honorable prime minister has given to all of us. Make sure that the most advanced chips are designed in India, fabricated, ATMP’d and finally getting into the product in India. Thank you, Cristiano, for starting that two nanometer chip here. Now we go to Jitu from Cisco. Please do more. Make in India.

Jeetu Patel

Namaste, Pradhan Mantriji. Bharat na pragati na joyi na bhuvva janam. And, you know, Cisco’s commitment to India goes back 30 years. We’ve had over 35 ,000 employees that we are lucky to employ over here. We have manufacturing that is now happening in India, not just for India, but for export. throughout the world. And we are also investing very heavily in skilling, where we’ve actually just last year trained about 800 ,000 Indians with skills in cybersecurity and AI and networking. So it’s an honor to play a small role in the success of India. And the next 30 years we feel are going to be far more exciting than even the past 30 that we’ve had. And, you know, as you start to democratize AI, we would love to make sure that we partner with India in providing the underlying infrastructure so that that does not become a constraint for the realization of benefits in AI.

And we also want to make sure that AI can be safe and secure for the use by every citizen in India and beyond. So it’s an honor to be an Indian and born in India and actually move to America and see the level of progress that’s been made. And I thank you for… For real. the partnership between the two countries.

Ashwini Vaishnaw

Thank you, Jeetu. You didn’t touch Make in India, though. You have to commit before, Honorable Prime Minister.

Jeetu Patel

We’re very committed to Make in India. We’ve actually already had a full note and we will continue to keep investing in there.

Ashwini Vaishnaw

Thank you. We go to Cybersecurity. Mr. Matthew Prince, CloudFare.

Matthew Prince

Thank you, Honorable Prime Minister. I appreciate your remarks and especially your vision. CloudFare, the company I run, is not an AI company directly, but instead we provide the rails and the guardrails for most of the AI companies and much of the Internet. Inspired by your remarks and as a bit of a neutral provider, I would propose a framework to judge our progress at this summit and as these powerful technologies evolve. First, there should be 500 ,000 AI companies, not five. As you said in your remarks this morning, this needs to be open and for the students of India and around the world to be able to extend it, embrace it, and it’s not to be captured. Second, there needs to be a business model for journalists, content creators, and small businesses, because left to its own, AI takes but doesn’t always give back.

Third, AI should embrace and enhance our unique culture and values, not homogenize them. We shouldn’t make the same mistakes we made with the Internet, where everything goes back to the United States and the global south is sometimes left behind. Fourth, AI should be a tool for all, including students, as you mentioned, and the poorest members of the global south. And that’s something that we work very hard at Cloudflare to ensure will happen. And so Cloudflare wants to be a partner with India to realize your vision and move forward in those fundamental framework goals. We’re investing here. Building. A research facility here. And making this one of the places we will build the technologies that will power the future.

But beyond that, we’ve taken the technologies of today and made sure that they’re accessible. For example, rolling out AI for Bharat across our entire network, supporting all of the languages of India and making it available at the lowest price as possible so that anyone in India can take care of this. And Indian startups are the second largest cohort in our incubator program. And we’re, in universities, training and providing free credits to use CloudFlare’s infrastructure so that in India, the next great AI startups and companies can be built. Thank you for your leadership, and please know we are at your service. Thank you.

Ashwini Vaishnaw

And Nikesh from Palo Alto, another leading cybersecurity company.

Nikesh Arora

Namaskar, Pradhan Muthuji. As a world’s leading cybersecurity company, our mission is to deliver this AI vision safely and securely. If what Dev is saying is going to happen, that we will have ten times the industrial revolution in India, we will have ten times the industrial revolution in India, and ten times the speed. we have to deliver that in such a way that it does not cause social disruption it is a challenge, as Sam said that if you are going to put this in the hands of 1 .8 billion people you have to make sure that we’re upskilling at the same time as these jobs start transforming from what they are today in line with what Matthew said in terms of making sure there is a role in the future for all the people who rely on their creative capabilities in addition to that as we go towards this future as rapidly as we are going and we start talking about agents who are going to act autonomously there is a large question of governance and accountability who is responsible for these agents who are you going to hold responsible if something goes wrong who is going to provide the moral backbone or the nuanced requirements that agents must have just the way humans have humans have human judgment they have collaboration capability they have skills which we have to find a way of making sure it gets imbibed into these agents as we start thinking of an agentic future Last but not the least, at the speed at which we are going, where we’re seeing an arms race between the AI leaders, they are not spraying as much heat to the needs of doing this in a secure manner.

There is a challenge that AI could go rogue on us if the kill switches are not developed while we’re building AI. There is a challenge that AI could be taken over by nation states or other companies who can cause harm if this is not built in a secure and safe manner. Towards that end, we have established an AI security competence center in Bangalore. We have over 1 ,500 people, and we’re going to make sure that India becomes a center where we build these capabilities from a governance accountability, from a cybersecurity, and from a social upskilling and social impact perspective. Thank you very much, Mr. Prime Minister, for your leadership, and we look forward to working with the government on this.

Ashwini Vaishnaw

Thank you, Nikesh. We go to the Captains of Indian Industry, Mr. Mukesh Ambani.

Mukesh Ambani

Most respected Prime Minister. Thank you for this roundtable. The Manav vision that you presented at the summit this morning, I’m sure will become the AI manifesto for the world. It provides for me the moral compass for humanity in today’s uncertain time. Prime Minister, under your leadership in 2014, you gave the call for digital India. At that time, India was 138th in the world in terms of mobile and broadband connectivity. Today, we are number one in the world under your leadership. We have deployed world -class digital infrastructure for over a billion people, efficiently, inclusive and transparently. And within a single decade, scale has no longer India’s constraint. It is now India’s unique advantage. India missed the bus at the time of the first industrial revolution.

Demis and all others think that the AI revolution is 10 times bigger because India is as ready as any other country in the world to take advantage of AI revolution. With your leadership, you forced us to put all our 5G networks. So as far as all the roads for AI are ready up to the last village in India and we are absolutely ready. Four days ago, you said that you will invite the world’s data and intelligence to reside in India. With this, what we are familiar with, which is sab ka saath, sab ka vikas. And that means like everybody together and everybody’s progress has now changed to my mind to dunya ka saath, dunya ka vikas.

You have given a vision for the world. Respected Prime Minister, earlier today, as Reliance and Jio, we committed that in this opportunity, over the next seven years, starting this year, we will invest 10 lakh crores of rupees in intelligence. And this is patient, disciplined, nation -building capital designed to create durable economic value and strategic resilience for decades to come. I am grateful to you for having spent time a few days ago and encouraged all the young people who are working on this. My second commitment is that, as Reliance, we will work on the more difficult social areas of education, healthcare, agriculture, apart from, of course, consumer and enterprises, which are there, and we will prioritize them.

Our view is that AI will be accelerating not only economic growth but also job creation the way we will handle this with your leadership. Guided by your Manav -centric AI vision, it will be Jio’s endeavor to make intelligence not only democratic but affordable for every Indian. And we will partner with all the startups, with all the research institutions and the global companies present here to not only treat India as a consumption market but to treat India as an innovation market where we can develop and make an India and serve India first and then serve the rest of the world. Thank you, Prime Minister, for all your leadership and inspiration. Thank you.

Ashwini Vaishnaw

Thank you. And Mr. Natarajan Chandrasekaran, Chairman of Tata Group, next to you.

Natarajan Chandrasekaran

Honorable Prime Minister, firstly, I would like to congratulate you for hosting this AI Impact Summit at such an aspirational scale and the convening power that India has displayed in getting all of us together. Thank you so much. The second point I want to make is while all the AI companies, technology companies, industrial companies and all of us will do everything that we need to do to advance AI and also make AI work for the society, the individuals and businesses. Your role, your leadership, which you exhibited this morning. In defining what AI should do. for responsibly making progress. First, you exhibited enormous confidence in the positive power of AI. And you also exhibited what the potential of AI could be.

And you clearly outlined what it means to have a human central leadership, a open, shared, collaborative leadership. This is setting the agenda for all countries and all businesses and societies. Thank you for that leadership. The second point I want to make is AI has two important characteristics, like many other characteristics. One is ambition. And by ambition I mean it helps those who are ambitious to achieve that ambition. Second one is skill. Under your leadership, India has the same two qualities, ambition and wanting to scale. I think with these two coming together, we are naturally well positioned to lead in the AI world for the good of the world. Third, Honourable Prime Minister, on behalf of the Tata Group, I want to say that we will do everything that we need to do in three important areas.

One is scaling the nation. As he said in this morning, it is about ensuring that every citizen is empowered. Second, we will build the AI infrastructure across all the layers from the hardware layer all the way to chips and data centres and agent TKI and data, etc. to make an impact. Third, we will build the AI infrastructure across all the layers from the hardware layer all the way to chips and data centres and agent TKI and data, etc. to businesses around the world and also help in India’s journey in social transformation. Thank you again for the opportunity and your vision and leadership. Thank you. We have

Ashwini Vaishnaw

Mr. Mittal from Airtel.

Sunil Bharti Mittal

Honorable Prime Minister, at the outset, let me congratulate you for an amazingly successful AI Impact Summit. India, and we are all as Indians fortunate to have you as a leader who understands the power of technology. I have watched you for long years, and the way you have guided this country through the power of technology is absolutely incredible. You directed some of us to launch 5G networks in the shortest possible time frame. I still recall your words, I want it launched in 10 months. And your inspirational leadership made it possible for us to make it happen. 5G across every square kilometer of this country. You knew that this country will need a bedrock of stability, a very strong network on which the fanciful new wave of technologies will ride.

Today, connectivity through fiber, submarine cables, data centers, towers all across the country is in place, Honorable Prime Minister. And you should be very proud that India today leads that race by putting a smartphone, a computer in the hands of every citizen of this country at $2 a month for unlimited amount of data that they use. We will keep on nourishing these arteries. We’ll keep on building the muscles of the data centers, fiber. And we will ensure that this wonderful invention of AI are brought to our billion customers in the country at the most cheapest frugal way. You have assembled an amazing set of people here in this room today. The world’s entire AI technology industry is sitting here.

they should note that India could do a moon mission, the Gaganyaan, with $74 million against the U .S. spending over $92 billion to do a similar moon landing. And India did the moon landing on the difficult side of the moon highly successful. The world leaders can pick up frugal innovation from our country and use the very large base that we offer the customer for their own benefit. Together, Honorable Prime Minister, your vision will resonate with the entire globe of making AI democratic, AI available to all for the benefit of humanity. India is connected to the destiny of India. Thank you, sir.

Ashwini Vaishnaw

Thank you, Mr.Mittal. And Mr. Nandan Nilakani, Chairman of Infosys, we would like to hear from you what you are doing for this transition.

Nandan Nilekani

No, I’ll talk about something. Prime Minister Modi, first of all, congratulations. Congratulations on a fantastic summit. It’s really been great. I want to talk about AI diffusion in India with an example. When I met Prime Minister Modi on 8th January and talked about applying AI to farmers, he said, why can’t we apply it to cows and cattle? Because if the cow is sick, it can’t tell you that it is sick. How can you solve this problem? And he gave us his vision of applying AI to agriculture and dairy. The same day, the PMO had a meeting along with Mr. Krishnan and Abhishek of METI with Amul and with some of my colleagues. And within three weeks, the application went live.

The meeting was on Jan 8th. It went live on February 11th. This is the world’s largest cooperative with 3 .6 million farmers, 2 billion milk transactions per year. and 40 million cattle. And today, all these farmers, a large number of them being women, use the Sarla Ben application and actually get real responses about their cattle, their problems, their pregnancy, their milk production, and so on. Look at the speed of diffusion. An idea that the PM had on Jan 8 has become reality on Feb 11. This, to me, is an example of the speed of execution of AI diffusion in India. And I get the same sense of excitement I got on December 30, 2016, when the PM launched the Bheem payment application on UPI.

And when he launched that application, he started something where today we have 21 billion transactions a month with 500 million users and the world’s largest payment system. I feel the same sense that AI is at that point. And with his leadership and vision of how AI can be used for the benefit of Indians, common man, farmers, workers, and so on, I think it’s going to really take off. Moreover, this is also designed for sovereign data. The data of Amul remains with Amul, and design is to make sure that it’s within India. So it’s really a great example, and I think by the time many of these people come back in a few months, there’ll be many, many more applications, and India will lead the world on showing how AI diffusion can matter to improve the lives of common man, farmers, students, patients, and so on.

And I think this is where the AI road, the race to the top is going to happen, and thank you for your leadership, sir. Thank you, Dandan.

Ashwini Vaishnaw

We have a respected industry leader from Sweden, Mr. Marcus Wallenberg.

Marcus Wallenberg

Honorable Prime Minister, I want to congratulate you on amassing such force behind the AI initiative. it is amazing to see what your inspirational leadership can do and it reminds me very much of when you launched Make in India you got a similar very positive wave behind you so I represent a number of companies who are active in India since many many years over 100 years actually in the case of Ericsson and I believe that this idea about pushing the AI initiative will be very very supportive for future business of these companies in India over the years because you have through many of your IT companies provided support all over the world for digitization and you have a lot of support from the IT companies but now when you’re moving into AI I think that a lot of the investments that already have been carried through the companies that we are close to, like ABB and AstraZeneca, et cetera, et cetera, will be even further enhanced.

And I think this will position, your initiative here will position India as an even better place for investment and development of their businesses. So thank you. And I also want to say that on behalf of Mr. Johansson, who was called upon earlier but not here, Saab is not a car company, it’s a defense company, and it’s also operating here. Very thankful for the good reception from the Indian government and continue to develop their efforts and products and services within the AI field and hope to serve the Indian government going forward. Thank you. Thank you.

Ashwini Vaishnaw

Thank you. My apologies for having that car company thing. We now go to Mr. Rishi Sunak, who’s currently leading the enrichment project, and the former Prime Minister of UK. Thank you, Minister.

Rishi Sunak

When I hosted the first AI summit in Bletchley Park a few years ago, I never imagined that the journey would take us to this incredible event here in New Delhi. And I think that’s testament to your leadership, Prime Minister Modi Ji, but also the vibrancy of the AI ecosystem that we’ve been hearing about, and of course the incredible optimism and energy of the Indian people that’s been on full display these past few days. And it’s that magical combination that I think is why, Minister, this morning you were able to remind us all that according to Stanford University’s authoritative global index, index, India has now been ranked a leading global AI superpower. What you were too polite to say this morning, to spare Demis and my blushes, is that you leapfrogged the UK to get into that position.

Although I did check, and I can gently point out that England remains just ahead of India in the ICC test rankings, so not all is lost. The special and distinctive thing about these summits is that they bring together not just leaders from government, but also all of you, leaders from industry. And when I did that originally, it was in recognition of the fact that this technology is largely being developed in the private sector. And to put it in context, just this year, the companies represented in this room will spend 20 times more to develop this technology than the United States spent on the entire. Manhattan Project. So all of you in this room, I occupy a very significant role.

in what is about to happen. And I know many of you personally, and I know all of you are driven by values, and you take that privilege position seriously and recognise that it comes with responsibilities. So maybe because, Minister, I can’t offer to invest a billion dollars to make something in India, I can instead actually join Prime Minister Modi to request something of all of you in this room, and that’s to think about two particular responsibilities that the Prime Minister outlined this morning in his manner of vision. And the first is to develop this technology responsibly, safely and securely. I still remember the meeting I had with Demis, Dario and Sam all together in Downing Street where we spoke through these issues a few years ago, and I think they would be the first to say that the pace of acceleration in this technology means that those conversations we have are even more relevant today.

And what I’d ask is that the Prime Minister that we maintain the transparency. the dialogue between government and all of your companies, engagement, and then at the appropriate moment, talk about governance as it’s required. We must maintain the trust and confidence of our citizens as this technology develops. And remember that our first duty as elected leaders is to ensure their safety. And then the second responsibility I’d ask of all of you is to ensure that this technology benefits everyone, everywhere. And we had a very compelling vision of that this morning. We talk a lot about AI raising the ceiling, but we need to make sure that it also lifts the floor for humanity. And for me in particular, I think that means in health care and education, because those are the two foundational things that give all of us dignity and opportunity.

And AI has the potential to be, I think, both the most uplifting, democratizing, and positive force that any of us have all known. Bye. it’s up to everyone in this room to make sure that that becomes a reality and deliver on the vision that the Prime Minister set out this morning. Thank you.

Ashwini Vaishnaw

Thank you so much, Rishi. We go to Mr. Enrico Bagnasco from Sparkle.

Enrico Bagnasco

Thank you and a special thank to you, Prime Minister Modi, for the vision you shared this morning and for this event. SPACL is one of the main international telecom carriers connecting currently 40 countries worldwide. We design, deploy and operate long distance submarine optical cables. So we deploy one of the key enabling infrastructure to support digital services and to support the AI revolution. We have been operating in India for 28 years out of our Mumbai offices. and work with great success to all the key Indian carriers represented here, by the way. And we’re now a fully licensed operator, so we plan to continue to remain, invest, and deploy our facilities here. We are currently deploying together with Google the Blue Raman subsea cable, which will connect with a new infrastructure, Milano, Italy, with Mumbai, India, through a new diversified route.

So thank you again for your vision. We’re here to stay and to support your deployment. Thank you very much.

Ashwini Vaishnaw

Thank you so much. Now Mr. Giordano Albertazzi from Vertiv.

Giordano Albertazzi

resolve was there before these meetings, hearing directly from you your vision, now the resolve is even stronger. So, count on Vertiv, just like many of our customers that sit around this table, can count on us for an expansion of our manufacturing, our engineering, our service presence in India. So, I couldn’t be more thrilled to be here today and certainly to be part of this adventure in India and globally. A big thank you. Thank you.

Ashwini Vaishnaw

We have His Excellency, Mr. Mansour Ibrahim Al -Mansouri from G42. The floor is yours.

Mansour Ibrahim Al Mansouri

Thank you, Prime Minister. It’s an honor to be here, and under your leadership, you have elevated technology from a sector to a nation, building infrastructure. Thank you. and you are building a foundation for sovereign intelligence and economic growth. And we in Abu Dhabi, we share the same convictions, where AI is no longer a productivity tool but a core infrastructure. And our company, G42, the national champion through its business units, is delivering two large factories of the future. First, a token factory to serve intelligence at scale, and second, agent factory to empower enterprise at scale. And this is not done in isolation, but with strategic partners, many of whom are in this room today. We are working in a great collaboration.

And Prime Minister, we believe that nations should always build the strongest intelligence infrastructure and cross -border partnership that will define the next century of economic growth. And the UAE believes India is a partner, and let us build. Let us build this partnership and lead the future together. Thank you.

Ashwini Vaishnaw

We now come to the investment category. Hemant from General Catalyst, please do share your investment plans.

Hemant Taneja

Shree, Prime Minister Modiji, congratulations on an amazing conference, and thank you for your leadership. I think focusing on trying to align AI with human centricity is something we need as a global perspective, and we need a lot more of it. This is India’s moment to lead. So when I think about the last 15 years, the work India did in leapfrogging in identity in financial payments with Azhara and UPI, I think India’s got a similar opportunity to do in the applied AI layer. And so… And the way India can lead is with one mindset of abundance. How does AI really empower everybody? in health care, in education, in a way that you can uplift the lives of not only the billion and a half people in India, but do that worldwide.

And I have a deep belief that the entrepreneurial ecosystem in India is going to deliver some incredible global leaders that are focusing on this problem. One of the things that’s really top of mind for me is India’s got one of the youngest demographics entering the workforce, and there’s an opportunity to lead in empowering that workforce with AI rather than resisting the diffusion of AI and the jobs created with it as a source of eliminating opportunity for them. If everybody had the productivity of AI in the Indian workforce, it would no doubt create an amazing country and an amazing opportunity for the rest of the world. We’re deeply believing in the entrepreneurial ecosystem. We’ve been investing in some incredible companies.

We’ve been investing in companies over the last few years. and we’re heavily doubling down on our investment, and we’ve agreed to invest about $5 billion in the next five years in the Indian entrepreneurial ecosystem. Thrilled to be doing that, and thank you for creating the conditions for that opportunity.

Ashwini Vaishnaw

Thank you, Mr. Taneja, for the $5 billion pledge that you have taken. Mr. Vinod Khosla, one of the most respected persons from the IT community.

Vinod Khosla

Thank you, and namaste. I’m most excited by what AI can do to not only help me meet the India 2047 vision, but far exceed it. I believe in the real era of abundance by them, but AI adoption needs the permission of the people. Capitalism, in my view, which is a great tool for progress, is by permission of democracy and the votes of the people. So because of that, I feel the first thing we should do and aggressively do is bring AI in services to the people. And my favorite few services is to add free AI tutors for every child in the country, all 250 million of them, free AI doctors for every one of its citizens, and free agronomists for all the small farmers that are so important, such an important part of the vote bank and the permission we need to apply AI.

It is very clear to me that the 2030s will be a chaotic era. There will be disruption. There will be large changes. And so before we get to that stage and people are tolerant about the progress and the uncertainty that it creates at the same time, that we should really have every single Indian benefiting from these services so they understand the power that AI can bring. So these, in my view, should be part of Aadhaar. Just like UPI is part of Aadhaar, AI doctors should be an Aadhaar service, AI tutors should be an Aadhaar service, and AI agronomists should be Aadhaar services. So the hard part in Aadhaar is done already, and I think this can bring real benefit to mass and large number of Indians and have them appreciate.

What AI can do. Having said that. On the investment side, India is just a wonderful area for us. We were one of the early investors in Sarvam as the sovereign AI model, something we believed in for a long time. Emergent, which is one of the fastest growing companies anywhere I’ve seen, has grown to 100 million in the last eight months. So India has great talent. We have healthcare companies, two of them making very radical innovations in how healthcare is applied. The way to make more money in healthcare is not to do more surgeries or more procedures. It is to have a model that’s aligned with consumers and the cost of providing service, which is generally the paying entity.

Wuhan, another one, has $40 million. Indian workers providing people like Swiggy and… Flipkart and others with workers. They’re hiring half a million workers a month now out of their 40 million database and AI is absolutely essential to talk to this many people about what job is the right fit for them. So thank you. I appreciate the time you’re spending and making sure that the ecosystem is rich. But I do think we should start with benefits for the people in these essential services. Thank you,

Ashwini Vaishnaw

Mr. Khosla. Lightspeed is very active here in India in the tech space. Ravi, your turn.

Ravi Mhatre

Hi, thank you, Prime Minister Modi, for organizing this really amazing global AI event that focused both on how AI can and benefit India but also the rest of the world. It’s an honor to be here. Lightspeed has invested in Indian technology startups for the past 17 years. We’ve committed close to a billion dollars in that time and have plans to significantly increase our investments as we move into the sort of AI generation of technology. We also, along with my colleague Vinod, are large investors in Sarvam, which is providing sovereign AI capabilities to India and has worked closely with your administration, as well as several other emerging artificial intelligence startups. And we believe there’s, in India, a dual opportunity as we move into the AI era.

The first, India represents the largest single AI consumption market in the democratic world. India’s opportunity is to build real -world applications at real -world scale, and that’s… essentially unmatched. We think that the Indian economy also has great breadth, including agriculture, manufacturing services, healthcare, education, financial inclusion, and government at scale. These are all places that create new opportunities for artificial intelligence innovation. The second major opportunity, we believe, for technology innovation where venture capital funding can accelerate innovation is India’s ability to provide technology talent density as one of its greatest exports for AI. We’ve already seen some of the largest AI developer communities form in India, and they are continuing to grow. And they’ve been showing the ability to build.

They’re transformative products which can serve both India and the rest of the world. Thank you.

Ashwini Vaishnaw

Thank you. And with this, we have covered practically all the layers of the AI stack. We have covered models. We have covered services, infra, compute, from funding to overall the conventional, the machine makers, the industry leaders. Thank you all for sharing your views. Now I’ll request Honorable Prime Minister Shendarendra Modiji, sir, kindly for your views and your guidance to us. Thank you.

Shri Narendra Modi

Friends, I understand that in governance, I have been working for many years. I once called Shikha. I said, you teach this to the people of my defense. So Shikha has, I had a daughter with her. So my effort is, I have met almost everyone. I have met everyone from time to time. I have spoken because I have, and whenever I have met you, I have tried to learn, to understand, to understand. My effort is that we should together, like a traveler, like a co -traveler, keep these targets and these directions in mind and achieve our goal. We should definitely try in that direction. And I have faith. that globally also, this talent in that talent pool can also be associated with India.

I have told Sanjay many times, Sanjay Shahi will be the most patent owner here. So I said to Sanjay, you should also this patent world, give them this kind of habit from lab. They are also understanding, they are doing the work. So I have disappointed whoever I told you. You have given time, you have contributed your experience, and because of that you must have seen that whenever I have to meet, I keep giving you some new work. And you give me after doing it. And the work that you do, the habit of making it work a little more is of a human. So it is also mine, that when you do this, then you make it work more.

This will be my profession. but I assure you that there will be a relationship of policies and whatever changes we have to make we are ready to take it down the ground work we have to do the government is ready and I am not working from a limited angle I believe that this is the way of humanity and that is why I have to give it strength the coming generations will be blessed and that is why I said this morning there is a group of people who see fear in this there is a group of people who see the future in this I represent that group of people who see the future in this and I am not saying that some people have fear and some people have fortune I am not a person of fear I am a person of faith in fortune and that is why I believe that the path we are on and the faith we are on we will get the right results I will once again you all took out time you gave valuable thoughts and commitment of cooperation commitment of moving forward so this is our partnership it is a different situation and for global good for a good for a good cause so I am sure that we will get the right results I am very grateful to you.

Thank you very much.

Related ResourcesKnowledge base sources related to the discussion topics (12)
Factual NotesClaims verified against the Diplo knowledge base (4)
Confirmedhigh

“Sundar Pichai thanked Prime Minister Modi and said India is poised to be a global AI leader because of its talent and ecosystem”

The knowledge base records Sundar Pichai opening his keynote by thanking Prime Minister Modi and expressing confidence that India will have an extraordinary AI trajectory, highlighting its talent and ecosystem [S5] and [S13].

Confirmedhigh

“Dario Amodei of Anthropic said India must play a central role in amplifying AI benefits and managing its risks”

Anthropic’s statements confirm Amodei’s view that India can be a partner and leader in addressing AI security and economic risks, emphasizing a central role for the country [S144].

Additional Contextmedium

“Google’s commitment will span TPUs, cloud infrastructure, research collaborations and partnerships across agriculture, healthcare and language access”

While the transcript notes Google’s broad partnership ambitions with India, the knowledge base does not detail specific technology components such as TPUs or sector-specific collaborations, providing only general confirmation of a partnership intent [S13].

Additional Contextmedium

“Anthropic is seeking deeper AI cooperation with India and sees the country as a partner in guiding global AI risk responses”

Anthropic’s public statements elaborate on its desire for deeper cooperation with India, reinforcing the claim of a partnership focus and adding nuance about specific collaborative goals [S144].

External Sources (146)
S1
Invest India Fireside Chat — Very good afternoon, everyone. I’m truly honored to run a fireside chat with Mr. Vinod Khosla. And throughout my Intel j…
S2
Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session — Thank you, Mr. Taneja, for the $5 billion pledge that you have taken. Mr. Vinod Khosla, one of the most respected person…
S3
Keynote-Demis Hassabis — -Demis Hassabis: Role – Co-founder and CEO of Google DeepMind; Titles – Sir, Nobel laureate; Areas of expertise – Artifi…
S4
Folding Science / DAVOS 2025 — This discussion focused on the intersection of artificial intelligence (AI) and biology, particularly in the context of …
S5
Keynote-Sundar Pichai — -Moderator: Role/Title: Event Moderator; Area of Expertise: Not mentioned -Mr. Dario Amote: Role/Title: Not mentioned; …
S6
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Ananya Birla Birla AI Labs — -Sundar Pichai: Role/Title: Not specified in transcript; Area of expertise: Technology (implied)
S7
State of Play: AI Governance / DAVOS 2025 — – Arthur Mensch: Co-founder and Chief Executive Officer, Mistral Arthur Mensch: I’m suggesting that this is the direct…
S8
The Role of Government and Innovators in Citizen-Centric AI — – Arthur Mensch- Jarek Kutylowski – Arthur Mensch- Roberto Viola
S9
Keynote by Marcus Wallenberg Chairman SEB & Saab — – Marcus Wallenberg: No specific title mentioned in the transcript, but appears to be a business leader with extensive e…
S10
Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session — -Marcus Wallenberg- Representative from Sweden (representing multiple companies including Ericsson, ABB, AstraZeneca)
S11
Keynote by Marcus Wallenberg Chairman SEB & Saab — India, on the other hand, as I see it or as I perceive it, has not gone primarily the R&D route, but primarily the way t…
S13
Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session — Speakers:Rishi Sunak, Mukesh Ambani, Shantanu Narayen, Natarajan Chandrasekaran, Shri Narendra Modi
S14
S15
Sparkle, University of Genoa, and SubOptic launch submarine communications program — Sparkle, the University of Genoa, and the SubOptic Foundation haveforgeda landmark partnership to advance education and …
S16
Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session — Thank you so much, Rishi. We go to Mr. Enrico Bagnasco from Sparkle.
S17
Technology in the World / Davos 2025 — – Dario Amodei: CEO of Anthropic Dario Amodei: I think both. I’m worried, I think, both about kind of the internatio…
S18
Keynote-Rishad Premji — -Mr. Dario Amote: Role/Title: Not specified; Area of expertise: Artificial intelligence (described as pioneer and though…
S19
Davos 2026 reveals competing visions for AI — AIhas dominateddebates at Davos 2026, matching traditional concerns such as geopolitics and global trade while prompting…
S20
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Jeetu Patel President and Chief Product Officer Cisco Inc — -Speaker: No specific role, title, or area of expertise mentioned in the transcript For closing the context gap, Patel …
S21
Partnering on American AI Exports Powering the Future India AI Impact Summit 2026 — Note: The transcript appears to conclude mid-sentence with the introduction of Jeetu Patel from Cisco, suggesting additi…
S22
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Jeetu Patel President and Chief Product Officer Cisco Inc — Impact:This analogy deepened the technical discussion by making the context problem relatable and urgent. It shifted the…
S23
Keynote-Rishi Sunak — -Moderator: Event moderator introducing speakers and facilitating the discussion Building Public Trust Through Implemen…
S24
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Ananya Birla Birla AI Labs — -Rishi Sunak: Role/Title: Not specified in transcript; Area of expertise: Not specified
S26
Keynote-Julie Sweet — -Moderator: Role/Title: Not specified, Area of expertise: Not specified Addressing the SME Challenge for Inclusive Grow…
S27
Industries in the Intelligent Age / DAVOS 2025 — – Julie Sweet – CEO of Accenture 2. HR Transformation: Julie Sweet argued that HR departments need to be reinvented for…
S28
Keynote-Julie Sweet — Overall Tone:The tone is consistently optimistic and forward-looking throughout, with Sweet maintaining confidence in AI…
S29
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Hemant Taneja General Catalyst — -Hemant Taneja: CEO of General Catalyst (venture capital firm), advocate for responsible innovation, focuses on bridging…
S30
Sticking with Start-ups / DAVOS 2025 — – Hemant Taneja: Chief Executive Officer and Managing Director at General Catalyst USA Hemant Taneja and Mohit Bhatnaga…
S31
S32
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Giordano Albertazzi — 1705 words | 123 words per minute | Duration: 830 secondss Human intelligence happen in the brain. But the brain doesn’…
S34
Keynote-Brad Smith — -Brad Smith: Role/Title: Vice Chair and President of Microsoft; Areas of expertise: Technology policy, privacy, cybersec…
S35
Brad Smith — As Microsoft’s vice chair and president, Brad Smith leads a team of more than 1,900 business, legal and corporate affair…
S36
A Digital Future for All (morning sessions) — – Brad Smith – Vice Chair and President, Microsoft Brad Smith: Well, I know the time is running out. Let me be brief,…
S37
Cracking the Code of Digital Health / DAVOS 2025 — – Roy Jakobs: President and Chief Executive Officer, Royal Philips 1. Systems Approach: Roy Jakobs emphasized the need …
S38
Keynote-Roy Jakobs — The discussion features Roy Jakobs, CEO of Philips, presenting his vision for artificial intelligence’s transformative r…
S39
Scaling AI Beyond Pilots: A World Economic Forum Panel Discussion — -Roy Jakobs: President and Chief Executive Officer of Royal Philips (Healthcare/Medical Technology) Roy Jakobs, Ryan Mc…
S40
Keynote Adresses at India AI Impact Summit 2026 — -Sanjay Mehrotra- CEO of Micron Technology And so we are here to listen to our distinguished guests as they present the…
S41
Keynote Adresses at India AI Impact Summit 2026 — Ladies and gentlemen, and now the Pax Silica Declaration is being signed between India and the United States of America….
S42
Partnering on American AI Exports Powering the Future India AI Impact Summit 2026 — Agreed with:Ambassador Sergio Gor, Secretary S. Krishnan, Sanjay Mehrotra, Dr. Randhir Thakur — U.S.-India partnership i…
S43
https://app.faicon.ai/ai-impact-summit-2026/keynote-nikesh-arora — It was a pleasure listening to you and all your wonderful ideas, giving solutions, AI solutions to India’s problems. And…
S45
Keynote-Nikesh Arora — At the current moment, the balance is tilted not in the favor of trust, inclusion, and security. It’s actually tilted in…
S46
Keynote-Rajesh Subramanian — -Frederick W. Smith: Role/Title: Founder of FedEx; Area of expertise: Not specified in current context (referenced by Ra…
S48
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Ananya Birla Birla AI Labs — -Mukesh Ambani: Role/Title: Business leader; Area of expertise: Business, industry
S49
S50
Keynote-Mukesh Dhirubhai Ambani — Ambani framed artificial intelligence as the cornerstone of India’s transformation into a fully developed nation by 2047…
S51
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Cristiano Amon — -Announcer: Role/Title: Event announcer/moderator; Areas of expertise: Not mentioned And Mr. Amon is leading Qualcomm’s…
S52
Lift-off for Tech Interdependence? / DAVOS 2025 — – Cristiano Amon: President and CEO at Qualcomm Cristiano Amon: What I’ll say is, technology is moving very, very fast…
S53
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Cristiano Amon — This discussion features Cristiano Amon, President and CEO of Qualcomm, presenting his vision for the next chapter of ar…
S54
Keynote-Rishad Premji — -Mr. Nandan Nilekani: Role/Title: Not specified; Area of expertise: Artificial intelligence (described as pioneer and th…
S55
High Level Session 2: Digital Public Goods and Global Digital Cooperation — – **Nandan Nilekani** – Co-founder and chairman of Infosys Technologies Limited (participated online) Nandan Nilekani, …
S56
https://dig.watch/event/india-ai-impact-summit-2026/fireside-conversation-01 — Thank you so much, Mr. Sikka, for your profound and very interesting remarks. And of course, your work at VNI also exemp…
S58
https://dig.watch/event/india-ai-impact-summit-2026/leaders-plenary-global-vision-for-ai-impact-and-governance-afternoon-session — We have His Excellency, Mr. Mansour Ibrahim Al -Mansouri from G42. The floor is yours.
S59
Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session — -Mansour Ibrahim Al Mansouri- His Excellency, G42 (UAE)
S60
S61
Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session — Thank you and now we go to the automobile sector Mr. Michael Johnson. I think he’s not there engine Mr. Sébastien Fabre….
S62
Keynote-Alexandr Wang — -Moderator: Role involves introducing speakers and facilitating the discussion A belief that anything is possible, and …
S64
Keynote-Alexandr Wang — The moderator introduces Alexander Wang by highlighting his impressive credentials and current role. This introduction e…
S65
Keynote-Sam Altman — -Moderator: Role/Title: Event moderator; Area of expertise: Not mentioned -Sam Altman: Role/Title: CEO of OpenAI; Area …
S66
Oversight of AI: Hearing of the US Senate Judiciary Subcommitee — 10“GPT-4 Is OpenAI’s Most Advanced System, Producing Safer and More Useful Responses.” OpenAI, https://openai.com/produc…
S67
Driving U.S. Innovation in Artificial Intelligence — 26. Sam Altman – CEO, OpenAI
S68
Open Internet Inclusive AI Unlocking Innovation for All — Very few individuals have done more to bring revolutionary and transformative technology into the hands of millions than…
S69
https://app.faicon.ai/ai-impact-summit-2026/open-internet-inclusive-ai-unlocking-innovation-for-all — Very few individuals have done more to bring revolutionary and transformative technology into the hands of millions than…
S70
Defending the Cyber Frontlines / Davos 2025 — – Matthew Prince: CEO of Cloudflare Matthew Prince: Absolutely. So Cloudflare’s, our mission is to help build a bette…
S71
THEBROADBAND BRIDGE — Sunil Bharti Mittal, Founder, Chairman and Group CEO, Bharti Enterprises
S72
AcknoWleDGment — – Mr Sunil Bharti Mittal, Mr Himar Arjun Singh, Bharti Enterprises – Mr Börje Ekhlom, Ms Elaine Weidman, Ericsson – Mr A…
S73
S74
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Takahito Tokita Fujitsu — -Announcer: Role as event announcer/host, expertise/title not mentioned -Vivek Mahajan: CTO (Chief Technology Officer) …
S75
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Takahito Tokita Fujitsu — 676 words | 101 words per minute | Duration: 400 secondss AI must not be a force that replaces people or becomes a thre…
S76
S77
Keynote-N Chandrasekaran — 972 words | 112 words per minute | Duration: 519 secondss AI is nothing artificial, it is real. Because it learns from …
S78
Keynote-N Chandrasekaran — Natarajan Chandrasekaran
S79
Announcement of New Delhi Frontier AI Commitments — -Shri Narendra Modi: Role/Title: Honorable Prime Minister of India, Area of expertise: Not specified
S80
Keynote-N Chandrasekaran — -Sri Narendra Modi ji: Prime Minister of India (referred to as “Honourable Prime Minister”) Honourable Prime Minister, …
S81
Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session — -Shri Narendra Modi- Honorable Prime Minister of India
S82
AI-Powered Chips and Skills Shaping Indias Next-Gen Workforce — -Ashwini Vaishnaw- Role/Title: Honorable Minister (appears to be instrumental in India’s semiconductor industry developm…
S83
Announcement of New Delhi Frontier AI Commitments — -Shri Ashwini Vaishnaw: Role/Title: Honorable Minister for Electronics and Information Technology, Area of expertise: El…
S84
AI and Global Power Dynamics: A Comprehensive Analysis of Economic Transformation and Geopolitical Implications — -Ashwini Vaishnaw- Minister for Economic Electronics and Information Technology of India
S85
© 2019, United Nations — Developments in AI, including machine learning, are enabled by the large amounts of digital data that can be a…
S86
From Innovation to Impact_ Bringing AI to the Public — So I see it not as a job reduction. I see it as opportunity for India to create a global AI-dominant nation… So a pers…
S87
Artificial intelligence (AI) – UN Security Council — The discussion highlighted that open-source models enable a wide range of entities, from startups to larger corporations…
S88
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Matthew Prince Cloudflare — Prince argues that AI should be a widely distributed capability rather than being monopolized by a few large firms. He l…
S89
Keynote-Demis Hassabis — This discussion features a keynote address by Sir Demis Hassabis, co-founder and CEO of Google DeepMind and Nobel laurea…
S90
Keynote-Sundar Pichai — In this comprehensive keynote address delivered in India, Sundar Pichai, CEO of Alphabet and Google, opened with “Namast…
S91
Skilling and Education in AI — And then the data that I’m submitting into the system, simply by interacting with AI, I’m submitting data and providing …
S92
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Hemant Taneja General Catalyst — Taneja argues for universal AI empowerment of India’s massive workforce influx, suggesting that equipping each new worke…
S93
AI Innovation in India — Bagla articulated a compelling vision of India’s unique advantages in the global AI landscape, asserting that India will…
S94
Keynote-Mukesh Dhirubhai Ambani — Ambani makes a bold prediction about India’s future position in global AI leadership. He expresses confidence that India…
S95
The Global Power Shift India’s Rise in AI & Semiconductors — So the goal of Genesis Project is to really, one, align public and private partnership, two, invest government resources…
S96
Welfare for All Ensuring Equitable AI in the Worlds Democracies — -Democratizing AI Access and Preventing Digital Divide: Concerns about AI’s economic value concentrating in Western econ…
S97
Welfare for All Ensuring Equitable AI in the Worlds Democracies — Evidence:Examples include impact on farmers, small schools, NGOs, and small hospitals. Google’s mission is to keep every…
S98
Keynote-Dario Amodei — Discussion point:Transition Management
S99
The strategic imperative of open source AI — Meta’s Chief AI Scientist, Yann LeCun, captured this shift clearly. Responding to those who see DeepSeek’s rise as ‘Chin…
S100
WS #145 Revitalizing Trust: Harnessing AI for Responsible Governance — The level of consensus among the speakers was relatively high, particularly on the benefits and potential applications o…
S101
DeepSeek: Some trade-related aspects of the breakthrough  — Although so far proprietary models have predominated in the market, open source has been gaining traction, as noted by Y…
S102
To share or not to share: the dilemma of open source vs. proprietary Large Language Models — Bilel Jamoussi:Great. Thank you. Google has a history of both open source contributions and proprietary developments. Ar…
S103
Driving Indias AI Future Growth Innovation and Impact — Summary:The main areas of disagreement center around regulatory approach (light-touch vs. balanced frameworks), implemen…
S104
The Innovation Beneath AI: The US-India Partnership powering the AI Era — Moderate disagreement level with significant implications for investment strategies, infrastructure development prioriti…
S105
Smaller Footprint Bigger Impact Building Sustainable AI for the Future — Abhishek Singh from India’s AI mission articulated a distinctly pragmatic approach, explicitly stating that India is “no…
S106
Keynote-Demis Hassabis — This discussion features a keynote address by Sir Demis Hassabis, co-founder and CEO of Google DeepMind and Nobel laurea…
S107
Keynote-Demis Hassabis — Anticipated arrival and impact of AGI Artificial intelligence He quantifies the transformative power of AGI, suggestin…
S108
Policy Network on Artificial Intelligence | IGF 2023 — This education should be accessible to all, regardless of their age or background. Additionally, the panel discussion sh…
S109
Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session — Arguments:AI has to be democratized. This has to be democratized. We have to put these tools in the hands of lots of peo…
S110
Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session — AI has to be democratized. This has to be democratized. We have to put these tools in the hands of lots of people
S111
Secure Finance Risk-Based AI Policy for the Banking Sector — The moderator emphasizes that AI governance should not be viewed through a completely different lens but should be integ…
S112
Open Forum #53 AI for Sustainable Development Country Insights and Strategies — Success of Aadhaar, UPI, and data layer implementations that enabled various sector applications to be built on top The…
S113
Panel Discussion AI in Digital Public Infrastructure (DPI) India AI Impact Summit — Explanation:There’s unexpected consensus that successful private sector AI innovation actually depends heavily on public…
S114
National Strategy for Artificial Intelligence — Artificial intelligence in the public sector can contribute to: Such assessments related to the use of AI in public adm…
S115
Global Perspectives on Openness and Trust in AI — “which is why in France and in Europe we’re very much in favor of open source as a competitive tool and as a way to leve…
S116
New study suggests leading AI models fall short of responsible AI standards set by EU AI Act — A recent study conducted by Stanford researchers has found thatmajor AI models are falling short of the responsible AI s…
S117
Global Perspectives on Openness and Trust in AI — I don’t know, is really the answer. Governance is such a broad word. There’s a lot of, for example, open source is reall…
S118
The strategic shift toward open-source AI — The release of DeepSeek’s open-source reasoning model in January 2025, followed by the Trump administration’s July endor…
S119
Building Indias Digital and Industrial Future with AI — Deepak Maheshwari from the Centre for Social and Economic Progress provided historical context, tracing India’s digital …
S120
Discussion Report: Sovereign AI in Defence and National Security — Policy and Regulatory Considerations Regulatory frameworks can be adapted to different national contexts The moderator…
S121
AI: Lifting All Boats / DAVOS 2025 — Hatem Dowidar: Look, I mean, there are markets that work closely, let’s say, have strong geopolitical ties. And I just…
S122
Day 0 Event #173 Building Ethical AI: Policy Tool for Human Centric and Responsible AI Governance — Chris Martin: Thanks, Ahmed. Well, everyone, I’ll walk through I think a little bit of this presentation here on what…
S123
AI, automation, and human dignity: Reimagining work beyond the paycheck — AI advancement is inevitable. The technology will continue to improve, costs will continue to fall, and adoption will co…
S124
AI and Digital Developments Forecast for 2026 — Human Rights and Dignity Kurbalija argues that current human rights discourse around AI is overly focused on bias and e…
S125
Global AI Governance: Reimagining IGF’s Role & Impact — Shamira Ahmed: Thanks, Liz. So, my name is Shamira Ahmed and I am the Executive Director of the Data Economy Policy Hub,…
S126
Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session — The foundation for this optimism lies in India’s remarkable digital transformation over the past decade. As Mukesh Amban…
S127
Keynote-Mukesh Dhirubhai Ambani — Ambani makes a bold prediction about India’s future position in global AI leadership. He expresses confidence that India…
S128
Keynote-Demis Hassabis — During his visit to India, Hassabis was particularly impressed by the country’s potential to become a global AI leader. …
S129
Keynote_ 2030 – The Rise of an AI Storytelling Civilization _ India AI Impact Summit — “The first is the fact that we have demographic energy.”[27]”This is certainly a category where India can lead and show …
S130
Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session — “And the Vizag project, the AI Hub, which is a $15 billion investment, is our start.”[1]. “And we will bring a full‑stac…
S131
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Hemant Taneja General Catalyst — This announcement serves as a concrete validation of all the preceding arguments, transforming the discussion from theor…
S132
Welfare for All Ensuring Equitable AI in the Worlds Democracies — -Democratizing AI Access and Preventing Digital Divide: Concerns about AI’s economic value concentrating in Western econ…
S133
Press Briefing by HMIT Ashwani Vaishnav on AI Impact Summit 2026 l Day 5 — Vaishnaw reported substantial investment commitments made during the summit, with over $250 billion pledged for infrastr…
S134
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Hemant Taneja General Catalyst — Entrepreneurship and Investment Commitment: Focus on startups as the most important institutions for rebuilding society,…
S135
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Matthew Prince Cloudflare — Thank you. Thank you. It’s an honor to be here at India’s AI Summit, and I look forward to what we’ll be doing in Geneva…
S136
Welfare for All Ensuring Equitable AI in the Worlds Democracies — Evidence:Examples include impact on farmers, small schools, NGOs, and small hospitals. Google’s mission is to keep every…
S137
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Takahito Tokita Fujitsu — This comment is insightful because it acknowledges that responsible AI development cannot be achieved by any single comp…
S138
Welcome Address — – Prime Minister Narendra Modi
S139
The Innovation Beneath AI: The US-India Partnership powering the AI Era — Thank you everyone We are up against Jan, we are up against her boss. So, but, let’s have fun in this panel. And the bro…
S140
Keynote-Dario Amodei — “of AI models, their potential for misuse by individuals and governments, and their potential for economic displacement….
S141
For the record: AI, creativity, and the future of music — Michael Nash advocates for collaborative, business-oriented approaches to AI development rather than regulatory restrict…
S142
IMF calls for new fiscal policies to address AI’s economic and environmental impacts — The International Monetary Fund (IMF) hasrecommendedfiscal policies for governments grappling with the economic impacts …
S143
Anthropic launches grants for developing new AI benchmark — Anthropic islaunchinga new program to fund the creation of new benchmarks for better assessing AI model performance and …
S144
Anthropic seeks deeper AI cooperation with India — The chief executive of Anthropic, Dario Amodei,has saidIndia can play a central role in guiding global responses to the …
S145
Amazon’s AI partnership with Anthropic cleared by UK regulator — TheUnited Kingdom’s Competition and Markets Authority (CMA) hasconfirmedthat Amazon’s $4 billion partnership with AI sta…
S146
Generative AI: Steam Engine of the Fourth Industrial Revolution? — In addition to its importance in warfare, AI holds great potential within the healthcare industry. It can significantly …
Speakers Analysis
Detailed breakdown of each speaker’s arguments and positions
A
Arthur Mensch
2 arguments181 words per minute307 words101 seconds
Argument 1
AI will drive multiple digits of global GDP and reshape economies
EXPLANATION
Arthur Mensch argues that AI will have a profound impact on the global economy, driving growth measured in multiple digits of GDP. He stresses that this transformation requires broad participation and access to AI technologies.
EVIDENCE
He stated that AI is going to change the economy profoundly and that AI will drive multiple digits of the global GDP in the coming years, emphasizing the need for everyone to participate and have access to AI. He also warned about the risk of excessive concentration of power and market concentration, advocating for open-source technology to ensure broad access and prevent extractive economies. [74-78]
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Analyses highlight that AI-driven productivity gains could lift GDP by multiple digits, supporting Mensch’s claim [S86].
MAJOR DISCUSSION POINT
AI will drive multiple digits of global GDP and reshape economies
Argument 2
Open‑source AI ensures broad access and prevents market concentration
EXPLANATION
Mensch highlights open‑source AI as a mechanism to democratize AI technology, allowing anyone to modify and deploy AI without external control. This approach mitigates the risk of excessive market concentration and ensures inclusive benefits.
EVIDENCE
He explained that open-source technology creates common goods for everybody, enabling broad access and preventing excessive concentration of power and market dominance, which could lead to higher prices and an extractive economy. [74-78]
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Open-source models are cited as leveling the playing field and curbing market concentration in several reports [S87][S2][S7].
MAJOR DISCUSSION POINT
Open‑source AI ensures broad access and prevents market concentration
AGREED WITH
Sam Altman, Matthew Prince, Vinod Khosla
DISAGREED WITH
Sundar Pichai, Demis Hassabis, Other corporate speakers
D
Demis Hassabis
2 arguments193 words per minute304 words94 seconds
Argument 1
AI’s impact will be ten‑times the Industrial Revolution, delivered in a decade
EXPLANATION
Hassabis compares the forthcoming AI revolution to the Industrial Revolution, but ten times larger and occurring ten times faster, within a single decade. He suggests this will unlock a new era of scientific discovery and economic growth.
EVIDENCE
He described AI’s impact as “sort of 10 times the industrial revolution… at 10 times the speed happening over a decade instead of a century,” estimating an impact equivalent to 100 times the industrial revolution and foreseeing a golden era of scientific discovery. [47-49]
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Hassabis described AI’s impact as ten times the industrial revolution within a decade in his keynote address [S3][S89].
MAJOR DISCUSSION POINT
AI’s impact will be ten‑times the Industrial Revolution, delivered in a decade
AGREED WITH
Sam Altman, Nikesh Arora, Rishi Sunak
Argument 2
DeepMind’s research investments complement Google’s infrastructure rollout
EXPLANATION
Hassabis notes that DeepMind’s breakthroughs, such as AlphaFold, exemplify how AI research can accelerate science and medicine, complementing Google’s broader infrastructure investments in India.
EVIDENCE
He referenced AlphaFold’s solution of the 50-year protein-folding challenge as an example of AI advancing science and medicine, linking this research success to Google’s larger investment and partnership plans in India. [43-46]
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
He highlighted DeepMind breakthroughs such as AlphaFold as complementary to Google’s broader infrastructure investments in India [S3].
MAJOR DISCUSSION POINT
DeepMind’s research investments complement Google’s infrastructure rollout
M
Mukesh Ambani
2 arguments121 words per minute487 words240 seconds
Argument 1
AI will accelerate growth, create jobs and build strategic resilience
EXPLANATION
Ambani asserts that AI will be a catalyst for economic acceleration, job creation, and long‑term strategic resilience for India. He ties this to India’s broader digital transformation and the nation’s capacity to lead globally.
EVIDENCE
He said AI will accelerate growth, create jobs and build strategic resilience, noting India’s readiness to become a global AI leader and emphasizing the role of AI in economic expansion and job creation. [280-304]
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Ambani framed AI as the cornerstone of India’s transformation and a driver of growth and strategic resilience [S50].
MAJOR DISCUSSION POINT
AI will accelerate growth, create jobs and build strategic resilience
Argument 2
Reliance’s ₹10 lakh crore AI investment for social sectors and democratized intelligence
EXPLANATION
Ambani announced a massive ₹10 lakh crore (approximately $10 trillion) investment over seven years to develop AI across social sectors such as education, healthcare, and agriculture, aiming to make AI democratic and affordable for all Indians.
EVIDENCE
He detailed Reliance and Jio’s commitment to invest 10 lakh crore rupees in AI over the next seven years, focusing on social sectors, democratizing intelligence, and partnering with startups and research institutions to serve India first and then the world. [296-303]
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
The summit’s plenary noted Reliance’s commitment of 10 lakh crore rupees over seven years for AI in social sectors [S2][S13].
MAJOR DISCUSSION POINT
Reliance’s ₹10 lakh crore AI investment for social sectors and democratized intelligence
DISAGREED WITH
Hemant Taneja, Vinod Khosla
S
Sundar Pichai
2 arguments153 words per minute239 words93 seconds
Argument 1
Full‑stack AI investment and partnership will boost India’s global leadership
EXPLANATION
Pichai outlines Google’s comprehensive commitment to India, spanning hardware (TPUs), infrastructure, research, and sector‑specific partnerships, beginning with the $15 billion Vizag AI Hub.
EVIDENCE
He highlighted Google’s full-stack commitment, including TPUs, infrastructure investments, research, and the $15 billion Vizag AI Hub, as well as partnerships across agriculture, healthcare, language access, skilling, and end-to-end collaborations with Indian companies. [7-10][12-14]
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Pichai outlined Google’s $15 billion Vizag AI Hub and full-stack AI commitment for India in his keynote [S5][S90].
MAJOR DISCUSSION POINT
Full‑stack AI investment and partnership will boost India’s global leadership
AGREED WITH
Brad Smith, Alexander Wang, Shantanu Narayen, Raj Subramaniam, Cristiano Amon, Sanjay Mehrotra, Mansour Ibrahim Al Mansouri
DISAGREED WITH
Arthur Mensch, Demis Hassabis, Other corporate speakers
Argument 2
Google’s skilling programs will prepare India’s talent for AI adoption
EXPLANATION
Pichai emphasizes that Google will work with the Indian government to develop AI skilling initiatives, ensuring the workforce is ready to adopt and build AI‑powered services.
EVIDENCE
He mentioned that Google will work on skilling, collaborating with the government to develop AI talent and capabilities across the country. [13-14]
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
He emphasized collaboration with the Indian government on AI skilling initiatives and talent development [S91].
MAJOR DISCUSSION POINT
Google’s skilling programs will prepare India’s talent for AI adoption
H
Hemant Taneja
2 arguments152 words per minute310 words122 seconds
Argument 1
AI can empower the entire Indian workforce, multiplying productivity
EXPLANATION
Taneja argues that AI will dramatically boost productivity across India’s massive workforce, turning the country into a powerhouse of AI‑enhanced output.
EVIDENCE
He stated that if everybody had the productivity of AI in the Indian workforce, it would create an amazing country and opportunity for the world, emphasizing AI’s role in empowering the workforce rather than displacing jobs. [444-452]
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Taneja argued that AI could dramatically boost productivity across India’s massive workforce, echoing broader productivity forecasts [S86][S92].
MAJOR DISCUSSION POINT
AI can empower the entire Indian workforce, multiplying productivity
Argument 2
General Catalyst’s $5 bn five‑year investment to empower Indian AI ecosystem
EXPLANATION
Taneja announced a commitment of $5 billion over five years to invest in Indian AI startups, underscoring confidence in India’s entrepreneurial ecosystem and its potential to generate global AI leaders.
EVIDENCE
He disclosed that General Catalyst has agreed to invest about $5 billion in the next five years in the Indian entrepreneurial ecosystem, highlighting past investments and a belief in India’s talent and startup potential. [453-455]
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
He announced a $5 billion, five-year investment in Indian AI startups, as reported in summit coverage [S92][S29].
MAJOR DISCUSSION POINT
General Catalyst’s $5 bn five‑year investment to empower Indian AI ecosystem
R
Raj Subramaniam
2 arguments161 words per minute224 words83 seconds
Argument 1
AI‑enabled logistics can cut costs dramatically, enhancing supply‑chain efficiency
EXPLANATION
Subramaniam explains that applying AI to FedEx’s logistics operations can reduce freight costs from 15 % to 8 %, making supply chains smarter and more efficient across India.
EVIDENCE
He noted that AI superpowers FedEx’s data to make supply chains smarter, and that the cost of logistics in India can be reduced from 15 % to 8% through AI-driven initiatives. [128-133]
MAJOR DISCUSSION POINT
AI‑enabled logistics can cut costs dramatically, enhancing supply‑chain efficiency
Argument 2
FedEx’s ₹10 000 crore hub in Navi Mumbai and logistics AI deployment
EXPLANATION
Subramaniam announced a major investment of about ₹10 000 crore in India, including the groundbreaking of a new hub in Navi Mumbai, to expand FedEx’s logistics network and AI‑driven capabilities.
EVIDENCE
He described the announcement of a new hub in Navi Mumbai with Patnavisji and the Adani Group, stating that FedEx is investing about ₹10 000 crore in India and plans to double that investment in the next three years. [136-139]
MAJOR DISCUSSION POINT
FedEx’s ₹10 000 crore hub in Navi Mumbai and logistics AI deployment
S
Sanjay Mehrotra
2 arguments137 words per minute368 words160 seconds
Argument 1
Memory and semiconductor capacity are the fuel for AI’s expansion
EXPLANATION
Mehrotra likens memory to fuel for AI, emphasizing that semiconductor production and advanced memory are essential enablers for AI’s growth and the digital economy.
EVIDENCE
He said, “If AI is the engine of the digital economy, then memory is the fuel of AI,” and highlighted Micron’s large-scale semiconductor clean-room and memory production in India. [192-200]
MAJOR DISCUSSION POINT
Memory and semiconductor capacity are the fuel for AI’s expansion
Argument 2
Micron’s 500 000 sq ft semiconductor clean‑room and memory production
EXPLANATION
Mehrotra detailed Micron’s investment in a massive 500,000 sq ft clean‑room facility in India, which will produce advanced memory chips, supporting AI workloads and the broader semiconductor ecosystem.
EVIDENCE
He described the clean-room’s size (equivalent to ten cricket fields), the steel and concrete used, the current workforce of 2,000 employees (ramping to 5,000), and that 10 % of Micron’s global production will be assembled and tested there, underscoring government support and strategic importance. [194-204]
MAJOR DISCUSSION POINT
Micron’s 500 000 sq ft semiconductor clean‑room and memory production
V
Vinod Khosla
3 arguments124 words per minute495 words238 seconds
Argument 1
AI should be leveraged for the benefit of every citizen, fostering inclusive prosperity
EXPLANATION
Khosla stresses that AI must be used to benefit all citizens, ensuring inclusive prosperity and preventing concentration of power.
EVIDENCE
He argued that AI should be democratized so that every Indian benefits, emphasizing inclusive growth and warning against excessive concentration of power. [460-466]
MAJOR DISCUSSION POINT
AI should be leveraged for the benefit of every citizen, fostering inclusive prosperity
AGREED WITH
Sunil Bharti Mittal, Mukesh Ambani
Argument 2
AI services (tutors, doctors, agronomists) should be integrated with Aadhaar for universal reach
EXPLANATION
Khosla proposes that free AI‑powered tutors, doctors, and agronomists be delivered as Aadhaar‑linked services, mirroring how UPI and other utilities are integrated with Aadhaar, to ensure universal access.
EVIDENCE
He suggested adding free AI tutors for every child, AI doctors for every citizen, and AI agronomists for small farmers as Aadhaar services, comparing this to how UPI is linked to Aadhaar, and emphasizing the need for universal benefit. [462-470]
MAJOR DISCUSSION POINT
AI services (tutors, doctors, agronomists) should be integrated with Aadhaar for universal reach
AGREED WITH
Sam Altman, Arthur Mensch, Matthew Prince
DISAGREED WITH
Adobe (Shantanu Narayen), Meta (Alexander Wang), Google (Sundar Pichai)
Argument 3
Sovereign AI models (e.g., Sarvam) keep data within national boundaries
EXPLANATION
Khosla highlights the importance of sovereign AI models that retain data within a country, citing Sarvam as an example of an Indian‑owned AI model that respects data sovereignty.
EVIDENCE
He mentioned early investment in Sarvam, a sovereign AI model, underscoring the principle that data should remain within national borders. [473-474]
MAJOR DISCUSSION POINT
Sovereign AI models (e.g., Sarvam) keep data within national boundaries
S
Shantanu Narayen
2 arguments151 words per minute201 words79 seconds
Argument 1
AI tools must be free for students to build a creative economy
EXPLANATION
Narayen announces that Adobe will make its AI‑powered products (Photoshop, Acrobat, Firefly) free for students, aiming to equip the next generation with creative tools and foster a vibrant creative economy.
EVIDENCE
He said Adobe announced that all AI products would be available free for students, enabling them to acquire skills for the creative economy. [115-119]
MAJOR DISCUSSION POINT
AI tools must be free for students to build a creative economy
Argument 2
Adobe’s free AI products for students and content‑authenticity initiative
EXPLANATION
In addition to free student tools, Adobe introduced a content authenticity initiative, including watermarking, to ensure trustworthy digital content.
EVIDENCE
He described the launch of a content authenticity initiative, including watermarking, to address content provenance and maintain trust in digital media. [120-122]
MAJOR DISCUSSION POINT
Adobe’s free AI products for students and content‑authenticity initiative
M
Matthew Prince
2 arguments165 words per minute378 words137 seconds
Argument 1
AI should be placed in the hands of 500 000 companies, with business models for creators and SMEs
EXPLANATION
Prince proposes a framework where half a million AI companies exist, each with viable business models that support creators, journalists, and small businesses, preventing AI from merely extracting value.
EVIDENCE
He called for 500,000 AI companies, emphasizing the need for business models that benefit creators, journalists, and SMEs, and warned that without such models AI would take value without giving back. [252-259]
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Prince advocated for a distributed AI ecosystem of 500,000 companies to support creators and SMEs, likening it to the printing press [S88].
MAJOR DISCUSSION POINT
AI should be placed in the hands of 500 000 companies, with business models for creators and SMEs
Argument 2
Cloudflare provides free credits and infrastructure to nurture Indian AI startups
EXPLANATION
Prince outlines Cloudflare’s commitment to support Indian AI startups by offering free credits, training, and infrastructure, facilitating the growth of the next generation of AI companies.
EVIDENCE
He noted that Cloudflare provides free credits, training, and infrastructure to Indian startups, runs an incubator program with many Indian participants, and is rolling out AI for Bharat across its network at low cost. [264-267]
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
He described Cloudflare’s free credits, training, and incubator support for Indian AI startups [S88].
MAJOR DISCUSSION POINT
Cloudflare provides free credits and infrastructure to nurture Indian AI startups
S
Sam Altman
1 argument193 words per minute224 words69 seconds
Argument 1
AI must be democratized globally, with sovereign approaches for each country
EXPLANATION
Altman stresses that AI should be made accessible worldwide, with each nation developing its own sovereign strategies to ensure equitable benefits and mitigate disruptions.
EVIDENCE
He said AI must be democratized, placed in the hands of many people, and that countries need sovereign approaches, emphasizing the need for iterative deployment and global cooperation. [85-99]
MAJOR DISCUSSION POINT
AI must be democratized globally, with sovereign approaches for each country
AGREED WITH
Arthur Mensch, Matthew Prince, Vinod Khosla
A
Alexander Wang
2 arguments173 words per minute295 words102 seconds
Argument 1
Meta will empower millions of small businesses via WhatsApp and digital governance tools
EXPLANATION
Wang describes Meta’s initiatives to support small businesses through WhatsApp, providing digital tools that enable commerce and governance services at scale across India.
EVIDENCE
He highlighted Meta’s support for tens of millions of small businesses via WhatsApp, the deployment of governance tools to citizens (e.g., 100 million subway tickets sold via WhatsApp), and the broader vision of digitization and AI. [60-66]
MAJOR DISCUSSION POINT
Meta will empower millions of small businesses via WhatsApp and digital governance tools
Argument 2
Meta’s collaboration with government for AI‑enabled citizen services
EXPLANATION
Wang notes Meta’s partnership with the Indian government to deliver AI‑driven citizen services, such as governance tools delivered through WhatsApp, reinforcing public‑private collaboration.
EVIDENCE
He mentioned partnering with the government to bring governance tools directly to citizens via WhatsApp, citing the example of state of Andhra Pradesh and the sale of subway tickets. [62-66]
MAJOR DISCUSSION POINT
Meta’s collaboration with government for AI‑enabled citizen services
J
Jeetu Patel
2 arguments160 words per minute245 words91 seconds
Argument 1
Cisco’s training of 800 000 Indians supports AI democratization and security
EXPLANATION
Patel highlights Cisco’s investment in skilling, having trained about 800,000 Indians in cybersecurity and AI, thereby fostering a secure and inclusive AI ecosystem.
EVIDENCE
He stated that Cisco trained roughly 800,000 Indians last year in cybersecurity and AI, emphasizing the role of this training in supporting AI democratization and secure deployments. [232-238]
MAJOR DISCUSSION POINT
Cisco’s training of 800 000 Indians supports AI democratization and security
Argument 2
Cisco’s AI infrastructure supports secure, scalable deployments for enterprises
EXPLANATION
Patel explains that Cisco is providing underlying AI infrastructure to ensure that AI deployments are not constrained, secure, and can scale for enterprises across India and beyond.
EVIDENCE
He said Cisco wants to partner with India to provide the underlying infrastructure so AI does not become a constraint, and to ensure AI is safe and secure for all citizens. [237-239]
MAJOR DISCUSSION POINT
Cisco’s AI infrastructure supports secure, scalable deployments for enterprises
S
Sunil Bharti Mittal
2 arguments150 words per minute362 words144 seconds
Argument 1
Affordable AI delivery to billions of users aligns with frugal innovation
EXPLANATION
Mittal emphasizes India’s frugal innovation model, noting that AI services should be delivered at low cost to billions, leveraging the country’s extensive connectivity infrastructure.
EVIDENCE
He highlighted India’s cheap smartphones and data plans ($2 per month for unlimited data), the extensive fiber and data-center network, and the need to bring AI to a billion customers in a frugal way. [330-340]
MAJOR DISCUSSION POINT
Affordable AI delivery to billions of users aligns with frugal innovation
AGREED WITH
Mukesh Ambani, Vinod Khosla
Argument 2
Airtel’s 5G rollout and frugal innovation model for nationwide connectivity
EXPLANATION
Mittal describes Airtel’s rapid 5G deployment across every square kilometre of India, positioning the network as a foundational layer for AI and other emerging technologies.
EVIDENCE
He recounted the Prime Minister’s directive to launch 5G in ten months, Airtel’s achievement of 5G across the country, and the broader frugal innovation approach that makes advanced technology accessible. [330-340]
MAJOR DISCUSSION POINT
Airtel’s 5G rollout and frugal innovation model for nationwide connectivity
T
Takahito Tokita
2 arguments121 words per minute255 words126 seconds
Argument 1
Data sovereignty, ethical AI, and protection of human dignity must accompany AI growth
EXPLANATION
Tokita stresses that AI development must respect data sovereignty, ensure ethical standards, and protect human dignity, advocating for safe, reliable data spaces and ethical AI research.
EVIDENCE
He discussed the need for data sovereignty, safe and reliable data spaces, protection of human dignity, and the importance of ethics alongside AI evolution. [141-146]
MAJOR DISCUSSION POINT
Data sovereignty, ethical AI, and protection of human dignity must accompany AI growth
AGREED WITH
Rishi Sunak, Nikesh Arora, Matthew Prince
Argument 2
Fujitsu’s joint work on data‑sovereign AI platforms
EXPLANATION
Tokita mentions Fujitsu’s collaboration with the Japanese government and Indian stakeholders to develop AI platforms that respect data sovereignty and support an AI‑led society.
EVIDENCE
He noted Fujitsu’s discussions with the Japanese government on AI-led society, emphasizing the need for high-performance, power-saving computing platforms and collaboration on data-sovereign AI. [141-146]
MAJOR DISCUSSION POINT
Fujitsu’s joint work on data‑sovereign AI platforms
D
Dario Amodei
1 argument152 words per minute371 words145 seconds
Argument 1
Sharing economic impact data enables balanced growth and disruption mitigation
EXPLANATION
Amodei argues that companies should publish economic statistics on AI usage and share data with governments to collectively understand AI’s economic transformation, allowing for mitigation of disruptions.
EVIDENCE
He said his company keeps economic statistics on model usage, encourages others to do the same, and that sharing data helps accentuate benefits and mitigate disruptions. [31-35]
MAJOR DISCUSSION POINT
Sharing economic impact data enables balanced growth and disruption mitigation
R
Roy Jakobs
2 arguments186 words per minute450 words144 seconds
Argument 1
Regulated, trustworthy AI is critical for medical devices and health outcomes
EXPLANATION
Jakobs emphasizes the necessity of regulation and trustworthiness for AI in medical equipment, ensuring safety, efficacy, and positive health impacts.
EVIDENCE
He discussed the need for transparent and regulated AI, the importance of trust, and the challenges of rapid AI development in medical contexts. [169-176]
MAJOR DISCUSSION POINT
Regulated, trustworthy AI is critical for medical devices and health outcomes
Argument 2
AI‑enabled medical equipment and health data platforms improve patient outcomes
EXPLANATION
Jakobs outlines Philips’ work in India, leveraging AI and software developed locally to enhance health data platforms, supporting primary care and improving outcomes for patients.
EVIDENCE
He described Philips’ investment of $1.7 billion in software and AI, collaborations with the Ministry of Health on data, and AI tools for primary care centers and ASHA workers to address health challenges early. [169-176]
MAJOR DISCUSSION POINT
AI‑enabled medical equipment and health data platforms improve patient outcomes
B
Brad Smith
1 argument149 words per minute308 words123 seconds
Argument 1
Digital sovereignty and cross‑border trust are needed for technology services
EXPLANATION
Smith argues that while nations protect digital sovereignty, there must be mechanisms to allow technology services to cross borders securely, positioning the US‑India partnership as a model.
EVIDENCE
He highlighted Microsoft’s role in advocating for Indian IT sector needs, the importance of cross-border technology trust, and the need to protect digital sovereignty while enabling services to flow globally. [158-162]
MAJOR DISCUSSION POINT
Digital sovereignty and cross‑border trust are needed for technology services
N
Nandan Nilekani
1 argument152 words per minute432 words170 seconds
Argument 1
AI‑driven agritech for dairy farmers (Amul) demonstrates rapid diffusion and impact
EXPLANATION
Nilekani shares the rapid rollout of an AI‑based application for Amul’s dairy farmers, illustrating how AI can quickly reach millions of users and improve agricultural productivity.
EVIDENCE
He recounted meeting the Prime Minister on Jan 8, the subsequent meeting with Amul, and the launch of the Sarla Ben application on Feb 11, which now serves 3.6 million farmers and 40 million cattle, providing real-time insights on health, pregnancy, and milk production. [354-367]
MAJOR DISCUSSION POINT
AI‑driven agritech for dairy farmers (Amul) demonstrates rapid diffusion and impact
E
Enrico Bagnasco
1 argument116 words per minute164 words84 seconds
Argument 1
Blue Raman subsea cable deployment with Google to boost AI‑ready connectivity
EXPLANATION
Bagnasco describes SPACL’s partnership with Google to deploy the Blue Raman subsea cable linking Italy and India, enhancing the digital infrastructure needed for AI services.
EVIDENCE
He noted that SPACL is deploying the Blue Raman subsea cable together with Google, connecting Milan, Italy, and Mumbai, India, as part of a diversified route to support AI-ready connectivity. [416-418]
MAJOR DISCUSSION POINT
Blue Raman subsea cable deployment with Google to boost AI‑ready connectivity
G
Giordano Albertazzi
1 argument117 words per minute78 words39 seconds
Argument 1
Vertiv’s expansion of manufacturing and services in India
EXPLANATION
Albertazzi announces Vertiv’s commitment to expand its manufacturing, engineering, and service presence in India, supporting the country’s AI and digital infrastructure growth.
EVIDENCE
He stated that Vertiv can be counted on for expansion of manufacturing, engineering, and service presence in India, expressing enthusiasm for the adventure. [423-424]
MAJOR DISCUSSION POINT
Vertiv’s expansion of manufacturing and services in India
M
Mansour Ibrahim Al Mansouri
1 argument0 words per minute0 words1 seconds
Argument 1
G42’s AI factories and cross‑border partnership with India
EXPLANATION
Al Mansouri outlines G42’s development of two large AI factories—a token factory for intelligence at scale and an agent factory for enterprise—built in partnership with Indian stakeholders.
EVIDENCE
He described G42 delivering two large factories of the future (a token factory and an agent factory), built with strategic partners present in the room, and emphasized the importance of sovereign intelligence infrastructure and cross-border partnership with India. [433-440]
MAJOR DISCUSSION POINT
G42’s AI factories and cross‑border partnership with India
R
Rishi Sunak
1 argument164 words per minute628 words229 seconds
Argument 1
Transparent governance and accountability are essential for responsible AI
EXPLANATION
Sunak calls for clear governance frameworks, transparency, and accountability in AI development to maintain public trust and ensure safety.
EVIDENCE
He urged maintaining transparency, dialogue between government and companies, and appropriate governance at the right moment to keep citizens’ trust and safety. [396-401]
MAJOR DISCUSSION POINT
Transparent governance and accountability are essential for responsible AI
AGREED WITH
Takahito Tokita, Nikesh Arora, Matthew Prince
N
Nikesh Arora
1 argument207 words per minute419 words121 seconds
Argument 1
AI security competence centre in Bangalore to address safety, governance and societal impact
EXPLANATION
Arora announces the establishment of a large AI security competence centre in Bangalore, staffed by 1,500 people, to develop governance, accountability, and cybersecurity capabilities for AI.
EVIDENCE
He detailed the creation of an AI security competence centre in Bangalore with over 1,500 staff, focusing on governance, accountability, cybersecurity, and social upskilling to ensure safe AI development. [274-276]
MAJOR DISCUSSION POINT
AI security competence centre in Bangalore to address safety, governance and societal impact
S
Sébastien Fabre
1 argument140 words per minute169 words72 seconds
Argument 1
Modular, sovereign AI architecture ensures independence and security
EXPLANATION
Fabre emphasizes that sovereign AI should be built on open, modular architectures that can be deployed on sovereign infrastructure, ensuring security and independence for critical applications.
EVIDENCE
He explained that sovereignty is about open, modular architecture that can be deployed on sovereign infrastructure using sovereign data, and that Safran has been in India for 65 years with a commitment to double presence by 2030. [181-186]
MAJOR DISCUSSION POINT
Modular, sovereign AI architecture ensures independence and security
C
Cristiano Amon
1 argument159 words per minute190 words71 seconds
Argument 1
Qualcomm’s 2‑nm chip design and R&D in India
EXPLANATION
Amon announces that Qualcomm has designed a 2‑nm chip in India, highlighting the country’s growing R&D capabilities and its role in the global semiconductor supply chain.
EVIDENCE
He noted that the first two-nanometer chip in India was designed by Qualcomm’s team, emphasizing the significance of India’s semiconductor R&D. [218-219]
MAJOR DISCUSSION POINT
Qualcomm’s 2‑nm chip design and R&D in India
J
Julie Sweet
1 argument152 words per minute195 words76 seconds
Argument 1
Microsoft highlights India’s large AI workforce and invests in training
EXPLANATION
Sweet points out that India has one of the world’s largest AI workforces and that Microsoft is investing in training programs to further develop AI talent across the country.
EVIDENCE
She stated that Microsoft has over 350,000 people in India, is growing, bringing companies through Global Capability Centers, and investing to train everyone, emphasizing the large AI workforce. [102-108]
MAJOR DISCUSSION POINT
Microsoft highlights India’s large AI workforce and invests in training
R
Ravi Mhatre
1 argument128 words per minute269 words125 seconds
Argument 1
Lightspeed’s continued funding and focus on India’s AI talent export
EXPLANATION
Mhatre describes Lightspeed’s long‑term investment in Indian AI startups and the country’s role as a talent exporter, supporting both domestic AI applications and global AI product development.
EVIDENCE
He mentioned Lightspeed’s $1 billion investment over 17 years, plans to increase investments, and highlighted India’s talent density, large developer community, and ability to build transformative products for both India and the world. [490-502]
MAJOR DISCUSSION POINT
Lightspeed’s continued funding and focus on India’s AI talent export
M
Marcus Wallenberg
2 arguments125 words per minute269 words128 seconds
Argument 1
The AI initiative will make India a more attractive destination for Swedish and other foreign investors, boosting future business opportunities.
EXPLANATION
Wallenberg argues that the Indian government’s focus on AI will enhance the investment climate for companies like Ericsson, ABB, and AstraZeneca, encouraging them to expand their operations and develop new AI‑driven products in India.
EVIDENCE
He explained that the AI push will be “very, very supportive for future business of these companies in India” and that it will “position India as an even better place for investment and development of their businesses” for firms such as Ericsson, ABB and AstraZeneca. He also noted that these companies have been active in India for over a century. [377-379]
MAJOR DISCUSSION POINT
AI initiative as a catalyst for foreign investment and business growth in India
Argument 2
Swedish defence company Saab will deepen its AI collaboration with the Indian government, contributing to India’s security and defence capabilities.
EXPLANATION
Wallenberg points out that Saab, although a defence firm rather than a car maker, intends to work with the Indian government on AI‑enabled defence solutions, reinforcing India’s strategic autonomy.
EVIDENCE
He clarified that “Saab is not a car company, it’s a defence company” and expressed the intention to “serve the Indian government going forward” with AI-related products and services. [380-381]
MAJOR DISCUSSION POINT
AI partnership with defence sector to strengthen India’s security capabilities
M
Mansour Ibrahim Al Mansouri
2 arguments150 words per minute176 words70 seconds
Argument 1
G42 is building two large AI factories in India to provide sovereign, large‑scale AI infrastructure for the country.
EXPLANATION
Al Mansouri states that G42 will deliver a “token factory” for intelligence at scale and an “agent factory” for enterprise AI, establishing core AI infrastructure that is owned and operated within India.
EVIDENCE
He described the two factories – a token factory for intelligence at scale and an agent factory to empower enterprises – as being built in collaboration with strategic partners present in the room, emphasizing the goal of creating sovereign AI infrastructure. [433-440]
MAJOR DISCUSSION POINT
Establishment of sovereign AI factories to underpin India’s AI ecosystem
Argument 2
Cross‑border partnership with India will enable both nations to lead the next century of economic growth through AI.
EXPLANATION
Al Mansouri highlights that the UAE and India share a common conviction that AI is a core national infrastructure and that joint collaboration will help both countries build the strongest intelligence infrastructure, driving future economic development.
EVIDENCE
He said, “the UAE believes India is a partner, and let us build this partnership and lead the future together,” stressing the importance of cross-border cooperation for AI-driven growth. [437-440]
MAJOR DISCUSSION POINT
Strategic India‑UAE partnership to accelerate AI‑driven economic growth
N
Natarajan Chandrasekaran
2 arguments122 words per minute376 words184 seconds
Argument 1
Tata Group will invest in building end‑to‑end AI infrastructure—from hardware and chips to data centres and AI agents—to support India’s AI ecosystem and global AI markets.
EXPLANATION
Chandrasekaran outlines Tata’s commitment to three pillars: scaling the nation, constructing AI hardware and data‑centre infrastructure, and extending AI capabilities to businesses worldwide, thereby creating a robust AI stack for India and beyond.
EVIDENCE
He listed the three important areas: scaling the nation, building AI infrastructure across hardware, chips, data centres and agents, and delivering AI solutions to businesses around the world. [322-326]
MAJOR DISCUSSION POINT
Tata’s comprehensive investment in AI infrastructure and services
Argument 2
India’s ambition and talent density give it a natural advantage to lead responsibly in AI on the global stage.
EXPLANATION
Chandrasekaran emphasizes that India’s combination of ambition and skilled workforce positions the country to set responsible AI agendas, promote open collaboration, and drive AI progress for the benefit of the world.
EVIDENCE
He noted that under the Prime Minister’s leadership, India possesses “ambition and wanting to scale” and that these qualities make it well-positioned to lead in AI for the good of the world. [317-321]
MAJOR DISCUSSION POINT
India’s ambition and talent as drivers of responsible global AI leadership
S
Shri Narendra Modi
2 arguments130 words per minute486 words223 seconds
Argument 1
The government will work hand‑in‑hand with industry and academia to create policies, implementation frameworks and trust‑building mechanisms that ensure inclusive and responsible AI development.
EXPLANATION
Modi stresses a collaborative approach, urging all stakeholders to travel together, maintain dialogue, and co‑create policies that balance innovation with safety, thereby fostering an ecosystem where AI benefits everyone.
EVIDENCE
He said, “we should together, like a traveler, like a co-traveler, keep these targets and these directions in mind and achieve our goal,” and later added that “we must maintain the trust and confidence of our citizens as this technology develops” while emphasizing partnership with industry. [516-525][527]
MAJOR DISCUSSION POINT
Collaborative governance and policy co‑creation for inclusive AI
Argument 2
AI is a transformative technology that will drive India’s future prosperity, and the nation must harness it through coordinated effort, democratic values and frugal innovation.
EXPLANATION
Modi portrays AI as a catalyst for economic and social upliftment, calling for democratic, affordable AI solutions that align with India’s tradition of frugal innovation and serve the broader population.
EVIDENCE
He referred to India’s ability to deliver large-scale projects frugally (e.g., moon mission) and asserted that AI will be a “new golden era” that benefits everyone, urging the country to act collectively to realise this vision. [343-349][527]
MAJOR DISCUSSION POINT
AI as a driver of inclusive, frugal-driven national development
A
Ashwini Vaishnaw
2 arguments125 words per minute737 words352 seconds
Argument 1
The summit has addressed every layer of the AI stack—from foundational models to services, infrastructure, compute and financing—demonstrating a holistic national strategy for AI development.
EXPLANATION
Vaishnaw summarises that the round‑table covered models, services, infrastructure, compute, and funding, indicating that India is pursuing an integrated approach that aligns all components of the AI ecosystem.
EVIDENCE
He explicitly stated, “we have covered practically all the layers of the AI stack. We have covered models. We have covered services, infra, compute, from funding to overall the conventional, the machine makers, the industry leaders.” [503-506]
MAJOR DISCUSSION POINT
Comprehensive coverage of the entire AI stack at the summit
Argument 2
The government is committed to fostering public‑private partnerships that accelerate AI adoption across sectors and ensure that AI benefits all citizens.
EXPLANATION
Throughout the round‑table, Vaishnaw repeatedly invites CEOs to present their AI plans, emphasizes partnership, and calls for collaborative action, reflecting a policy stance that prioritises joint initiatives between government and industry.
EVIDENCE
He introduced each speaker, thanked them, and repeatedly said “we look forward to partnering” and “the floor is yours,” signalling an ongoing commitment to partnership across agriculture, healthcare, education, and other sectors. Examples include his introductions of Sundar Pichai, Dario Amodei, Demis Hassabis, and many others, each followed by remarks about collaboration. [1-4][19-20][36-38][70-71][100-102][140-142][164-166][224-226][242-245][269-271][277-279][305-307][383-384]
MAJOR DISCUSSION POINT
Government’s proactive role in building public‑private AI partnerships
Agreements
Agreement Points
AI will have a transformative economic and societal impact comparable to a ten‑fold Industrial Revolution within a decade
Speakers: Demis Hassabis, Sam Altman, Nikesh Arora, Rishi Sunak
AI’s impact will be ten‑times the Industrial Revolution, delivered in a decade AI must be democratized… this will be a seismic shift… ten times the size of the Industrial Revolution, ten times faster If what Dev is saying is going to happen, that we will have ten times the industrial revolution in India, and ten times the speed the impact … ten times the industrial revolution … ten times the speed
All four speakers highlighted that the coming AI wave will be orders of magnitude larger and faster than the historic Industrial Revolution, reshaping economies and societies within a single decade [47-49][85-89][270-272][386-389]
POLICY CONTEXT (KNOWLEDGE BASE)
This view mirrors the assessment of Demis Hassabis at the India AI summit, who quantified AGI’s impact as roughly ten times the Industrial Revolution and expected it within a decade [S107].
AI must be democratized and made broadly accessible to all citizens
Speakers: Sam Altman, Arthur Mensch, Matthew Prince, Vinod Khosla
AI must be democratized globally, with sovereign approaches for each country Open‑source AI ensures broad access and prevents market concentration First, there should be 500,000 AI companies… AI should be a tool for all, including students, as you mentioned, and the poorest members of the global south AI services (tutors, doctors, agronomists) should be integrated with Aadhaar for universal reach
The speakers agreed that AI should be open, affordable and delivered as a public good – through open-source models, massive numbers of AI companies, and integration with national platforms like Aadhaar – to ensure every individual benefits [252-259][74-78][85-89][462-470]
POLICY CONTEXT (KNOWLEDGE BASE)
Multiple plenary sessions emphasized that AI tools should be put in the hands of many, including students and the poorest, and that free AI services (tutors, doctors, agronomists) could be layered onto Aadhaar, reflecting a strong democratization agenda [S109][S110][S112].
Large‑scale skilling and capacity development are essential for AI adoption
Speakers: Sundar Pichai, Julie Sweet, Jeetu Patel, Brad Smith
And going forward, we’ll work on skilling, working with the government We’re investing to train everyone… we have over 350,000 people here, and we’re bringing other companies here through the Global Capability Centers as we advise our clients on their talent strategy… We’re investing to train everyone We’ve actually just last year trained about 800,000 Indians with skills in cybersecurity and AI and networking We are skilling
All highlighted programmes to develop AI talent – Google’s skilling partnership, Microsoft’s training of 350k staff, Cisco’s 800k-person up-skilling effort, and Microsoft’s broader skilling commitment – as a cornerstone of India’s AI future [13-14][102-108][232-238][154-155]
POLICY CONTEXT (KNOWLEDGE BASE)
Panel discussions highlighted the need for education and capacity-building to be accessible to all ages and backgrounds as a prerequisite for responsible AI deployment [S108].
Responsible governance, ethics and data sovereignty must accompany AI development
Speakers: Rishi Sunak, Takahito Tokita, Nikesh Arora, Matthew Prince
Transparent governance and accountability are essential for responsible AI Data sovereignty, ethical AI, and protection of human dignity must accompany AI growth There is a large question of governance and accountability who is responsible for these agents AI should embrace and enhance our unique culture… we shouldn’t make the same mistakes we made with the Internet
The speakers converged on the need for clear governance frameworks, ethical safeguards and respect for data sovereignty to maintain public trust as AI scales [396-401][141-146][270-274][252-259]
POLICY CONTEXT (KNOWLEDGE BASE)
Speakers called for AI governance to be embedded within existing regulatory frameworks and for clear data-sovereignty rules, especially in national security contexts [S108][S111][S120].
Public‑private partnership and substantial investment are driving AI infrastructure in India
Speakers: Sundar Pichai, Brad Smith, Alexander Wang, Shantanu Narayen, Raj Subramaniam, Cristiano Amon, Sanjay Mehrotra, Mansour Ibrahim Al Mansouri
Full‑stack AI investment and partnership will boost India’s global leadership We are investing… partnering… skilling… we are very bullish on the prospects for AI leadership We’re very excited for continued partnership… empower small businesses… we are also excited to partner with the government of India to bring governance to citizens through WhatsApp All of our AI products… will be available free for students We are investing about 10,000 crores in India… we announced the groundbreaking of our hub in Navi, Mumbai the first two‑nanometer chip in India has been designed by our team 500,000 sq ft clean‑room… 10% of Micron’s global production will be assembled and tested here we are delivering two large factories of the future… token factory… agent factory… cross‑border partnership with India
Multiple leaders described coordinated investments – from Google’s $15 bn Vizag AI Hub and end-to-end partnerships, Microsoft’s multi-year AI commitment, Meta’s WhatsApp-based services, Adobe’s free student tools, FedEx’s ₹10 000 crore logistics hub, Qualcomm’s 2-nm chip design, Micron’s massive semiconductor clean-room, and G42’s AI factories – all illustrating a shared belief in public-private collaboration to build India’s AI stack [7-10][152-155][58-66][115-119][136-139][218-219][194-200][433-440]
POLICY CONTEXT (KNOWLEDGE BASE)
Analyses of the Indian AI ecosystem note that successful private-sector innovation relies heavily on public-sector data and infrastructure support, shaping investment strategies and policy focus [S113][S103][S104].
AI solutions should be delivered affordably, following India’s tradition of frugal innovation
Speakers: Sunil Bharti Mittal, Mukesh Ambani, Vinod Khosla
Affordable AI delivery to billions of users aligns with frugal innovation Reliance’s … AI will be democratic and affordable for every Indian AI should be leveraged for the benefit of every citizen, fostering inclusive prosperity
Mittal emphasized low-cost AI for a billion users, Ambani pledged democratic and affordable AI, and Khosla called for AI services that benefit every citizen, reflecting a consensus on frugal, inclusive deployment [330-340][301-303][460-466]
POLICY CONTEXT (KNOWLEDGE BASE)
India’s AI mission explicitly rejects the “parameter race” and pursues pragmatic, cost-effective models that address societal problems, reflecting a frugal-innovation mindset [S105].
Similar Viewpoints
Both stress that an open, widely distributed AI ecosystem – via open‑source models and a large number of independent AI companies – is essential to avoid concentration of power and to ensure broad societal benefit [74-78][252-259]
Speakers: Arthur Mensch, Matthew Prince
Open‑source AI ensures broad access and prevents market concentration First, there should be 500,000 AI companies… AI should be a tool for all
Both highlight substantial physical‑infrastructure investments (logistics hub and advanced chip design) as critical enablers for AI‑driven economic growth in India [136-139][218-219]
Speakers: Raj Subramaniam, Cristiano Amon
FedEx’s ₹10,000 crore hub in Navi Mumbai and logistics AI deployment Qualcomm’s 2‑nm chip design and R&D in India
Both call for strong governance, data sovereignty and ethical safeguards to accompany rapid AI development [396-401][141-146]
Speakers: Rishi Sunak, Takahito Tokita
Transparent governance and accountability are essential for responsible AI Data sovereignty, ethical AI, and protection of human dignity must accompany AI growth
Both stress that AI must be delivered affordably and at scale, consistent with India’s frugal innovation ethos [330-340][301-303]
Speakers: Sunil Bharti Mittal, Mukesh Ambani
Affordable AI delivery to billions of users aligns with frugal innovation Reliance’s … AI will be democratic and affordable for every Indian
Unexpected Consensus
Defense and aerospace firms aligning with sovereign AI architecture and data sovereignty
Speakers: Sébastien Fabre, Takahito Tokita
Modular, sovereign AI architecture ensures independence and security Data sovereignty, ethical AI, and protection of human dignity must accompany AI growth
Although representing very different sectors (defence vs technology), both emphasized that AI must run on open, modular, sovereign infrastructure and respect data sovereignty, showing an unexpected cross-sector agreement on AI governance principles [181-186][141-146]
POLICY CONTEXT (KNOWLEDGE BASE)
Policy discussions on sovereign AI stress the need for defence and aerospace sectors to adopt national AI architectures that safeguard data sovereignty [S120].
Telecom submarine cable deployment being framed as essential AI infrastructure
Speakers: Enrico Bagnasco, Sundar Pichai
Blue Raman subsea cable deployment with Google to boost AI‑ready connectivity Full‑stack AI investment and partnership will boost India’s global leadership
A telecom carrier (SPACL) and Google’s AI hub leader both positioned high-capacity subsea connectivity as a foundational layer for AI services, an unexpected convergence of telecom and AI strategy discussions [416-418][7-10]
POLICY CONTEXT (KNOWLEDGE BASE)
Historical reviews trace India’s digital backbone to its first submarine cable in 1858, underscoring the long-standing view that robust telecom links are foundational for modern AI infrastructure [S119].
Overall Assessment

The round‑table displayed a strong, multi‑sectoral consensus that AI will be a transformative, high‑impact technology; that its benefits must be democratized through open‑source models, widespread access and public‑private investment; that massive up‑skilling and capacity‑building programmes are essential; and that responsible governance, ethics and data sovereignty are non‑negotiable. A shared commitment to affordable, frugal delivery and to building the full AI stack—from chips and memory to services and governance—underpinned the dialogue.

High consensus across government, multinational corporations, venture capitalists and sectoral leaders, indicating a unified strategic direction for India’s AI ecosystem that combines ambitious investment with inclusive, responsible deployment.

Differences
Different Viewpoints
Open‑source AI versus proprietary, full‑stack partnership models
Speakers: Arthur Mensch, Sundar Pichai, Demis Hassabis, Other corporate speakers
Open‑source AI ensures broad access and prevents market concentration Full‑stack AI investment and partnership will boost India’s global leadership
Mensch argues that open-source technology is needed to avoid excessive concentration of power and to let everyone benefit from AI [74-78]. In contrast, Google’s Sundar Pichai outlines a proprietary, full-stack commitment that includes TPUs, a $15 billion Vizag AI Hub and end-to-end partnerships, without mentioning an open-source approach [7-10][12-14]. This reflects a clash between an open-source, commons-based model and a corporate, proprietary partnership model.
POLICY CONTEXT (KNOWLEDGE BASE)
Leading AI scientists argue that open-source models are overtaking proprietary ones, positioning open-source as a strategic imperative in the global AI race [S99][S101][S102][S115][S118].
Embedding AI services in the national Aadhaar infrastructure versus delivering AI through private‑sector platforms
Speakers: Vinod Khosla, Adobe (Shantanu Narayen), Meta (Alexander Wang), Google (Sundar Pichai)
AI services (tutors, doctors, agronomists) should be integrated with Aadhaar for universal reach AI tools should be offered free to students or via private platforms to empower users
Khosla proposes that free AI tutors, doctors and agronomists be delivered as Aadhaar-linked services, similar to UPI, to guarantee universal access [462-470]. Adobe announces that its AI-powered products will be free for students [115-119], while Meta highlights WhatsApp-based tools for small businesses and citizen services [60-66]. Google’s remarks focus on partnerships and skilling without a public-utility model [12-14]. The disagreement lies in whether AI should be provided as a public, identity-linked utility or through private-sector offerings.
POLICY CONTEXT (KNOWLEDGE BASE)
Debates highlighted the tension between leveraging Aadhaar as a public AI layer (e.g., free services) and allowing private platforms to operate independently, reflecting broader public-private dynamics in AI deployment [S109][S112][S113].
Scale and focus of AI investment in India
Speakers: Mukesh Ambani, Hemant Taneja, Vinod Khosla
Reliance’s ₹10 lakh crore AI investment for social sectors and democratized intelligence General Catalyst’s $5 billion five‑year investment in Indian AI startups Khosla’s early investments in sovereign AI models and broader capital commitments
Ambani announces a massive ₹10 lakh crore (≈$10 trillion) AI investment over seven years focused on education, healthcare and agriculture [296-303]. Taneja pledges $5 billion over five years to fund Indian AI startups [453-455]. Khosla mentions early investments in sovereign AI models like Sarvam and other capital deployments [458-465]. The speakers differ sharply on the magnitude and allocation of resources for AI development in India.
POLICY CONTEXT (KNOWLEDGE BASE)
Stakeholders differ on whether India should chase large-parameter models or prioritize current-level, application-driven AI, with discussions on investment priorities and institutional focus [S105][S103][S104].
Unexpected Differences
Contrasting views on the role of human dignity in AI
Speakers: Takahito Tokita, Mukesh Ambani
AI should function for people and must not have human dignity Manav vision places human‑centric values at the core of AI development
Tokita states that AI should function for people and must not have human dignity, emphasizing a purely functional view of AI [143]. In contrast, Ambani repeatedly references the “Manav” (human) vision as a moral compass for AI, highlighting human-centric leadership [280-283]. This unexpected clash reflects differing philosophical stances on whether AI should be guided by human dignity or treated as a neutral tool.
POLICY CONTEXT (KNOWLEDGE BASE)
Thought leaders argue that AI governance must go beyond bias mitigation to protect fundamental human rights and dignity, emphasizing a human-centric approach to AI policy [S123][S124].
Open‑source stance from a European AI company versus typical proprietary approaches
Speakers: Arthur Mensch, Most other corporate speakers (e.g., Google, Meta, Microsoft)
Open‑source AI ensures broad access and prevents market concentration Corporate partners emphasize proprietary stacks, investments and private platforms
Mensch, representing a European firm, uniquely advocates for open-source AI as a means to avoid concentration and ensure common goods [74-78]. The majority of other speakers (Google, Meta, Microsoft, etc.) focus on proprietary investments, partnerships and private-sector solutions without highlighting open-source, making this an unexpected divergence.
POLICY CONTEXT (KNOWLEDGE BASE)
European policy circles champion open-source as a competitive and trust-building tool, contrasting with the proprietary models dominant elsewhere [S115][S99].
Overall Assessment

The summit displayed broad consensus on AI’s transformative potential and the need for inclusive, secure, and skilled deployment. However, clear disagreements emerged around the preferred model for AI provision (open‑source vs proprietary), the mechanism for universal access (public‑utility/Aadhaar integration vs private‑sector platforms), and the scale of financial commitment (trillion‑rupee national programmes versus multi‑billion‑dollar venture investments). Unexpected philosophical differences on human dignity and an unusual open‑source stance from a European firm added nuance to the debate.

Moderate to high. While participants largely agree on AI’s importance, the divergent views on openness, public versus private delivery, and investment magnitude could affect policy alignment and coordination. These disagreements may lead to parallel tracks of development—government‑driven public‑utility initiatives alongside private‑sector proprietary ecosystems—potentially fragmenting the AI landscape unless reconciled through coordinated frameworks.

Partial Agreements
All speakers share the goal of making AI widely accessible and inclusive. Altman calls for global democratization and sovereign strategies [85-99]; Khosla stresses inclusive prosperity and universal services [460-466]; Prince proposes a massive ecosystem of 500 000 AI companies with models that benefit creators and SMEs [252-259]; Narayen announces free AI products for students to nurture a creative economy [115-119]. However, they diverge on the mechanisms: sovereign‑linked services, free student tools, ecosystem scale, and policy frameworks.
Speakers: Sam Altman, Vinod Khosla, Matthew Prince, Shantanu Narayen
AI must be democratized globally, with sovereign approaches for each country AI should be leveraged for the benefit of every citizen, fostering inclusive prosperity AI should be placed in the hands of 500 000 companies, with business models for creators and SMEs AI tools must be free for students to build a creative economy
All three emphasize the need for robust governance and safety in AI deployment. Arora announces a dedicated AI security competence centre in Bangalore [274-276]; Sunak calls for transparent governance, dialogue and accountability to maintain public trust [396-401]; Jakobs stresses regulation and trustworthiness for AI‑enabled medical equipment [169-176]. They agree on the importance of governance but differ on the institutional approach—industry‑led competence centre, governmental transparency, and sector‑specific regulation.
Speakers: Nikesh Arora, Rishi Sunak, Roy Jakobs
AI security competence centre in Bangalore to address safety, governance and societal impact Transparent governance and accountability are essential for responsible AI Regulated, trustworthy AI is critical for medical devices and health outcomes
Takeaways
Key takeaways
AI is viewed as a transformative economic engine that can add multiple digits to global GDP and outpace the Industrial Revolution in speed and scale. India is positioned to become a global AI leader thanks to its talent pool, startup ecosystem, and recent infrastructure investments. Broad consensus on the need to democratize AI access – free tools for students, AI services for billions, and open‑source models to avoid market concentration. Significant private‑sector commitments to invest in India across the full AI stack: data centres, chips, memory, connectivity, and sector‑specific applications. Partnerships between government and industry are essential for skilling, research, regulatory frameworks, and responsible deployment. Governance, ethics, security, and data sovereignty are critical concerns that must be addressed alongside rapid AI rollout. Sector‑specific AI use cases (agri‑tech, healthcare, logistics, creative economy, defense) are already being piloted and scaled. Talent development and large‑scale skilling programmes are a priority to prepare the workforce for AI‑augmented productivity.
Resolutions and action items
Google to establish a $15 bn Vizag AI Hub, provide TPUs, infrastructure, research collaborations, and launch skilling programmes in India. Anthropic will share economic impact data and collaborate on agritech and other social‑impact projects. DeepMind (Google) will continue research investments and promote AI‑driven scientific breakthroughs in India. Meta will expand AI tools for small businesses via WhatsApp, support digital governance, and deepen partnership with Indian ministries. Adobe will make Photoshop, Acrobat, and Firefly free for students and launch a content‑authenticity (watermarking) initiative. FedEx will invest ₹10 000 crore in a new hub at Navi Mumbai and apply AI to cut logistics costs from 15 % to 8 %. Fujitsu will collaborate on data‑sovereign AI platforms and joint research on AI ethics and safety. Microsoft will work with the US and Indian governments to ensure cross‑border technology trust and promote AI talent development. Reliance (Jio) commits ₹10 lakh crore over seven years to AI investments in social sectors (education, health, agriculture) and to make AI affordable for all Indians. Tata Group will build end‑to‑end AI infrastructure from chips to data centres and support AI‑driven social transformation. Airtel will continue 5G rollout and promote frugal innovation to deliver AI services at low cost nationwide. Qualcomm will design and develop 2‑nm chips in India, expanding its R&D footprint. Micron will complete a 500,000 sq ft semiconductor clean‑room, scaling memory production to 10 % of its global output in India. Blue Raman subsea cable project (Google + SPACL) will enhance connectivity for AI workloads. Vertiv will expand manufacturing and services in India to support AI‑critical infrastructure. G42 will establish two AI factories (token and agent) in partnership with Indian stakeholders. General Catalyst pledged $5 bn over five years to fund Indian AI startups and ecosystem. Lightspeed will increase investments in Indian AI ventures and support sovereign AI models like Sarvam. Cisco will continue AI and cybersecurity training for 800 000 individuals and provide infrastructure for secure AI deployments. Cloudflare will offer free credits, AI‑for‑Bharat services, and support Indian startups through its incubator. AI security competence centre in Bangalore (Palo Alto) will focus on governance, accountability, and safety of autonomous agents. Commitments to open‑source AI development (Mistral) and to keep AI models and data within national boundaries (Sarvam).
Unresolved issues
Specific regulatory frameworks and timelines for AI governance, data protection, and accountability of autonomous agents remain undefined. Mechanisms for ensuring equitable distribution of AI‑generated economic value and preventing excessive market concentration were discussed but not concretized. Details on how AI safety ‘kill‑switches’ and cross‑border security standards will be implemented were not finalized. The process for integrating AI services (tutors, doctors, agronomists) with Aadhaar and other national platforms needs further planning. Clarification on standards for AI transparency, content provenance, and verification across industries is still pending. Agreement on a unified global AI governance model or coordination mechanism among participating countries was not reached.
Suggested compromises
Adoption of open‑source AI models to broaden access and mitigate market concentration (Arthur Mensch). Development of modular, sovereign AI architectures that can run on national data centres while allowing cross‑border collaboration (Sébastien Fabre, Takahito Tokita). Balancing security and cross‑border technology flow through bilateral trust frameworks (Brad Smith). Combining frugal innovation with large‑scale AI deployment to keep services affordable for the masses (Sunil Bharti Mittal). Co‑creating AI governance frameworks that involve both government oversight and industry self‑regulation (Rishi Sunak, Nikesh Arora).
Thought Provoking Comments
We will bring a full‑stack commitment to India, all the way from TPUs to infrastructure investments to research and models, starting with the $15 billion Vizag AI Hub, and we will partner across agriculture, healthcare, language access, skilling and governance.
Sets a concrete, multi‑layered roadmap for Google’s involvement, moving beyond rhetoric to specific investments and partnership domains.
Established the benchmark for other CEOs to announce tangible projects; shifted the discussion from abstract potential to concrete commitments, prompting others to detail their own investments and collaborations.
Speaker: Sundar Pichai (Google)
We should share economic statistics and data on how AI is used, so that governments and companies can jointly track AI’s economic impacts, accentuate the good parts and mitigate disruptions.
Introduces the idea of collaborative data sharing as a mechanism for responsible AI rollout and economic monitoring.
Inspired subsequent remarks on transparency and data governance (e.g., Matthew Prince’s framework, Nikesh Arora’s accountability points) and added a layer of policy‑oriented discussion.
Speaker: Dario Amodei (Anthropic)
Excessive concentration of power and market concentration can lead to extractive economies; we should bet on open‑source technology to ensure everyone can access, modify, and deploy AI without external control.
Highlights a systemic risk—concentration of AI power—and proposes open‑source as a structural remedy.
Prompted other speakers to address governance and openness (e.g., Sam Altman’s democratization, Matthew Prince’s call for many AI companies, and Nikesh Arora’s focus on accountability).
Speaker: Arthur Mensch (Mistral AI)
AI has to be democratized. Countries need sovereign approaches, and the largest democracy should lead in putting tools in the hands of people and figuring out the path forward together.
Frames democratization as a national‑level responsibility and ties it to sovereignty, moving the conversation from corporate to societal scale.
Set the tone for later emphasis on inclusive access (Vinod Khosla’s Aadhaar services, Matthew Prince’s 500 k AI companies) and reinforced the need for policy frameworks.
Speaker: Sam Altman (OpenAI)
AI will be about ten times the impact of the Industrial Revolution, but happening over a decade instead of a century – essentially a 100‑fold acceleration of transformative change.
Provides a vivid, quantitative framing of AI’s potential magnitude, sharpening the urgency of the discussion.
Elevated the stakes for all participants, leading many to stress speed of execution (e.g., Nandan Nilekani’s rapid deployment story) and the necessity of robust governance.
Speaker: Demis Hassabis (DeepMind)
Data sovereignty requires safe, reliable data spaces and protecting human dignity; AI must function for people and not erode dignity, so ethics must evolve alongside technology.
Brings the often‑overlooked ethical dimension of data sovereignty and human dignity into the technical dialogue.
Reinforced the ethical thread that later speakers (Nikesh Arora, Matthew Prince) expanded into governance, accountability, and cultural preservation.
Speaker: Takahito Tokita (Fujitsu)
We need 500,000 AI companies, business models for journalists and small businesses, AI that enhances rather than homogenizes culture, and a framework that prevents the same US‑centric dominance we saw with the Internet.
Offers a concrete, multi‑point framework for inclusive AI ecosystem building, directly addressing concentration and cultural concerns.
Served as a reference point for later calls for democratization and responsible deployment, influencing the tone of responsibility taken by Rishi Sunak and Nikesh Arora.
Speaker: Matthew Prince (Cloudflare)
Governance and accountability for autonomous agents are critical; we need kill‑switches, clear responsibility, and an AI security competence centre to build these capabilities in India.
Highlights emerging technical‑risk issues (autonomous agents, security) and proposes institutional solutions.
Shifted the conversation toward concrete security measures and regulatory infrastructure, echoing earlier concerns about concentration and ethics.
Speaker: Nikesh Arora (Palo Alto Networks)
The idea of AI for cattle health was proposed on Jan 8 and went live on Feb 11 – a month later – showing how fast AI can diffuse when the government, a cooperative (Amul), and tech teams align.
Provides a real‑world, rapid‑deployment example that illustrates the speed of AI adoption possible in India.
Validated the earlier claims about acceleration, encouraging others (e.g., Rishi Sunak, Hemant Taneja) to stress speed and scalability of AI solutions.
Speaker: Nandan Nilekani (Infosys)
AI services like free tutors, doctors, and agronomists should become Aadhaar services, just as UPI did, to ensure every citizen benefits and gives democratic permission to AI.
Links AI deployment to existing national identity infrastructure, proposing a concrete pathway for universal access.
Reinforced the democratization narrative, influencing the closing remarks about inclusive AI and prompting thoughts on integrating AI with public services.
Speaker: Vinod Khosla (Khosla Ventures)
We must maintain transparency and dialogue between government and companies, ensure AI lifts the floor for humanity—especially in health and education—while also protecting safety and trust.
Summarizes the dual imperative of responsible development and inclusive benefit, framing it as a global leadership challenge.
Served as a turning point that refocused the discussion on ethical stewardship and inclusive outcomes, prompting participants to reiterate commitments to safety, regulation, and societal uplift.
Speaker: Rishi Sunak (Former UK Prime Minister)
Overall Assessment

The discussion was driven forward by a series of high‑impact interventions that moved the dialogue from aspirational rhetoric to concrete, actionable themes. Early commitments (Sundar Pichai) set a baseline of tangible investment, while Dario Amodei’s call for shared economic data introduced a collaborative governance lens. Warnings about concentration (Arthur Mensch) and the push for open‑source opened a debate on power dynamics, which was deepened by Sam Altman’s democratization framing and reinforced by Matthew Prince’s detailed ecosystem blueprint. Ethical and sovereignty concerns raised by Tokita and later expanded by Nikesh Arora added a security‑and‑accountability dimension. Nandan Nilekani’s rapid‑deployment example and Rishi Sunak’s call for transparency and inclusive uplift crystallized the urgency and responsibility of the AI rollout. Collectively, these pivotal comments redirected the conversation toward concrete investments, open‑source safeguards, sovereign data, inclusive access, and robust governance, shaping a narrative that balances ambition with responsibility.

Follow-up Questions
How can India systematically track and share the economic impacts of AI across sectors, including data on model usage and productivity effects?
Understanding AI’s acceleration of growth and potential disruptions requires shared economic statistics from companies and the government to inform policy and mitigate risks.
Speaker: Dario Amodei (Anthropic)
What measures can prevent excessive market concentration in AI and ensure broad, equitable access to AI technologies?
Excessive concentration could lead to high prices and extractive economies; open‑source approaches are suggested to democratize AI benefits.
Speaker: Arthur Mensch (Mistral AI)
How can AI be democratized through sovereign approaches that allow each country to develop and control its own AI ecosystem?
Ensuring that AI tools are in the hands of many people and nations is essential to avoid disruption and to enable inclusive benefits.
Speaker: Sam Altman (OpenAI)
What standards and technologies are needed for content authenticity and provenance (e.g., watermarking) of AI‑generated media?
Authenticity mechanisms are important for accountability, preventing misuse, and maintaining trust in AI‑created content.
Speaker: Shantanu Narayen (Adobe)
What frameworks are required for data sovereignty, safe data spaces, and ethical AI governance that protect human dignity?
Data sovereignty and ethical safeguards are critical to ensure AI serves people without compromising privacy or dignity.
Speaker: Takahito Tokita (Fujitsu)
How can cross‑border digital sovereignty be balanced with the need for technology and AI services to flow freely across countries?
Trade, security, and sovereignty concerns must be addressed to allow AI services to operate globally while protecting national interests.
Speaker: Brad Smith (Microsoft)
How can India unlock large domestic data sets for research and development while ensuring privacy and regulatory compliance?
Access to extensive data is needed to advance AI in health and other sectors, but must be managed transparently and securely.
Speaker: Roy Jakobs (Philips)
What architecture and policies are needed to build sovereign AI infrastructure that runs on modular, open systems using sovereign data?
A sovereign AI framework ensures control over AI deployment on national hardware and data, reducing dependence on foreign platforms.
Speaker: Sébastien Fabre (Safran)
What ecosystem targets and support models (e.g., for journalists, creators, small businesses, cultural preservation, and the poorest) are needed to achieve a thriving AI economy of 500,000 companies?
A comprehensive framework is required to make AI inclusive, culturally relevant, and beneficial for all economic participants.
Speaker: Matthew Prince (Cloudflare)
What governance, accountability, and safety mechanisms (including kill‑switches) are required for autonomous AI agents to prevent misuse and ensure responsibility?
As AI agents become more autonomous, clear accountability structures and safety controls are essential to avoid rogue behavior.
Speaker: Nikesh Arora (Palo Alto Networks)
How should transparent dialogue and governance between government and industry be structured to ensure AI safety and that AI benefits everyone, especially in health and education?
Maintaining trust and ensuring AI lifts the floor for all citizens requires ongoing collaboration and clear regulatory frameworks.
Speaker: Rishi Sunak (Former UK Prime Minister)
What is the impact of AI on India’s large, young workforce, and how can AI be leveraged to boost productivity without causing social disruption?
Research is needed on AI’s effects on employment and productivity to guide policies that empower workers rather than displace them.
Speaker: Hemant Taneja (General Catalyst)
How can free AI services such as tutors, doctors, and agronomists be integrated with the Aadhaar platform to provide universal access?
Embedding AI services in Aadhaar could ensure mass adoption, trust, and equitable benefits across the population.
Speaker: Vinod Khosla (Khosla Ventures)
What strategies are needed to develop AI language support and cultural nuance adaptation for India’s diverse linguistic landscape?
Effective AI deployment must address multilingual needs and cultural contexts to be inclusive and widely usable.
Speaker: Arthur Mensch (Mistral AI)

Disclaimer: This is not an official session record. DiploAI generates these resources from audiovisual recordings, and they are presented as-is, including potential errors. Due to logistical challenges, such as discrepancies in audio/video or transcripts, names may be misspelled. We strive for accuracy to the best of our ability.

Keynote-Roy Jakobs

Session at a glanceSummary, keypoints, and speakers overview

Summary

Roy Jacobs opened by asserting that artificial intelligence will have its greatest impact in healthcare because the sector urgently needs it [11-13]. He noted that rising demand, chronic disease, workforce strain and high patient expectations are putting immense pressure on health systems, which in turn accelerates AI adoption [17-19]. Jacobs pointed out that AI is already embedded in clinical workflows, augmenting expertise, improving imaging precision, enabling earlier detection and extending care beyond hospital walls [20-26].


He described the first wave of AI as focused on removing friction-automating repetitive tasks, reducing clicks and making systems more intuitive-to give clinicians back valuable time [36-42]. An autonomous MRI example illustrates AI positioning patients, selecting optimal protocols, continuously monitoring image quality and delivering precise scans, which speeds imaging, provides AI-driven insights, expands capacity, reduces variability and lowers cost [46-71].


In a smart hospital room, data from multiple devices flow into a unified platform where AI analyses vital signs, suppresses false alarms and alerts staff early to deterioration, supporting preventive action [85-87]. Jacobs emphasized that trust requires transparent, continuously validated AI operating within evolving regulatory frameworks and clear communication to clinicians and patients [87].


He highlighted India’s digital health initiatives-such as the Ayushman Bharat Digital Health Mission-that create interoperable records and longitudinal data ideal for robust AI, while the country’s diverse care settings provide a large-scale testbed for globally applicable solutions [89-103]. Philips has invested in India through innovation campuses, R&D, manufacturing and a workforce of over 4,000 engineers whose work supports both local and worldwide healthcare products [104-112]. The Philips Future Healthcare Index reports that 76 % of Indian clinicians and 79 % of patients are optimistic about AI’s ability to improve outcomes, indicating strong readiness for adoption [121-122].


Jacobs concluded that success will be measured by tangible outcomes-earlier disease detection, fewer complications and more time for care providers-positioning AI as a pivotal driver of future global health, and he described a shift from reactive, fragmented care to predictive, connected, continuous health systems that will shape the wellbeing of billions [116-122][122].


Keypoints

AI will be the most transformative force in healthcare, driven by mounting system pressures and the need for better access, early detection, and cost reduction. The CEO notes that “healthcare systems are under immense pressure… The pressure is only accelerating, and that will also accelerate the adoption of data and AI-driven innovation” and that AI is already “helping to provide better access to care… improving imaging precision… enabling earlier detection”[11-20][21-26].


The first wave of AI focuses on removing friction for clinicians by automating documentation, prioritising worklists, and making workflows more intuitive. He emphasizes that clinicians “lack… time” and that “the first wave of AI is about removing friction, automating repetitive steps, reducing clicks, supporting documentation, making systems more intuitive”[30-42].


Concrete AI-enabled solutions are already being deployed, such as autonomous MRI scanning and AI-driven smart hospital rooms that predict deterioration and cut false alarms. He illustrates an “autonomous MRI scanning” workflow that “positions her first-time right accurately and selects the optimal scan protocol… the scan is now able to see the scan” and later describes a “smart healing environment” where “AI continuously analyzes vital signs… false alarms are reduced… a subtle deterioration pattern… is detected by AI hours before it becomes visible”[46-66][80-87].


Trust, transparency, and continuous regulatory oversight are essential for AI adoption in healthcare. The speaker stresses that “Healthcare runs on trust… AI in healthcare needs to be transparent… validated continuously… clinicians need to understand how systems arrive at their recommendations… regulators need confidence that safety and efficacy is rigorously monitored”[87].


India is positioned as a global test-bed and innovation engine for AI-enabled healthcare, offering rich, longitudinal data, diverse care settings, and strong digital infrastructure. He highlights the “Ashman Bharat Digital Health Mission… interoperable health records… real-world complexity at scale… urban and rural, public and private… the diversity of care environments creates an unparalleled testbed” and notes Philips’ extensive R&D, manufacturing, and AI engineering presence across Bengaluru and Pune, with solutions that “shape… deployed around the world”[89-107][110-115].


Overall purpose:


The discussion serves to articulate Philips’ strategic vision for AI-driven healthcare, showcase current and near-future technologies, underscore the necessity of trustworthy governance, and rally Indian stakeholders as pivotal partners in scaling these innovations globally.


Overall tone:


The CEO’s tone is initially enthusiastic and persuasive, celebrating AI’s potential and Philips’ breakthroughs. It then shifts to a serious, cautionary tone when addressing trust, validation, and regulatory alignment. The speech concludes with a hopeful and collaborative tone, emphasizing partnership with India and a shared responsibility to improve billions of lives.


Speakers

Roy Jacobs – Role/Title: President and Chief Executive Officer, Royal Philips (CEO of Philips) – Area of Expertise: Healthcare technology, artificial intelligence in healthcare, medical imaging, digital health [S1][S2]


Speaker 1 – Role/Title: Moderator / event host (introducing the keynote speaker) – Area of Expertise: [S4][S6]


Additional speakers:


(none)


Full session reportComprehensive analysis and detailed insights

The moderator opened the session with a brief welcome, thanked Mr Alexander Wang for his remarks, highlighted the importance of AI-driven innovation, and then introduced Philips CEO Roy Jacobs [1-7].


Roy Jacobs began by stating that artificial intelligence will have its greatest impact in healthcare because the sector “needs it” and is under “immense pressure” from rising demand, chronic disease, stretched workforces and heightened patient expectations [8-13][16-20]. He explained that this pressure is accelerating the adoption of data-centric, AI-driven solutions that are already embedded in clinical workflows, augment human expertise, improve imaging precision, enable earlier detection and extend care beyond hospital walls [20-26].


He then described the “first wave of AI”, which is focused on removing friction for clinicians. Jacobs noted that clinicians repeatedly cite a lack of time for thinking, explaining and connecting with patients [30-33]; AI can automate repetitive documentation, reduce clicks, make systems more intuitive, and prioritize work-lists so that the most urgent cases surface without autonomous decision-making [36-43].


To illustrate the tangible benefits of this wave, Jacobs presented an autonomous MRI workflow. In the scenario a patient checks in smoothly, her clinical data are pre-loaded, AI positions her accurately, selects the optimal protocol and continuously monitors image quality, adjusting parameters in real time [46-71]. He linked this capability to Philips’ broader innovations, including a dedicated AI engine, advanced automation, and helium-free MRI hardware that makes scanners more sustainable, portable and suitable for deployment outside traditional hospitals [74-78].


Jacobs next outlined a “smart healing environment” inside a hospital room. A unified AI platform ingests data from myriad devices, continuously analyses vital signs and clinical trends, suppresses false alarms and flags subtle deterioration patterns hours before they become visible [80-87]. This “agentic AI” operates within defined guardrails, always under human oversight, turning raw data into actionable alerts for nurses and predictive insights for physicians.


Emphasising that trust is the foundation of any health-technology deployment, Jacobs warned that AI must be transparent, validated continuously-not approved just once-and must evolve with regulatory frameworks while protecting patient data [87-89].


Turning to India, Jacobs highlighted the country’s role as a global test-bed and innovation engine. The Ayushman Bharat Digital Health Mission is creating interoperable health records and longitudinal patient data, providing the “big-data play” essential for robust AI models [89-95]. India’s mix of urban and rural settings, public and private providers, and tertiary and primary care facilities offers “real-world complexity at scale”, allowing solutions built here to be resilient and globally applicable [96-103]. Philips has been present in India for 97 years and now employs more than 4,000 engineers across innovation campuses in Bengaluru and Pune, developing AI algorithms and software platforms that are deployed worldwide [104-112][113-115].


Looking ahead, Jacobs argued that success will be judged by measurable health outcomes rather than the number of algorithms deployed: earlier disease detection, fewer avoidable complications, shorter waiting times, greater access and more clinician-patient interaction [116-121]. He cited the Philips Future Healthcare Index, which shows that 76 % of Indian healthcare professionals and 79 % of patients are optimistic that AI can improve outcomes [121-122].


In his closing remarks, Jacobs reiterated that AI’s greatest impact will be realised when it delivers concrete benefits to billions of lives, shifting healthcare from a reactive, fragmented model to a predictive, connected and continuous system. He framed this transformation as a shared responsibility and an “exciting opportunity” that Philips is committed to pursuing together with partners in India and around the world [122-125].


Session transcriptComplete transcript of the session
Speaker 1

Thank you, Mr. Alexander Wang. Our heartfelt gratitude to you for your remarks, for sharing your findings as well. And well, your vision for AI and innovation, it does inspire many of us. Well, ladies and gentlemen, please join me now in welcoming Mr. Roy Jacobs, CEO of Philips. Mr. Jacobs leads one of the world’s most storied healthcare technology companies through a pivotal transformation, putting AI diagnostics and patient monitoring at the center of Philips’ mission. His work sits at the critical intersection of AI and human health, where the stakes could not be higher. Ladies and gentlemen, please welcome CEO of Philips, Mr. Roy Jacobs.

Roy Jacobs

Thank you so much. Good afternoon. It’s a true honor to be here among so many brilliant minds and very bold ambitions. and building of the visionary path of the Prime Minister, I want to share how we are walking that path, designing and developing in India and delivering to the world. And since I have your attention, let me share something that I want to leave behind. We believe that artificial intelligence will have its biggest impact in healthcare. Because healthcare needs it. And maybe some of you are skeptical, but let me try to convince you otherwise. For decades, healthcare has been adopting technology cautiously. And there are different reasons for this. For today, something fundamental has changed in healthcare.

Healthcare systems are under immense pressure. Rising demand. chronic disease, stretched workforces, and high expectations of patients and society. The pressure is only accelerating, and that will also accelerate the adoption of data and AI -driven innovation. AI is already today helping to provide better access to care. It’s embedded in clinical workflows. It’s augmenting human expertise. It’s improving imaging precision. It’s enabling earlier detection. It’s extending care beyond walls. And so the task now is, how do we build an intelligent care system that is predictive, trusted, more effective, and accessible for people everywhere? Thank you. Because ultimately, it’s not about AI or technology. It’s about the people that are served by technology and AI. Just consider for a moment the pressure care teams are under.

If you ask clinicians what they lack most, the answer is almost always time. Time to think. Time to explain. Time to connect with patients. In many health systems, nurses and physicians spend hours on documentation and administrative tasks. Hours not spent on the patient. AI has the full potential to change that equation. The first wave of AI is about removing friction, automating repetitive steps, reducing clicks, supporting documentation, making systems more intuitive. The first wave of AI is about making systems more intuitive. This may not be flashy, but it is transformative and it does generate huge impact. When AI listens to a clinical conversation and drafts structured documentation in the background that is not replacing a clinician, it’s giving time back to the clinician.

When AI prioritizes worklists so that the most urgent cases rise to the top, it’s not making decisions independently. It’s supporting better clinical judgment. This is AI working quietly in the background so clinicians can focus on what matters most. And this is real today. Let me make it even more tangible. Because AI is… is already reshaping what’s possible across healthcare. Imagine a patient in a regional hospital. She has been waiting for weeks for an MRI scan. The backlog is long. The technologists are stressed. Clinicians are not available to support. Traditionally, an MRI required significant manual setup, precise parameter selection, and highly specialized expertise to conduct a scan. Variability was inevitable and access was limited. Now imagine something completely different.

The patient arrives for her MRI and checks in smoothly. Her clinical information is already available to the system that will conduct the scan. AI helps position her first -time right accurately and selects the optimal scan protocol. It’s tailored to her autonomy. And as the scan runs, the image quality is continuously monitored and adjusted automatically. The scan is now able to see the scan. The scan is now able to see the scan. The scan is now able to see the scan. The scan is now able to see the scan. The scan is now able to see the scan. This is autonomous MRI scanning. The result? Clear, precise images, delivered efficiently and consistently. The radiologist receives AI -driven insights and quantitative biomarkers.

And Philips is working to make that a reality. It means more patients diagnosed earlier. Earlier detection of chronic and complex conditions. It expands capacity and access to care. It reduces variability. It scales expertise. And over time, it helps reduce the overall cost of care. And at Philips, we have been laying the groundwork for this future. Through advanced automation, integrated MRI workflows, and an AI engine, that scans faster while enhancing quality. We are also leaders in helium -free MRI systems. That’s a hardware breakthrough. And together, that does not only make it more sustainable, but it allows us to install MRI systems outside of the hospital, closer to patients, where it was previously not visible to go to.

That means even more access, more resilience, and more patients served. But diagnosis is not the only part of the story. Let’s enter a hospital room. There is no lack of data. Actually, there’s an overload of data. Many devices, alarms going off, dashboards to be read. Now image a hospital room transformed into a smart healing environment. Data from devices flow into a unified platform. Where AI continuously analyzes vital signs and clinical trends. False alarms are reduced True risks are elevated early A subtle deterioration pattern in patient is detected by AI hours before it comes visible The nurse is alerted within context, not with noise And the physician receives predictive insights Preventative action can and will be taken The true patient crisis never fully materializes That is the power of agentic AI Software that can perceive, reason and act within defined guardrails Always under human oversight It reduces burden It anticipates risks And it turns data into action That’s how we can give time back With all of this With all of this excitement We must be clear about one thing Healthcare runs on trust AI in healthcare needs to be transparent It must be validated continuously Not approved just once It must operate within regulatory frameworks And those evolve as the technology evolves Clinicians need to understand how systems arrive at their recommendations Patients need to know how their data is protected And regulators need confidence that safety and efficacy is rigorously monitored all the time Innovation and governance must advance together With speed Because trust determines adoption And adoption determines the success of valuable application of AI in healthcare If they move at different speeds Trust erodes If they move in alignment, adoption accelerates.

Now, I’m very happy to be back in India. India represents a remarkable opportunity in this transformation, standing at the intersection of skill, digital infrastructure, and ambition. The country’s digital infrastructure, including initiatives under the Ashman Bharat Digital Health Mission, is laying the foundation for interoperable health records and longitudinal patient data, the foundation for big data play. Unique health ideas and digital registries create the possibility for the continuity of care for population at scale. This matters enormously, as we have heard earlier today. AI systems thrive on structured, high -quality longitudinal data. When patients can be followed across settings, AI systems thrive on structured, high -quality longitudinal data. From primary care, to hospital, to diagnostics, and to the home, the power of predictive analytics increases dramatically.

But India also brings something else to the table. Real -world complexity at scale. Urban and rural settings. Public and private systems. High -end tertiary care systems and primary health centers. The diversity of care environments creates an unparalleled testbed for scalable, resilient solutions. Solutions built for India’s scale and constraints have the potential to inform global models of care. And at Philips, we see India as a global innovation engine. We are already here for 97 years. And through our Philips Innovation Campus in Bengaluru and the Healthcare Innovation Center in Pune, we have made significant investments in R &D, manufacturing, digital platforms, AI engineering and clinical collaboration. Our teams in India contribute to global development as much as to local development.

And they do that across imaging, monitoring and connected care. The work done here does not stay here alone. It shapes solutions deployed around the world. For example, AI algorithms developed and validated with diverse data sets here improve robustness across geographies. That’s done by more than 4 ,000 engineers that we have in India, developing for India and for the world. These software platforms, engineered here, connect ecosystems in multiple markets. Clinical workflow solutions co -created with Indian partners inform designs that scale globally, exactly as the Prime Minister says we are doing that today. And this is the kind of distributed innovation model that healthcare needs. Not isolated breakthroughs, but integrated, globally connected ecosystems. If we look ahead, in 10 years, success will not be defined by the number of algorithms deployed.

It will be defined by the outcomes they generate. Earlier detection of disease. Fewer avoidable complications. Shorter waiting times. Greater access. More times for clinicians, nurses, and technicians. The Philips Future Healthcare Index, one of the world’s largest recurring health research initiatives, tells a clear story. 76 % of Indian healthcare professionals are optimistic that AI can help them improve patient outcomes. 79 % of the Indian patients are optimistic that AI can improve their health. Their personal healthcare. This shows us that patients in India are ready for this And that Indian healthcare professionals are asking for this And so are policymakers and government as we heard today The promise is real It’s for today Healthcare will move from reactive to predictive From fragmented to connected From episodic to continuous The systems we build today will shape the health of billions tomorrow So let me return to where I started AI will have its greatest impact in healthcare for the world And when we look at this in a decade from now We will look at the outcomes and impact it has delivered It will look at the outcomes and impact it has delivered It will not be remembered for what is optimized on the screen but for the billions of lives that we could improve with it.

That’s the responsibility. That’s the exciting opportunity. And that’s the future that we are committed to building together with all our partners, together here in India and across the world. Thank you so much.

Related ResourcesKnowledge base sources related to the discussion topics (10)
Factual NotesClaims verified against the Diplo knowledge base (5)
Confirmedmedium

“Alexander Wang is a representative from Meta who participated in the session.”

The knowledge base lists Alexander Wang as a Representative from Meta in the Leaders’ Plenary session [S25].

Confirmedmedium

“Roy Jacobs is the CEO of Philips.”

Roy Jacobs is identified as President and Chief Executive Officer of Royal Philips (Healthcare/Medical Technology) in the World Economic Forum panel discussion [S7].

Confirmedhigh

“Roy Jacobs stated that artificial intelligence will have its greatest impact in healthcare because the sector desperately needs technology.”

The knowledge base records Jacobs arguing that AI will have its greatest impact in healthcare because the sector desperately needs technology [S2] and [S3].

Confirmedhigh

“Jacobs described an autonomous MRI workflow where the patient checks in smoothly, clinical data are pre‑loaded, AI positions the patient accurately, selects the optimal protocol, and continuously monitors image quality, adjusting parameters in real time.”

A similar scenario is detailed in the knowledge base, describing smooth check-in, pre-loaded clinical information, AI-assisted positioning, optimal protocol selection, and real-time image-quality monitoring [S28].

Additional Contextlow

“AI‑driven imaging solutions improve image quality and streamline clinical workflow.”

Philips has launched an AI-powered spectral CT system that integrates AI across the imaging chain to enhance image quality, reduce noise, and streamline workflow, providing additional context to the claim about AI improving imaging precision [S30].

External Sources (30)
S1
Cracking the Code of Digital Health / DAVOS 2025 — – Roy Jakobs: President and Chief Executive Officer, Royal Philips 2. Data Liquidity and Interoperability: Roy Jakobs h…
S2
Keynote-Roy Jakobs — The discussion features Roy Jakobs, CEO of Philips, presenting his vision for artificial intelligence’s transformative r…
S3
Keynote-Roy Jakobs — The discussion features Roy Jakobs, CEO of Philips, presenting his vision for artificial intelligence’s transformative r…
S4
Keynote-Martin Schroeter — -Speaker 1: Role/Title: Not specified, Area of expertise: Not specified (appears to be an event moderator or host introd…
S5
Responsible AI for Children Safe Playful and Empowering Learning — -Speaker 1: Role/title not specified – appears to be a student or child participant in educational videos/demonstrations…
S6
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Vijay Shekar Sharma Paytm — -Speaker 1: Role/Title: Not mentioned, Area of expertise: Not mentioned (appears to be an event host or moderator introd…
S7
Scaling AI Beyond Pilots: A World Economic Forum Panel Discussion — – Amin Nasser- Julie Sweet – Julie Sweet- Amin Nasser Development | Economic Focus on outcomes and value creation rat…
S8
Agentic AI and the new industrial diplomacy — How this looks in practice:The European AI Act,which came into force in2024, classifies many industrial AI systems as ‘h…
S9
Agentic AI in Focus Opportunities Risks and Governance — Thank you. So NetApp actually, as you said, multi -cloud, we both power public cloud as well as private cloud. Many of t…
S10
AI for Bharat’s Health_ Addressing a Billion Clinical Realities — Soi explains that while the long-term goal is institutional adoption where AI becomes intrinsic to the organization, cur…
S11
How Trust and Safety Drive Innovation and Sustainable Growth — Summary:All speakers agreed that trust is the foundational requirement for AI adoption. Without trust, people simply won…
S12
Digital Health at the crossroads of human rights, AI governance, and e-trade (SouthCentre) — Technological innovation has led to a significant transformation in health systems, particularly through advancements in…
S13
Panel Discussion AI in Healthcare India AI Impact Summit — Chris argues that AI can significantly reduce the administrative workload that currently consumes 70% of clinicians’ tim…
S14
WS #288 An AI Policy Research Roadmap for Evidence-Based AI Policy — Jason Tucker: Thank you. So I wear two hats. I’m an academic, but I also work in public policy. And this is why I’m sort…
S15
Keynote-Roy Jakobs — Healthcare systems are under immense pressure. Rising demand. chronic disease, stretched workforces, and high expectatio…
S16
Keynote-Roy Jakobs — Thank you so much. Good afternoon. It’s a true honor to be here among so many brilliant minds and very bold ambitions. a…
S17
AI for Bharat’s Health_ Addressing a Billion Clinical Realities — Soi explains that while the long-term goal is institutional adoption where AI becomes intrinsic to the organization, cur…
S18
Panel Discussion AI in Healthcare India AI Impact Summit — Evidence:In the U.S., only 30% of a clinician’s time is spent on patient care, with the rest on paperwork and administra…
S19
AI Automation in Telecom_ Ensuring Accountability and Public Trust India AI Impact Summit 2026 — Discussion point:Predictive network maintenance and self-healing systems
S20
How Small AI Solutions Are Creating Big Social Change — Antoine Tesniere explains that healthcare has a long history with AI, and many small AI models are already validated and…
S21
Catalyzing Global Investment in AI for Health_ WHO Strategic Roundtable — Verified AI extends beyond accuracy to encompass complete transparency in decision-making processes. Brey advocated for …
S22
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Lt Gen Vipul Shinghal — Third, ensuring transparency in AI systems:Commanders must understand the data sources, training methodologies, and deci…
S23
From principles to practice: Governing advanced AI in action — Juha argues that trust is the sine qua non for AI technology uptake, which in turn is necessary for AI benefits to mater…
S24
Cracking the Code of Digital Health / DAVOS 2025 — Key points included the need for better data liquidity and interoperability to fully leverage AI’s potential in healthca…
S25
Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session — -Alexander Wang- Representative from Meta -Roy Jakobs- CEO, Philips
S26
Thinking through Augmentation — While Ucuzoglu is optimistic about the long-term impact of transformative technology, he acknowledges that it is not an …
S27
Skilling and Education in AI — Thank you so much. It is a very, very pertinent question. And I truly believe that in the past, we have been able to do …
S28
https://dig.watch/event/india-ai-impact-summit-2026/keynote-roy-jakobs — The patient arrives for her MRI and checks in smoothly. Her clinical information is already available to the system that…
S29
AI performance in image-based medical diagnosis is equivalent to human performance, study finds — A review conducted by UK researchers argues that,when it comes to classifying diseases using medical images, the perform…
S30
Philips launches AI-powered spectral CT system — Philips hasunveiled Verida, the world’s first detector-based spectral CT fully powered by AI. The system integrates AI a…
Speakers Analysis
Detailed breakdown of each speaker’s arguments and positions
R
Roy Jacobs
14 arguments114 words per minute1672 words872 seconds
Argument 1
AI will have its biggest impact because healthcare is under immense pressure and needs efficiency (Roy Jacobs)
EXPLANATION
Jacobs argues that the unprecedented pressure on health systems—rising demand, chronic disease, workforce strain and high patient expectations—creates a critical need for AI to improve efficiency. He believes this pressure will accelerate AI adoption, making AI’s impact in healthcare the greatest of any sector.
EVIDENCE
He states that “We believe that artificial intelligence will have its biggest impact in healthcare. Because healthcare needs it” and explains that “Healthcare systems are under immense pressure. Rising demand, chronic disease, stretched workforces, and high expectations of patients and society. The pressure is only accelerating, and that will also accelerate the adoption of data and AI-driven innovation” [11-13][16-20].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
The keynote notes that healthcare systems face rising demand, chronic disease burden, stretched workforces and high patient expectations, which accelerates AI adoption [S3].
MAJOR DISCUSSION POINT
AI’s transformative impact on healthcare
AGREED WITH
Speaker 1
Argument 2
AI is already improving access, embedding in clinical workflows, enhancing imaging precision, and enabling earlier detection (Roy Jacobs)
EXPLANATION
Jacobs lists concrete ways AI is already adding value: expanding patient access, becoming part of everyday clinical processes, sharpening imaging quality, and supporting earlier disease detection. These examples illustrate that AI’s benefits are not theoretical but already observable in health care.
EVIDENCE
He enumerates that “AI is already today helping to provide better access to care. It’s embedded in clinical workflows. It’s augmenting human expertise. It’s improving imaging precision. It’s enabling earlier detection” [20-25].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Jacobs explicitly states that AI is helping provide better access, is embedded in clinical workflows, improves imaging precision and enables earlier detection [S3].
MAJOR DISCUSSION POINT
AI’s transformative impact on healthcare
Argument 3
AI automates documentation and repetitive steps, giving clinicians more time for patient care (Roy Jacobs)
EXPLANATION
Jacobs describes the first wave of AI as removing friction by automating routine documentation and other repetitive tasks, thereby freeing clinicians to focus on direct patient interaction. This automation is presented as a quiet but powerful transformation.
EVIDENCE
He explains that “The first wave of AI is about removing friction, automating repetitive steps, reducing clicks, supporting documentation, making systems more intuitive” and gives the example that “When AI listens to a clinical conversation and drafts structured documentation in the background… it’s giving time back to the clinician” [36-40].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
The first wave of AI is described as removing friction by automating repetitive steps, reducing clicks and supporting documentation [S3]; a World Economic Forum panel also highlights ambient listening technology that automates note-taking for clinicians [S7].
MAJOR DISCUSSION POINT
First wave of AI: automation and workflow friction reduction
Argument 4
AI prioritizes worklists to support, not replace, clinical judgment, improving decision‑making efficiency (Roy Jacobs)
EXPLANATION
Jacobs highlights that AI can reorder worklists so urgent cases surface first, assisting clinicians without making autonomous decisions. This support enhances decision‑making speed while preserving clinician authority.
EVIDENCE
He notes that “When AI prioritizes worklists so that the most urgent cases rise to the top, it’s not making decisions independently. It’s supporting better clinical judgment” [40-42].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Jacobs explains that AI can reorder worklists so urgent cases rise to the top, supporting better clinical judgment without making independent decisions [S3].
MAJOR DISCUSSION POINT
First wave of AI: automation and workflow friction reduction
Argument 5
AI positions patients, selects optimal scan protocols, and continuously monitors image quality for autonomous MRI scanning (Roy Jacobs)
EXPLANATION
Jacobs describes an autonomous MRI workflow where AI handles patient positioning, chooses the best scan protocol, and continuously adjusts image quality during acquisition, eliminating many manual steps and variability.
EVIDENCE
He paints the scenario: “The patient arrives… AI helps position her first-time right accurately and selects the optimal scan protocol. As the scan runs, the image quality is continuously monitored and adjusted automatically. This is autonomous MRI scanning” [54-64].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
The autonomous MRI scenario-patient positioning, protocol selection and real-time image quality monitoring-is detailed in the keynote [S3].
MAJOR DISCUSSION POINT
Autonomous MRI and imaging breakthroughs
Argument 6
Helium‑free MRI hardware makes scanners more sustainable and portable, expanding access beyond traditional hospital settings (Roy Jacobs)
EXPLANATION
Jacobs points out that Philips’ helium‑free MRI technology reduces environmental impact and enables installation of scanners outside conventional hospitals, thereby reaching patients in remote or underserved locations.
EVIDENCE
He states that “We are also leaders in helium-free MRI systems. That’s a hardware breakthrough. And together, that does not only make it more sustainable, but it allows us to install MRI systems outside of the hospital, closer to patients, where it was previously not visible to go to” [75-78].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Jacobs highlights helium-free MRI as a hardware breakthrough that reduces environmental impact and enables installation outside hospitals, expanding patient reach [S3].
MAJOR DISCUSSION POINT
Autonomous MRI and imaging breakthroughs
Argument 7
Unified AI platform consolidates device data, reduces false alarms, and detects subtle patient deterioration hours before it becomes visible (Roy Jacobs)
EXPLANATION
Jacobs envisions a smart hospital room where data from many devices feed into a single AI platform that filters noise, cuts false alarms, and identifies early signs of patient decline, enabling proactive care.
EVIDENCE
He describes that “Data from devices flow into a unified platform where AI continuously analyzes vital signs and clinical trends. False alarms are reduced. A subtle deterioration pattern in patient is detected by AI hours before it becomes visible” [85-88].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
The smart-room concept describes a unified AI platform that filters device data, cuts false alarms and flags early deterioration patterns hours in advance [S3].
MAJOR DISCUSSION POINT
AI‑driven predictive monitoring in smart hospital rooms
Argument 8
Agentic AI operates within defined guardrails under human oversight, turning data into actionable alerts for nurses and physicians (Roy Jacobs)
EXPLANATION
Jacobs stresses that the AI acting in the smart room is ‘agentic’—it can perceive, reason, and act—but always within pre‑set limits and under clinician supervision, converting raw data into contextual alerts.
EVIDENCE
He explains that “That is the power of agentic AI. Software that can perceive, reason and act within defined guardrails. Always under human oversight. It reduces burden. It anticipates risks. And it turns data into action” [85-88].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Jacobs defines agentic AI as perceiving, reasoning and acting within guardrails under human oversight [S3]; the EU AI Act’s high-risk classification and required human-in-the-loop controls provide regulatory context for such guardrails [S8].
MAJOR DISCUSSION POINT
AI‑driven predictive monitoring in smart hospital rooms
Argument 9
AI systems must be transparent, continuously validated, and compliant with evolving regulatory frameworks (Roy Jacobs)
EXPLANATION
Jacobs argues that trust in AI hinges on transparency, ongoing validation, and alignment with regulatory standards that evolve alongside the technology, ensuring safety and efficacy.
EVIDENCE
He asserts that “AI in healthcare needs to be transparent. It must be validated continuously, not approved just once. It must operate within regulatory frameworks that evolve as the technology evolves” [84-87].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
The keynote stresses the need for continuous validation, transparency and alignment with evolving regulatory frameworks [S3]; the European AI Act further mandates transparency documentation for high-risk AI systems [S8].
MAJOR DISCUSSION POINT
Trust, transparency, and regulatory alignment
Argument 10
Alignment of innovation speed with governance builds trust, which is essential for adoption and successful AI deployment (Roy Jacobs)
EXPLANATION
Jacobs emphasizes that when the pace of AI innovation matches the development of governance and oversight, trust is maintained, leading to faster adoption and effective deployment of AI solutions.
EVIDENCE
He notes that “Innovation and governance must advance together with speed. Because trust determines adoption and adoption determines the success of valuable application of AI in healthcare. If they move at different speeds trust erodes; if they move in alignment, adoption accelerates” [84-87].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Jacobs argues that synchronized innovation and governance preserve trust and accelerate adoption [S3].
MAJOR DISCUSSION POINT
Trust, transparency, and regulatory alignment
Argument 11
India’s digital health infrastructure, diverse care settings, and large population provide a unique testbed for scalable AI solutions (Roy Jacobs)
EXPLANATION
Jacobs highlights India’s extensive digital health initiatives, varied urban‑rural and public‑private environments, and massive population as an ideal laboratory for developing AI solutions that can scale globally.
EVIDENCE
He points out that “India’s digital infrastructure, including initiatives under the Ashman Bharat Digital Health Mission, is laying the foundation for interoperable health records… The diversity of care environments creates an unparalleled testbed for scalable, resilient solutions” [89-102].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
The keynote references India’s digital health initiatives (e.g., Ayushman Bharat Digital Health Mission), interoperable records and diverse care environments as an ideal testbed [S2].
MAJOR DISCUSSION POINT
India as a strategic testbed and innovation engine
Argument 12
Indian R&D teams develop AI algorithms and platforms that improve global robustness and are deployed worldwide (Roy Jacobs)
EXPLANATION
Jacobs states that engineers and researchers in India create AI models and software that, because of diverse local data, are more robust across geographies and are integrated into Philips’ global product portfolio.
EVIDENCE
He gives the example that “AI algorithms developed and validated with diverse data sets here improve robustness across geographies… more than 4,000 engineers… Clinical workflow solutions co-created with Indian partners inform designs that scale globally” [110-114].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Jacobs notes that AI algorithms built and validated with diverse Indian data improve robustness across geographies and are integrated into Philips’ global portfolio, involving thousands of engineers [S3].
MAJOR DISCUSSION POINT
India as a strategic testbed and innovation engine
Argument 13
Success will be judged by outcomes such as earlier disease detection, fewer complications, shorter wait times, and greater access (Roy Jacobs)
EXPLANATION
Jacobs argues that the true metric of AI’s value will be tangible health outcomes—earlier diagnoses, reduced complications, faster service, and broader access—rather than the sheer number of algorithms deployed.
EVIDENCE
He says “If we look ahead, in 10 years, success will not be defined by the number of algorithms deployed. It will be defined by the outcomes they generate: earlier detection of disease, fewer avoidable complications, shorter waiting times, greater access” [116-121].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Future success metrics are described as outcomes-earlier detection, fewer complications, reduced wait times and expanded access [S3]; a World Economic Forum panel also emphasizes outcome-focused value creation over mere technology deployment [S7].
MAJOR DISCUSSION POINT
Future success measured by outcomes, not algorithm count
Argument 14
Survey data shows high optimism among Indian healthcare professionals (76%) and patients (79%) that AI will improve health outcomes (Roy Jacobs)
EXPLANATION
Jacobs cites a Philips research index indicating that a strong majority of Indian clinicians and patients are confident that AI can enhance health outcomes, underscoring readiness for AI adoption.
EVIDENCE
He reports that “The Philips Future Healthcare Index… tells a clear story. 76 % of Indian healthcare professionals are optimistic that AI can help them improve patient outcomes. 79 % of the Indian patients are optimistic that AI can improve their health” [121-122].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
The Philips Future Healthcare Index reports 76% of Indian clinicians and 79% of patients are optimistic about AI’s impact on health outcomes [S3].
MAJOR DISCUSSION POINT
Future success measured by outcomes, not algorithm count
Agreements
Agreement Points
AI is seen as a critical driver for improving healthcare efficiency and outcomes
Speakers: Speaker 1, Roy Jacobs
AI will have its biggest impact because healthcare is under immense pressure and needs efficiency (Roy Jacobs)
Speaker 1 praised the vision for AI and innovation, while Roy Jacobs argued that AI will have its biggest impact in healthcare due to mounting system pressures and the need for efficiency [4-7][11-13][16-20].
POLICY CONTEXT (KNOWLEDGE BASE)
This consensus aligns with policy discussions on AI governance in health, as highlighted in the South-Centre report on digital health, human rights and e-trade [S12], and was reinforced at the AI Impact Summit where AI was shown to reduce clinicians’ administrative workload and improve outcomes [S13]. It also mirrors the broader framing in AI policy roadmaps that AI will revolutionise healthcare [S14].
Similar Viewpoints
Both arguments emphasize that the first wave of AI removes workflow friction by automating routine tasks and intelligently prioritising work, thereby freeing clinicians to focus on patient interaction [36-40].
Speakers: Roy Jacobs
AI automates documentation and repetitive steps, giving clinicians more time for patient care (Roy Jacobs) AI prioritizes worklists to support, not replace, clinical judgment, improving decision‑making efficiency (Roy Jacobs)
Both arguments stress that trust, transparency, and synchronized regulatory governance are essential for the successful adoption of AI in healthcare [84-87].
Speakers: Roy Jacobs
AI systems must be transparent, continuously validated, and compliant with evolving regulatory frameworks (Roy Jacobs) Alignment of innovation speed with governance builds trust, which is essential for adoption and successful AI deployment (Roy Jacobs)
Unexpected Consensus
Overall Assessment

The discussion shows limited but clear consensus: both speakers acknowledge AI’s pivotal role in healthcare, and Roy Jacobs repeatedly reinforces themes of workflow automation and the necessity of trust and governance. The agreement is concentrated around AI’s potential to alleviate system pressures and improve outcomes, while the rest of the talk expands on technical and strategic details.

Moderate consensus – agreement exists on the overarching importance of AI for healthcare and on the need for trustworthy, well‑governed deployment, but detailed technical visions are presented primarily by Roy Jacobs.

Differences
Different Viewpoints
Unexpected Differences
Overall Assessment

The transcript contains an introductory segment by Speaker 1 that only offers thanks and welcomes Roy Jacobs ([1-7]), followed by a lengthy keynote from Roy Jacobs presenting a unified vision of AI’s role in healthcare ([8-125]). No opposing statements or contrasting viewpoints are presented by any speaker, resulting in an absence of identifiable disagreement or partial disagreement.

Minimal – the discussion is essentially a one‑sided presentation, so there is no substantive conflict that could affect consensus on the topics of AI, healthcare transformation, or related policy considerations.

Takeaways
Key takeaways
AI will have its greatest impact in healthcare because the sector faces mounting pressure from demand, chronic disease, workforce strain, and patient expectations. The first wave of AI focuses on automating documentation and repetitive tasks, reducing workflow friction and giving clinicians more time for patient care. AI‑driven autonomous MRI can position patients, select optimal protocols, and continuously monitor image quality, while helium‑free hardware makes scanners more sustainable and portable. Smart hospital rooms using a unified AI platform can consolidate device data, cut false alarms, and detect early signs of patient deterioration, delivering actionable alerts under human oversight. Trust, transparency, continuous validation, and alignment with evolving regulatory frameworks are essential for AI adoption in healthcare. India’s digital health infrastructure, diverse care settings, and large population make it an ideal testbed and innovation engine for scalable AI solutions that can be exported globally. Future success will be measured by health outcomes—earlier detection, fewer complications, reduced wait times, and increased access—rather than the number of algorithms deployed.
Resolutions and action items
None identified
Unresolved issues
How to ensure continuous, real‑time validation of AI models while keeping pace with rapid innovation. Establishing clear regulatory pathways that evolve alongside AI technology without creating trust gaps. Addressing data privacy and security concerns as AI systems ingest large volumes of patient data across disparate settings. Managing the integration of AI tools into heterogeneous Indian healthcare environments (urban vs. rural, public vs. private). Overcoming lingering skepticism among clinicians and patients regarding AI’s role and reliability.
Suggested compromises
Balancing rapid AI innovation with equally rapid regulatory and governance processes to maintain trust and accelerate adoption. Deploying AI as supportive, non‑replacing tools (e.g., documentation assistance, work‑list prioritization) that operate within defined guardrails under human oversight.
Thought Provoking Comments
We believe that artificial intelligence will have its biggest impact in healthcare. Because healthcare needs it.
Sets a bold, overarching premise that frames the entire discussion, positioning AI not as a peripheral tool but as a core solution to systemic healthcare challenges.
Establishes the central thesis, prompting the audience to view subsequent examples and arguments through the lens of AI’s essential role, thereby steering the conversation toward concrete healthcare applications.
Speaker: Roy Jacobs
The first wave of AI is about removing friction, automating repetitive steps, reducing clicks, supporting documentation, making systems more intuitive. This may not be flashy, but it is transformative and it does generate huge impact.
Shifts focus from high‑profile, headline‑grabbing AI feats to the everyday, productivity‑driving benefits that directly affect clinicians’ workloads.
Redirects the narrative toward practical, near‑term implementations, encouraging listeners to consider AI’s value in routine clinical work rather than only in breakthrough technologies.
Speaker: Roy Jacobs
Imagine a patient in a regional hospital… AI helps position her first‑time right accurately and selects the optimal scan protocol… The scan is now able to see the scan… This is autonomous MRI scanning.
Provides a vivid, concrete scenario that illustrates how AI can overhaul a complex workflow (MRI imaging), making the abstract concept of AI‑driven care tangible.
Creates a turning point where the discussion moves from abstract benefits to a specific, visualizable use case, sparking interest in technical feasibility and patient‑centric outcomes.
Speaker: Roy Jacobs
AI continuously analyzes vital signs and clinical trends… False alarms are reduced, true risks are elevated early… The nurse is alerted within context, not with noise… The physician receives predictive insights… The true patient crisis never fully materializes.
Introduces the concept of “agentic AI” that not only informs but also acts within defined guardrails, highlighting a shift from decision‑support to proactive risk mitigation.
Deepens the conversation by adding a layer of complexity—how AI can intervene autonomously—prompting considerations of safety, oversight, and the balance between automation and human control.
Speaker: Roy Jacobs
Healthcare runs on trust. AI in healthcare needs to be transparent. It must be validated continuously, not approved just once. Clinicians need to understand how systems arrive at their recommendations. Patients need to know how their data is protected. Regulators need confidence that safety and efficacy are rigorously monitored all the time.
Challenges the optimistic narrative by foregrounding ethical, regulatory, and trust issues that are critical for real‑world adoption.
Marks a turning point where the tone shifts from enthusiasm to caution, prompting the audience to contemplate governance frameworks and the necessity of aligning innovation speed with regulatory oversight.
Speaker: Roy Jacobs
India represents a remarkable opportunity… its digital infrastructure, diverse care environments, and real‑world complexity at scale make it an unparalleled testbed for scalable, resilient solutions. Solutions built for India’s scale can inform global models of care.
Reframes the discussion geographically, positioning India not just as a market but as a strategic innovation engine that can shape worldwide healthcare AI.
Introduces a new thematic focus on localization and global scalability, encouraging participants to think about cross‑regional data diversity, implementation challenges, and the exportability of solutions.
Speaker: Roy Jacobs
Success will not be defined by the number of algorithms deployed but by the outcomes they generate—earlier detection of disease, fewer avoidable complications, shorter waiting times, greater access, more time for clinicians.
Shifts the metric of progress from quantitative deployment to qualitative health outcomes, urging a results‑oriented mindset.
Steers the conversation toward impact measurement and accountability, influencing how stakeholders might evaluate AI projects and prioritize investments.
Speaker: Roy Jacobs
76 % of Indian healthcare professionals are optimistic that AI can help them improve patient outcomes. 79 % of Indian patients are optimistic that AI can improve their health.
Provides data‑driven validation of the earlier optimism expressed, grounding the vision in measurable stakeholder sentiment.
Reinforces the earlier claims about readiness and acceptance, bolstering credibility and encouraging the audience to view AI adoption as both feasible and welcomed.
Speaker: Roy Jacobs
Overall Assessment

Roy Jacobs’ remarks structured the discussion around a clear, progressive narrative: starting with a bold claim about AI’s pivotal role in healthcare, moving through practical, everyday benefits, then illustrating transformative use cases (autonomous MRI, predictive monitoring), before confronting the essential issues of trust, regulation, and ethical deployment. By inserting the India‑specific testbed narrative, he broadened the scope from a global vision to a concrete, actionable platform for innovation. Finally, his shift to outcome‑based success metrics reframed the conversation from technology‑centric to impact‑centric. These pivotal comments collectively redirected the audience’s focus, deepened the analytical depth of the dialogue, and set a roadmap that balances ambition with responsibility.

Follow-up Questions
How can AI systems in healthcare be made transparent and continuously validated to maintain trust?
Ensuring ongoing validation and transparency is critical for clinician and patient confidence, which drives adoption of AI solutions.
Speaker: Roy Jacobs
How can regulatory frameworks evolve in step with rapidly advancing AI technologies to ensure safety and efficacy?
Regulators must keep pace with AI innovations to prevent trust erosion and to enable timely deployment of beneficial tools.
Speaker: Roy Jacobs
What strategies are needed to protect patient data while allowing AI-driven insights?
Balancing data privacy with the need for high‑quality longitudinal data is essential for building robust AI models.
Speaker: Roy Jacobs
How can AI be integrated into clinical workflows to reduce documentation burden without adding new complexities?
Effective workflow integration can give clinicians back time, a primary need identified in the discussion.
Speaker: Roy Jacobs
What methods can be employed to collect and standardize structured, high‑quality longitudinal health data across diverse Indian settings?
India’s varied care environments provide a testbed, but consistent data is required for AI scalability and accuracy.
Speaker: Roy Jacobs
What metrics should be used to evaluate AI success based on patient outcomes rather than the number of algorithms deployed?
Shifting focus to measurable health outcomes will better demonstrate AI’s real-world value.
Speaker: Roy Jacobs
How can AI solutions developed and validated in India be scaled and adapted for global deployment?
India’s scale and complexity offer insights, but research is needed to confirm transferability to other markets.
Speaker: Roy Jacobs
What are the long‑term impacts of autonomous MRI scanning on diagnostic accuracy, cost, and patient access?
Understanding the clinical and economic effects of autonomous MRI is vital before widespread rollout.
Speaker: Roy Jacobs
How can AI reduce false alarms and elevate true risks in smart hospital rooms to improve patient safety?
Optimizing alarm management can lessen clinician fatigue and enable earlier intervention.
Speaker: Roy Jacobs
What algorithms and data sources are needed for AI to detect subtle patient deterioration patterns hours before they become clinically visible?
Early detection can prevent crises, but requires research into predictive signals and validation.
Speaker: Roy Jacobs
How can the Philips Future Healthcare Index be expanded to track actual outcome improvements over time rather than just optimism levels?
Linking index data to concrete health outcomes will validate AI’s impact and guide future investments.
Speaker: Roy Jacobs
What quantitative evidence can demonstrate the amount of clinician time saved through AI‑driven documentation and workflow automation?
Measuring time savings will provide tangible proof of AI’s value proposition for healthcare teams.
Speaker: Roy Jacobs

Disclaimer: This is not an official session record. DiploAI generates these resources from audiovisual recordings, and they are presented as-is, including potential errors. Due to logistical challenges, such as discrepancies in audio/video or transcripts, names may be misspelled. We strive for accuracy to the best of our ability.

Keynote-Julie Sweet

Session at a glanceSummary, keypoints, and speakers overview

Summary

The session opened with Speaker 1 introducing Julie Sweet, CEO of Accenture, and highlighting the firm’s position as a leading global AI and technology transformation company with a massive workforce ([1-6]).


Sweet thanked Indian leaders for convening the summit, noted Accenture’s 350,000-plus employees in India and its extensive AI workforce across multiple regions, and outlined three guiding perspectives: AI as a growth engine, the unprecedented agenda ahead, and the primacy of human leadership ([7-14][15-18]).


She recalled how Accenture used robotic process automation in 2013 to create thousands of jobs and, over the decade, grew from 275,000 staff and $29 billion revenue to over 750,000 staff and $70 billion, illustrating that embracing new technologies drives prosperity ([20-22]).


A recent C-suite survey across 20 countries showed that 78 % of companies are already using AI and 80 % view its greatest value as growth, reinforcing the business case for AI adoption ([24]).


Sweet argued that AI must make the “impossible possible,” citing how large language models will transform retail engagement and how AI could cut drug development timelines from nine years to much shorter periods, thereby creating new products and saving lives ([26-28][30-33]).


She emphasized that small and medium-sized enterprises (SMEs) generate about 50 % of global GDP and 70 % of employment in the Global South, and that ensuring their access to AI technology and talent will unlock substantial business opportunities ([36-38]).


To achieve this, she highlighted public-private partnerships such as funding internships for college students at SMEs, which improve hiring outcomes and provide cutting-edge talent to these firms ([43-46]).


Sweet warned that advanced AI’s power heightens the need for global collaboration, faster action, and stronger public-private cooperation to translate AI into productivity gains ([48-51]).


She called on companies to reinvent processes, invest in reshaping workforces, and create sustained entry-level jobs with AI-native skills, noting Accenture will hire more entry-level staff this year with redesigned training programs ([52-64]).


Governments must also adapt by embedding AI in education from primary school, fostering lifelong learning, and establishing common safety and industry standards-especially in sectors like pharma where divergent regulations hinder scaling breakthroughs ([65-70][71-75]).


Central to her message was the belief that “humans in the lead, not humans in the loop,” meaning leaders must decide how to use AI responsibly and collaboratively rather than relying on technology alone ([75-79]).


She concluded by invoking Accenture’s eight leadership essentials, stressing confidence, humility, and collective accountability as the foundation for a future where AI benefits all ([80-86]).


Speaker 1 closed by echoing Sweet’s tagline that AI should make the impossible possible, underscoring the summit’s overarching optimism ([88]).


Keypoints

Major discussion points


AI as a catalyst for growth and “making the impossible possible.”


Julie Sweet frames AI as the sole path to global prosperity, urging CEOs to showcase new products, services, and performance that were previously unattainable – from retail-focused LLM “malls” to dramatically faster drug development [15-18][26-33][34-35].


Ensuring inclusive access for small- and medium-sized enterprises (SMEs) and the Global South.


She stresses that 50 % of world GDP and 70 % of Global South employment come from SMEs, calling for public-private partnerships, internship programs, and talent pipelines to give these firms the AI tools and expertise they need [36-48].


A systemic reinvention of companies, governments, and individuals.


Companies must redesign processes, invest in AI-native entry-level roles, and adopt lifelong learning; governments need to embed AI in education and co-create standards that enable safe, cross-border scaling of high-impact sectors such as pharma [52-74].


Human leadership-“humans in the lead, not just in the loop.”


Sweet argues that technology is merely a tool; decisive, humble, and collaborative leaders must set the agenda, uphold safety standards, and drive responsible, widespread AI adoption [75-86].


Overall purpose / goal of the discussion


The session was intended to convey Accenture’s strategic vision for AI at scale: to harness AI as a growth engine, democratize its benefits across all enterprise sizes and regions, and mobilize coordinated action among businesses, governments, and individuals-anchored by strong, ethical leadership-to realize AI’s promise for “the benefit of all” [14-18].


Overall tone and its evolution


– The opening remarks are formal and appreciative, thanking Indian leaders and highlighting the summit’s significance [7-10].


– Sweet’s address then shifts to an optimistic, confident tone, emphasizing AI-driven prosperity and concrete industry examples [15-33].


– Mid-speech the tone becomes urgent and prescriptive, calling for immediate partnerships, workforce transformation, and global standards [48-51][52-74].


– The concluding segment adopts an inspirational and humble tone, stressing leadership virtues-excellence, confidence, humility-and a collective responsibility to shape a better future [75-86].


Overall, the conversation moves from gratitude to optimism, through urgency, and ends on a rallying, humble call to action.


Speakers

Speaker 1


– Role/Title: Event moderator / host (introduces the keynote speaker) [S1]


– Area of Expertise:


Julie Sweet


– Role/Title: Chair and CEO, Accenture [S5]


– Area of Expertise: AI and technology transformation, business strategy, digital innovation


Additional speakers:


Full session reportComprehensive analysis and detailed insights

The session opened with Speaker 1 formally introducing the next presenter, Julie Sweet, Chair and CEO of Accenture, and underscoring the firm’s stature as a leading global AI and technology-transformation organisation that deploys hundreds of thousands of professionals across every sector of the world economy [1-6].


Julie Sweet began by thanking Prime Minister Modi, Minister Vaishnav and the summit organisers, and she highlighted the breadth of the international audience as evidence of the need for broad partnerships to harness AI’s potential while managing its risks [7-10]. She then noted that Accenture employs more than 350,000 “reinventors” in India and that the company maintains one of the world’s largest AI workforces, tightly linked to AI hubs in the United States, Europe, the Middle East and Japan [11-13].


She framed the remainder of her address around three guiding perspectives: (i) AI as an engine for growth; (ii) an unprecedented agenda that requires reinvention of work, collaboration and learning; and (iii) humans in the lead, not merely in the loop [14-18].


To illustrate the power of technology-driven reinvention, Sweet recalled the 2013 Oxford study that warned 47 % of U.S. jobs could be automated and the subsequent hype around robotic process automation (RPA) [19]. She explained how Accenture and the broader IT-services industry embraced RPA, digital tools and classical AI, creating thousands of new jobs and enabling clients to invest in further innovation [20]. The lesson she drew was that organisations that adopt new technologies and channel them into growth and productivity prosper [21-22]; she argued that advanced AI should follow the same trajectory [23-24].


She then argued that AI’s greatest value lies in making the impossible possible [26-28]. In retail, large-language models will become “the new mall”, offering a wholly novel way to engage customers that did not exist in 2022 [30-32]. In pharmaceuticals, AI can compress the average nine-year drug-development cycle to a fraction of that time, accelerating life-saving treatments and boosting sales [33-34]. These early examples merely hint at AI’s capacity to create new drugs, materials and products across industries [35].


A central pillar of her vision is inclusive access for small- and medium-sized enterprises (SMEs). Sweet highlighted that SMEs generate roughly 50 % of global GDP and account for 70 % of employment in the Global South, according to the speaker [36-38]. She warned that merely creating business opportunities for SMEs will be insufficient without coordinated public-private partnerships [41-42].


To illustrate such partnerships, she described Accenture’s collaboration with the U.S. college system, which funds internships for students at SMEs [43-46]. This “win-win” model improves graduates’ employment prospects while delivering cutting-edge AI talent to smaller firms, thereby reinforcing the need to keep SMEs at the centre of AI deployment [47].


Sweet stressed that the unprecedented power of today’s AI heightens the urgency for global collaboration, faster action and stronger public-private cooperation to translate AI into productivity gains [48-51]. Companies, she argued, must be willing to reinvent how they operate, their processes, and how they have been doing work for decades, and they must invest in reshaping their workforces [52-57]. She announced a concrete hiring commitment: “We will hire into more entry-level jobs this year than last year” [34]. Moreover, “the skills we require and the way we’re onboarding those individuals is fundamentally different” [35-36]. Entry-level roles are the pipeline for future leaders, but AI is reshaping what those roles look like, requiring intentional redesign of job descriptions and training programmes [58-64].


Governments must also reinvent their roles. They need to partner with the private sector, become the “best credential for why AI matters”, and embed AI learning from primary school onward, fostering lifelong learning because “formal education is no longer the destination” [65-70]. This aligns with broader calls for systemic educational reform to support AI-native talent [S12][S14].


A further imperative is the creation of harmonised global standards that cover safety and sector-specific impacts, especially in high-stakes fields such as pharma. Without aligned regulations, breakthroughs in drug discovery cannot be scaled, leaving the most vulnerable populations behind [71-75][S36].


Underlying all these recommendations is the philosophy that technology, no matter how powerful, is only a tool; it is leaders who decide how to use those tools [75-79][S35]. Sweet concluded by invoking Accenture’s eight leadership essentials, urging leaders to act with excellence, confidence and humility, to hold themselves accountable for delivering on the promise of AI, and to recognise that the challenge cannot be met alone [80-86].


Speaker 1 closed the segment by thanking Julie Sweet and highlighting her tagline, “AI should make the impossible possible,” as a guiding theme for the summit [87-88].


Session transcriptComplete transcript of the session
Speaker 1

Thank you, Mr. Ankur Wara. Your perspectives on leveraging AI for social impact have undoubtedly added depth to the summit. And ladies and gentlemen, our next speaker is Ms. Julie Sweet, Chair and CEO, Accenture. Ms. Julie Sweet has repositioned Accenture as one of the world’s largest AI and technology transformation companies, deploying hundreds of thousands of professionals across every sector of the global economy. Her perspective on what AI adoption actually looks like at scale beyond the hype is grounded in hard operational reality. So please welcome the CEO of Accenture, Ms. Julie Sweet.

Julie Sweet

Thank you, Prime Minister Modi, Minister Vaishnav, and your outstanding teams for convening us for this critical summit around AI. The breadth of distinguished guests from around the world is a recognition of the importance of broad global partnerships to capture the incredible potential of AI and address the risks. It is also a recognition of the importance of India in our AI -enabled future. At Accenture, we’re incredibly proud to have over 350 ,000 and growing reinventors here in India. We also have one of the largest AI workforces in the world, tightly integrated with our growing AI hubs in the US, Europe, the Middle East, and Japan. And I want to take this time to thank all of our people in India for your incredible commitment to value.

And to our clients. Today, I want to leave you with three perspectives that we believe will help us ensure that AI’s immense potential. is captured for the benefit of all. First, using AI as an engine for growth is the only path for global prosperity for all. Second, the agenda ahead of us is unprecedented. Companies, countries, and individuals must reinvent how they work, how they work together, and how they learn. And finally, it is humans in the lead, not humans in the loop, that will determine our future. As we turn to the imperative for growth, I want to take you back for a moment to 2013. Oxford University had just published a widely read study that said based on technology progress at that time, 47 % of U .S.

jobs would be automatable. dire headlines and predictions soon followed one of those technologies was robotic process automation or rpa and there were predictions that it services would be badly damaged because it would automate so many jobs and in fact we used rpa to automate thousands of jobs and we also as an industry embraced the new technologies of digital and classical ai and we created many many more jobs we helped our clients adopt rpa and those who did created investment capacity to invest in new technologies and to grow and in fact the it services industry has thrived over the last decade including many of india’s most successful companies that you’ve heard from today At Accenture alone in 2013, we were roughly 275 ,000 people and $29 billion in revenue, and today we’re over 750 ,000 and growing and $70 billion in revenue.

What the last decade has taught us is a critical lesson. When companies and countries embrace new technologies and then use them to drive growth and productivity, they prosper. Advanced AI should be the same. In fact, in our latest quarterly survey of C -suites across 20 countries, they agree. 78 % of all companies are using C -suites. 80 % say AI’s greatest value is in growth. Now, as we think about what growth should look like, there’s two important considerations. First. AI should make the impossible possible. AI should make the impossible possible. If in a few years as a CEO, you cannot point to new products and services, new levels of performance that were not possible before, then you have not captured potential of AI.

Think about the consumer and retail industries. LLMs are about to become the new mall. This is an entirely new way to engage customers and to engage in commerce that did not exist in 2022. If you think about pharma, we see a path toward bringing drugs to market much faster than the average of nine years. Not possible before, which means that life -saving drugs will get to people faster and pharma will have accelerated sales. Growth. Growth. And we are just beginning to understand how AI will create new drugs, new materials, new products across industries. A second consideration around growth is that we must commit to providing access to the technology and the talent for small and medium -sized enterprises.

If we are to use AI as an engine for growth, we need to make sure that the engine for growth, these types, these size enterprises, have access. 50 % of the world’s GDP are small and medium -sized enterprises. And in the global south, it’s 70 % of employment. To do so, there will be lots of business opportunities. So many industries will serve small and medium -sized enterprises. But that will not be enough. Private and public partnerships will be critical to making sure there’s access. For example, we’re working with the U .S. college system where we’re funding internships of college students at small and medium -sized enterprises. It’s a win -win. Statistically, if you have an internship, you have a better chance of getting a job.

And it’s providing these enterprises access to some of the cutting -edge talent. And so we must make sure that we’re continuing to focus on the small and medium -sized enterprises. Now, I know, and we all know, that advanced AI is much more powerful than the technology advancements of the last decade. And, of course, that means that the impact is more profound. But that doesn’t change the critical lesson that AI must be used for growth and productivity. What it does change are the sets of actions, the time frame, the need for global collaboration, the need for more public and private partnerships, and the urgency of what we must do in order for AI to drive growth.

So companies, companies must be willing to reinvent how they operate, their processes, how they’ve been doing work for the last decades. Underneath the headlines of a failure of AI is mostly a failure to reinvent. Companies have to invest to reshape their workforces. And companies must commit to creating. Creating sustained entry -level jobs. Now, entry -level jobs makes economic sense. They’re the only way to create future leaders. And they bring needed, truly AI -native talent to each of our organizations. But AI fundamentally is changing what an entry -level job looks like. And so a commitment means we have to be intentional about changing the roles, investing in training, which is exactly what Accenture is doing. We will hire into more entry -level jobs this year than last year.

But the skills we require and the way we’re onboarding those individuals is fundamentally different. Now, countries must also reinvent. They must reinvent their role and how they work with the private sector. They have to themselves as governments become the best credential for why AI matters. They must work with the private sector to help create lifelong learning because education is no longer a destination. We have to have lifelong learning. India is doing a great job of embedding AI into the educational system, starting in primary school, and governments across the world will need to do so. At the same time, as countries are thinking differently, individuals have to think differently and recognize that formal education is no longer the destination.

But perhaps the biggest fundamental change that must be made is that companies and countries need to pound the table for global standards. These standards should apply to safety, but also to the industries where AI can make the greatest impact. For example, in pharma, if one country is allowing pharma companies to use the latest technologies to discover drugs, they should be able to make the greatest impact. If one country is allowing pharma companies to use drugs and then test drugs, but other countries don’t follow suit, it means that you won’t be able to scale, you won’t be able to bring it. And we know that most often that impacts the most vulnerable. now we have a view of course that we have to reinvent but as we think about that reinvention or to our future is the fundamental belief that it is humans in the lead not humans in the loop that will shape that future we should not confuse how you deploy ai responsibly of course all of our compliance programs have humans they have technology that doesn’t change the critical lesson that we’ve learned over and over again technology no matter how powerful is only a tool it is simply a tool it is leaders who decide how to use those tools It is leaders who decide to commit to reinvent, who dedicate their time to making sure that people come along the journey.

And it is leaders who must choose to work together to ensure the safe, widespread adoption of AI. There are lots of headlines today that predict less. Less jobs, less opportunity, less human relevance. We are here because we see a future of more. At Accenture, we live by eight leadership essentials, the qualities we believe we need to run our company. And one of them is particularly important. We expect leaders to lead with excellence, confidence, and humility. As we look to our collective future, we should have the confidence to have the unwavering belief that together we can make a future that is better for all. We also must hold ourselves individually and collectively accountable for executing on that belief with a high bar of excellence because our people around the world are counting on that excellence.

And finally, we must all have the humility to know that we cannot do this alone. Thank you very much.

Speaker 1

Thank you so much, Ms. Julie Sweet. I think I can take a tagline out of her address, which says that AI should make the impossible possible.

Related ResourcesKnowledge base sources related to the discussion topics (24)
Factual NotesClaims verified against the Diplo knowledge base (2)
Confirmedhigh

“Accenture is a leading global AI and technology‑transformation organisation that deploys hundreds of thousands of professionals across every sector of the world economy.”

The knowledge base describes Accenture as “one of the world’s largest AI and technology transformation companies” and highlights its global reach, confirming its leading status and large workforce [S6].

Confirmedhigh

“Accenture employs more than 350,000 “reinventors” in India and maintains one of the world’s largest AI workforces.”

The source reports that Accenture has “over 350,000 employees in India and growing,” and notes that India’s talent pool is central to the company’s global AI strategy, supporting the claim about the size of its Indian AI workforce [S52].

External Sources (61)
S1
Keynote-Martin Schroeter — -Speaker 1: Role/Title: Not specified, Area of expertise: Not specified (appears to be an event moderator or host introd…
S2
Responsible AI for Children Safe Playful and Empowering Learning — -Speaker 1: Role/title not specified – appears to be a student or child participant in educational videos/demonstrations…
S3
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Vijay Shekar Sharma Paytm — -Speaker 1: Role/Title: Not mentioned, Area of expertise: Not mentioned (appears to be an event host or moderator introd…
S4
Keynote-Julie Sweet — -Moderator: Role/Title: Not specified, Area of expertise: Not specified Addressing the SME Challenge for Inclusive Grow…
S5
Industries in the Intelligent Age / DAVOS 2025 — – Julie Sweet – CEO of Accenture 2. HR Transformation: Julie Sweet argued that HR departments need to be reinvented for…
S6
Keynote-Julie Sweet — Overall Tone:The tone is consistently optimistic and forward-looking throughout, with Sweet maintaining confidence in AI…
S7
AI Governance Dialogue: Steering the future of AI — No meaningful disagreement analysis possible – the transcript consists of a keynote address by Doreen Bogdan Martin with…
S8
Using AI to tackle our planet’s most urgent problems — Vogels closed with a powerful statement: “Remember, maps aren’t just tools for navigation. They’re tools for justice, he…
S9
How to make AI governance fit for purpose? — Chuen Hong Lew: I try to sleep well at night. I think that’s very important. So, if all of you are not getting your seve…
S10
Technology in a Turbulent World — Humility and open conversation are emphasised by Julie Sweet as essential qualities for successful leadership. Accenture…
S11
How AI Drives Innovation and Economic Growth — Kremer argues that while there are forces that may widen gaps, AI has significant potential to narrow development dispar…
S12
AI/Gen AI for the Global Goals — Priscilla Boa-Gue argues for the creation of supportive policy environments to foster AI startups. This includes develop…
S13
Is AI a catalyst for development? — The Economist argues that AI has the potential to revolutionise developing countries by transforming their economies and…
S14
How AI Drives Innovation and Economic Growth — “So, you know, for all countries, but especially for emerging markets and developing economies, AI can be a game changer…
S15
Addressing the gender divide in the e-commerce marketplace – a policy playbook for the global South (IT for Change) — Another key point raised is the need to make developmental and gender topics transversal to current trade agreements. It…
S16
Unlocking potential: Addressing inclusivity barriers in e-commerce trade to deliver sustainable impact in communities everywhere (United Kingdom) — Cross-border trade offers numerous benefits to businesses, including increased sustainability, resilience, employment op…
S17
Creating Eco-friendly Policy System for Emerging Technology — Governments, businesses, foundations, universities, and others should design interventions geared at achieving a sustain…
S18
The fading of human agency in automated systems — Crucially, a human presence does not guarantee agency if the system is designed around compliance rather than contestati…
S19
Scaling AI Beyond Pilots: A World Economic Forum Panel Discussion — It is human in the lead, not human in the loop.
S20
Open Mic & Closing Ceremony — The overall tone was formal yet appreciative. There was a sense of accomplishment and gratitude expressed throughout, wi…
S21
Leaders’ Plenary | Global Vision for AI Impact and Governance Morning Session Part 2 — The tone was consistently collaborative, optimistic, and forward-looking throughout the session. Delegates maintained a …
S22
Powering AI Global Leaders Session AI Impact Summit India — “Thanks.”[1]. “Thank you.”[2]. “Thank you so much.”[5]. “Thank you for your partnership.”[7]. Expressing thanks and app…
S23
Summit Opening Session — The tone throughout is consistently formal, diplomatic, and collaborative. Speakers maintain an optimistic and forward-l…
S24
WS #288 An AI Policy Research Roadmap for Evidence-Based AI Policy — Anne Flanagan: Hello, apologies that I’m not there in person today. I’m in transit at the moment, hence my picture on yo…
S25
Laying the foundations for AI governance — Artemis Seaford: So the greatest obstacle, in my opinion, to translating AI governance principles into practice may actu…
S26
What policy levers can bridge the AI divide? — Hubert Vargas Picado: and we go to His Excellency, your title has innovation, can you tell us more about what are the be…
S27
Is AI a catalyst for development? — The Economist argues that AI has the potential to revolutionise developing countries by transforming their economies and…
S28
Keynote-Julie Sweet — “First, using AI as an engine for growth is the only path for global prosperity for all.”[1]”When companies and countrie…
S29
How AI Drives Innovation and Economic Growth — “So, you know, for all countries, but especially for emerging markets and developing economies, AI can be a game changer…
S30
Keynote-Julie Sweet — Sweet illustrates this transformative potential through specific industry examples. In consumer and retail sectors, she …
S31
How AI Drives Innovation and Economic Growth — So thank you very much, Jeanette. It’s a great pleasure to be here speaking to all of you this afternoon. Over the past …
S32
Unlocking potential: Addressing inclusivity barriers in e-commerce trade to deliver sustainable impact in communities everywhere (United Kingdom) — Access to the Global South for ongoing trade with the Global North depends on supporting MSMEs and the informal sector i…
S33
Addressing the gender divide in the e-commerce marketplace – a policy playbook for the global South (IT for Change) — Another key point raised is the need to make developmental and gender topics transversal to current trade agreements. It…
S34
[Tentative Translation] — In addition, (4) the government will create a new industrial base in which diverse entities, such as corporations, unive…
S35
The fading of human agency in automated systems — Crucially, a human presence does not guarantee agency if the system is designed around compliance rather than contestati…
S36
Scaling AI Beyond Pilots: A World Economic Forum Panel Discussion — It is human in the lead, not human in the loop.
S37
Open Mic & Closing Ceremony — The overall tone was formal yet appreciative. There was a sense of accomplishment and gratitude expressed throughout, wi…
S38
High-Level Track Facilitators Summary and Certificates — ## Opening Remarks and Summit Achievements
S39
Leaders’ Plenary | Global Vision for AI Impact and Governance Morning Session Part 2 — The tone was consistently collaborative, optimistic, and forward-looking throughout the session. Delegates maintained a …
S40
Opening of the session — The tone began very positively and constructively, with the Chair commending delegations for focused, specific intervent…
S41
Any other business /Adoption of the report/ Closure of the session — Finally, Colombia shows gratitude towards the chairs and the diplomatic teams. This appreciation transcends mere formali…
S42
How AI Drives Innovation and Economic Growth — The tone was notably optimistic yet pragmatic, described as representing “hope” rather than the “fear” that characterize…
S43
Ad Hoc Consultation: Tuesday 6th February, Afternoon session — The tone of urging could change the nature of the actions expected under the paragraph.
S44
New Technologies and the Impact on Human Rights — The discussion maintained a collaborative and constructive tone throughout, despite addressing complex and sometimes con…
S45
Presentation of outcomes to the plenary — In summary, the speaker emphasised that the discussions converged on the necessity for a strong, swift, collaborative ef…
S46
(Interactive Dialogue 1) Summit of the Future – General Assembly, 79th session — The overall tone was one of urgency and calls for action, with many speakers emphasizing the need for immediate reforms …
S47
(Interactive Dialogue 4) Summit of the Future – General Assembly, 79th session — The overall tone was one of urgency and determination. Many speakers emphasized that “the future starts now” and stresse…
S48
Closing Ceremony — The discussion maintains a consistently positive and collaborative tone throughout, characterized by gratitude, celebrat…
S49
Business Engagement Session: Sustainable Leadership in the Digital Age – Shaping the Future of Business — The discussion maintained a consistently collaborative and optimistic tone throughout. It began with academic framing bu…
S50
Leaders TalkX: Future-ready: enhancing skills for a digital tomorrow — The discussion maintained a consistently positive, collaborative, and inspiring tone throughout. Panelists were enthusia…
S51
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Jeetu Patel President and Chief Product Officer Cisco Inc — -His Honorable Prime Minister, Mr. Narendra Modi: Prime Minister of India (mentioned but did not speak in this transcrip…
S52
Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session — Julie Sweet from Accenture highlighted another crucial advantage: India’s human capital. With over 350,000 employees in …
S53
Open Forum #58 Collaborating for Trustworthy AI an Oecd Toolkit and Spotlight on AI in Government — Jibu Elias: Thank you very much, Yoichi-san. It’s an honor to be here to share my experience building a responsible AI e…
S54
We are the AI Generation — In her conclusion, Martin articulated that the fundamental question should not be “who can build the most powerful model…
S55
The Global Economic Outlook — Georgieva emphasizes the importance of making artificial intelligence accessible to all, not just a privileged few. She …
S56
Opening keynote — Bogdan-Martin conveyed optimism about the emerging AI regulatory landscape, viewing it as an opportunity to ensure that …
S57
Generative AI: Steam Engine of the Fourth Industrial Revolution? — Shared the curriculum created with Oxford university with U.S counterpart Sweet warns against complacency as every indu…
S58
World in Numbers: Jobs and Tasks / DAVOS 2025 — Andrew Ng: for humans to still do. Yeah, I’m glad you asked that. So it turns out that AI-assisted coding is incredi…
S59
REGULATING THE DIGITAL ECONOMY: DILEMMAS, TRADE OFFS AND POTENTIAL OPTIONS — This is not an easy task, especially for many developing countries, where STEM graduates are difficult to fi…
S60
Rights and Permissions — FIGURE 1.1 Estimates of the percentage of jobs at risk from automation ## vary widely Job loss predictions do not accur…
S61
Media Hub — Minister Bah mentioned ongoing conversations with the Minister of Education about adapting educational frameworks to int…
Speakers Analysis
Detailed breakdown of each speaker’s arguments and positions
S
Speaker 1
1 argument137 words per minute122 words53 seconds
Argument 1
AI should make the impossible possible.
EXPLANATION
Speaker 1 highlighted the slogan that AI must enable outcomes that were previously unattainable, emphasizing its transformative potential.
EVIDENCE
Speaker 1 repeated the phrase “AI should make the impossible possible” as a memorable tagline summarising Julie Sweet’s address [88].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
The keynote transcript notes Sweet repeatedly used the phrase “AI should make the impossible possible” to illustrate breakthrough potential [S6].
MAJOR DISCUSSION POINT
AI as an engine for growth and productivity
AGREED WITH
Julie Sweet
J
Julie Sweet
11 arguments122 words per minute1564 words765 seconds
Argument 1
AI must be used to drive growth and productivity; embracing new technologies leads to prosperity – Julie Sweet
EXPLANATION
Julie Sweet argued that using AI as an engine for growth is essential for global prosperity, and that companies and countries that adopt new technologies see increased productivity and economic success.
EVIDENCE
She stated that “using AI as an engine for growth is the only path for global prosperity for all” and that “when companies and countries embrace new technologies and then use them to drive growth and productivity, they prosper” [15][22].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Sweet emphasizes AI as a growth engine and links prosperity to adopting new technologies in the keynote [S6].
MAJOR DISCUSSION POINT
AI as an engine for growth and productivity
Argument 2
AI should make the impossible possible, delivering new products, services, and performance levels – Julie Sweet
EXPLANATION
She emphasized that AI should enable breakthroughs that were previously impossible, such as novel consumer experiences and faster drug development, thereby creating new value for businesses.
EVIDENCE
Julie Sweet repeated the mantra “AI should make the impossible possible” and explained that CEOs must point to new products, services, or performance that were not feasible before [26-28]; she illustrated this with examples of LLMs creating a “new mall” for retail and AI accelerating drug-to-market timelines in pharma [30-34].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
She expands the mantra with examples such as new retail experiences and accelerated drug development, illustrating breakthroughs previously unattainable [S6].
MAJOR DISCUSSION POINT
AI as an engine for growth and productivity
Argument 3
78 % of C‑suite respondents say AI’s greatest value is in growth – Julie Sweet
EXPLANATION
Julie Sweet cited a recent survey showing that a large majority of senior executives view AI primarily as a driver of growth.
EVIDENCE
She referenced Accenture’s latest quarterly survey of C-suite leaders across 20 countries, noting that 78 % of companies are using C-suite input and 80 % say AI’s greatest value is in growth [24].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
The Accenture quarterly survey cited in the speech shows 78 % of executives view AI’s greatest value as growth [S6].
MAJOR DISCUSSION POINT
AI as an engine for growth and productivity
Argument 4
Provide AI technology and talent access to SMEs, which account for 50 % of global GDP and 70 % of employment in the Global South – Julie Sweet
EXPLANATION
She argued that to harness AI’s growth potential, small and medium‑sized enterprises must be given both the tools and skilled workforce, because they represent a substantial share of the world economy and employment.
EVIDENCE
Julie Sweet highlighted that “50 % of the world’s GDP are small and medium-sized enterprises” and “70 % of employment in the global south” comes from SMEs, underscoring the scale of the opportunity [36-38].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
She highlights SMEs represent 50 % of global GDP and 70 % of Global South employment, underscoring the need for AI access [S4][S6].
MAJOR DISCUSSION POINT
Inclusion of small and medium‑sized enterprises (SMEs) and talent development
Argument 5
Public‑private partnerships such as U.S. college internships for SMEs create win‑win talent pipelines – Julie Sweet
EXPLANATION
She described a collaborative model where Accenture funds internships for college students placed in SMEs, benefiting both the enterprises (by gaining cutting‑edge talent) and the students (by improving job prospects).
EVIDENCE
She gave the example of working with the U.S. college system to fund internships at SMEs, noting that internships increase employment chances and provide enterprises with access to advanced talent, describing it as a “win-win” [42-47].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Sweet describes U.S. college internship programs funded for SMEs as win-win talent pipelines that benefit both students and enterprises [S6].
MAJOR DISCUSSION POINT
Inclusion of small and medium‑sized enterprises (SMEs) and talent development
Argument 6
Commit to creating sustained entry‑level, AI‑native jobs and invest in training for them – Julie Sweet
EXPLANATION
Julie Sweet called for companies to deliberately create and nurture entry‑level positions that are designed for AI‑native talent, emphasizing training and a re‑imagined onboarding process.
EVIDENCE
She said companies must “commit to creating sustained entry-level jobs” and invest in training, noting Accenture will hire more entry-level staff than the previous year and that the required skills and onboarding are fundamentally different [55-63].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
She calls for companies to create sustained entry-level AI-native positions and invest in training and onboarding for new talent [S6].
MAJOR DISCUSSION POINT
Inclusion of small and medium‑sized enterprises (SMEs) and talent development
Argument 7
Humans must be “in the lead,” not merely “in the loop,” as leaders decide how tools are used – Julie Sweet
EXPLANATION
She stressed that ultimate responsibility for AI lies with human leaders who set direction and make decisions, rather than merely supervising automated systems.
EVIDENCE
In a lengthy passage she contrasted “humans in the lead” with “humans in the loop,” asserting that leaders decide how tools are used and must work together to ensure safe AI adoption [75-76].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
She contrasts “humans in the lead” versus “humans in the loop,” stressing that ultimate responsibility rests with human leaders [S6].
MAJOR DISCUSSION POINT
Human leadership and responsible AI governance
Argument 8
Develop global standards for AI safety and industry‑specific impact (e.g., pharma) to ensure equitable scaling – Julie Sweet
EXPLANATION
She advocated for internationally harmonised standards covering safety and sector‑specific applications, so that breakthroughs in one country can be scaled globally and benefit vulnerable populations.
EVIDENCE
She called for standards that apply to safety and to high-impact industries, giving pharma as an example where inconsistent national rules would hinder scaling and affect the most vulnerable [72-75].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
She advocates for internationally harmonised AI safety standards, especially for high-impact sectors like pharmaceuticals, to enable equitable scaling [S6].
MAJOR DISCUSSION POINT
Human leadership and responsible AI governance
Argument 9
Leaders should embody excellence, confidence, and humility to guide AI adoption responsibly – Julie Sweet
EXPLANATION
She outlined Accenture’s leadership essentials, emphasizing that leaders need to act with excellence, confidence, and humility to collectively achieve a better AI‑enabled future.
EVIDENCE
Julie Sweet referenced Accenture’s eight leadership essentials, highlighting the need for excellence, confidence, and humility, and called for accountability and collaborative humility in AI deployment [81-86].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Accenture’s leadership essentials of excellence, confidence, and humility are outlined in the keynote and related commentary [S4][S10].
MAJOR DISCUSSION POINT
Human leadership and responsible AI governance
Argument 10
Countries need to reinvent their roles, work with the private sector, and embed AI into education with lifelong learning – Julie Sweet
EXPLANATION
She argued that governments must reshape how they interact with industry, champion lifelong learning, and integrate AI education from primary school onward to keep pace with rapid technological change.
EVIDENCE
She stated that countries must reinvent their role, collaborate with the private sector, and create lifelong learning pathways, noting India’s efforts to embed AI from primary school as a model [64-70].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
She urges governments to reshape private-sector collaboration, become AI exemplars, and embed AI education from primary school onward, citing India as a model [S4][S6].
MAJOR DISCUSSION POINT
Global collaboration and standards for AI adoption
Argument 11
Harmonized global standards enable cross‑border scaling of AI innovations, especially in high‑impact sectors like pharmaceuticals – Julie Sweet
EXPLANATION
She emphasized that coordinated international standards are essential for scaling AI‑driven breakthroughs, such as new drug discovery, across borders without regulatory fragmentation.
EVIDENCE
She explained that global standards should cover safety and industry impact, using pharma as an example where divergent national policies would block scaling and harm vulnerable populations [72-75].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
She stresses coordinated global standards are needed for scaling AI breakthroughs such as drug discovery across borders [S6].
MAJOR DISCUSSION POINT
Global collaboration and standards for AI adoption
Agreements
Agreement Points
AI should make the impossible possible
Speakers: Speaker 1, Julie Sweet
AI should make the impossible possible. AI should make the impossible possible, delivering new products, services, and performance levels — Julie Sweet
Both Speaker 1 and Julie Sweet emphasize that AI must enable outcomes that were previously unattainable, positioning it as a transformative engine for growth [26-28][88].
Similar Viewpoints
Julie Sweet consistently argues that AI should be harnessed as a growth engine, made accessible to SMEs, supported by public‑private partnerships, accompanied by workforce development, human leadership, and global standards to ensure responsible and inclusive adoption [15][22][26-28][36-38][42-47][55-63][71-76][81-86].
Speakers: Julie Sweet
AI must be used to drive growth and productivity; embracing new technologies leads to prosperity — Julie Sweet AI should make the impossible possible, delivering new products, services, and performance levels — Julie Sweet Provide AI technology and talent access to SMEs, which account for 50 % of global GDP and 70 % of employment in the Global South — Julie Sweet Public‑private partnerships such as U.S. college internships for SMEs create win‑win talent pipelines — Julie Sweet Commit to creating sustained entry‑level, AI‑native jobs and invest in training — Julie Sweet Humans must be “in the lead,” not merely “in the loop” — Julie Sweet Develop global standards for AI safety and industry‑specific impact (e.g., pharma) — Julie Sweet Leaders should embody excellence, confidence, and humility to guide AI adoption responsibly — Julie Sweet Countries need to reinvent their roles, work with the private sector, and embed AI into education with lifelong learning — Julie Sweet
Unexpected Consensus
Overall Assessment

The primary point of consensus is the shared emphasis on AI making the impossible possible, reflecting a common belief in AI’s transformative potential. Beyond this, the discussion is dominated by Julie Sweet’s extensive agenda, with limited direct overlap from Speaker 1.

Limited but clear consensus on the core slogan; broader agreement on detailed policy measures is absent, suggesting that while the vision of AI as a breakthrough tool is shared, concrete strategies remain speaker‑specific.

Differences
Different Viewpoints
Unexpected Differences
Overall Assessment

The transcript shows strong alignment rather than conflict. Julie Sweet’s detailed arguments about AI as a growth engine, SME inclusion, talent development, and human leadership are echoed by Speaker 1’s brief endorsement of the “AI should make the impossible possible” tagline. No speaker presents a contrasting viewpoint, resulting in an overall atmosphere of consensus.

Minimal disagreement; the lack of opposing positions suggests smooth collaboration and shared objectives, which bodes well for coordinated policy and industry action on AI.

Partial Agreements
Both speakers highlight the same core slogan that AI must enable outcomes that were previously unattainable, signalling a shared vision of AI as a breakthrough catalyst [26-28][88].
Speakers: Speaker 1, Julie Sweet
AI should make the impossible possible, delivering new products, services, and performance levels — Julie Sweet AI should make the impossible possible — Speaker 1
Takeaways
Key takeaways
AI must be used as an engine for growth and productivity; embracing new technologies leads to prosperity. AI should make the impossible possible by enabling new products, services, and performance levels that were previously unattainable. A large majority of C‑suite executives (78 %) view AI’s greatest value as driving growth. Inclusion of SMEs is critical because they represent 50 % of global GDP and 70 % of employment in the Global South; they need access to AI technology and talent. Public‑private partnerships (e.g., U.S. college internships for SMEs) are a viable way to build talent pipelines and provide AI access. Creating sustained entry‑level, AI‑native jobs and investing in training is essential for future leadership and economic health. Human leadership must be “in the lead,” not merely “in the loop”; leaders decide how AI tools are applied. Global, industry‑specific AI standards (including safety) are needed to enable cross‑border scaling, especially in sectors like pharmaceuticals. Countries must reinvent their role, collaborate with the private sector, and embed AI into education through lifelong learning.
Resolutions and action items
Accenture will increase hiring of entry‑level AI‑native positions this year compared with the prior year. Accenture will expand internship programs linking U.S. colleges with small and medium‑sized enterprises to create talent pipelines. Accenture commits to work with governments and industry partners to develop public‑private frameworks that provide AI technology and talent access to SMEs. Accenture will advocate for and participate in the creation of harmonized global AI safety and industry‑specific standards.
Unresolved issues
How to design and implement globally harmonized AI standards that are accepted across different regulatory regimes. Specific mechanisms for scaling AI‑driven pharmaceutical innovations internationally while ensuring safety and equitable access. Concrete policies and funding models to support widespread AI adoption by SMEs, especially in the Global South. Detailed strategies for embedding lifelong learning and AI curricula into national education systems beyond pilot initiatives. Addressing broader societal concerns about potential job displacement and ensuring inclusive economic benefits.
Suggested compromises
Adopt public‑private partnership models as a middle ground to provide AI resources to SMEs while sharing risk and investment. Balance stringent safety standards with the need for rapid innovation by creating tiered or sector‑specific regulatory frameworks. Position humans as leaders (in the lead) rather than merely overseers (in the loop) to reconcile responsible AI governance with technological advancement.
Thought Provoking Comments
AI should make the impossible possible.
Echoes the keynote’s central mantra, signaling that the audience internalized the core message and will likely frame future discussions around this principle.
Acts as a brief turning point that transitions the session from the keynote to the next agenda item, cementing the “impossible‑possible” narrative as the summit’s guiding theme.
Speaker: Speaker 1 (after Julie Sweet’s speech)
When companies and countries embrace new technologies and then use them to drive growth and productivity, they prosper. Advanced AI should be the same.
She draws a direct historical parallel between past tech waves (RPA, digital) and today’s AI, grounding the hype in concrete evidence that adoption leads to prosperity.
Creates a turning point from describing AI hype to presenting a proven growth narrative, encouraging the audience to view AI adoption as an economic imperative rather than a speculative risk.
Speaker: Julie Sweet
We must commit to providing access to the technology and the talent for small and medium‑sized enterprises… 50 % of the world’s GDP and 70 % of employment in the Global South come from SMEs.
Highlights inclusion and equity, shifting the focus from large corporations to the broader economic base that sustains most of the world’s workforce.
Introduces a new topic—SME empowerment—and calls for public‑private partnerships, prompting listeners to consider policy and investment strategies beyond corporate roll‑outs.
Speaker: Julie Sweet
Humans in the lead, not humans in the loop, will determine our future.
Moves the discourse from technical safety mechanisms to a leadership philosophy, emphasizing proactive human agency over passive oversight.
Marks a tonal shift toward responsibility and governance, influencing subsequent remarks about standards, lifelong learning, and the role of leaders in shaping AI outcomes.
Speaker: Julie Sweet
Companies must reinvent how they operate, reshape their workforces, and create sustained entry‑level, AI‑native jobs.
Challenges the conventional view that AI eliminates jobs, proposing instead a re‑design of entry‑level roles to cultivate future AI talent.
Deepens the analysis of workforce implications, leading the audience to contemplate training, hiring practices, and the long‑term talent pipeline rather than short‑term displacement.
Speaker: Julie Sweet
Countries need to embed AI into education from primary school and promote lifelong learning; education is no longer a destination.
Calls for systemic educational reform, expanding the conversation from corporate strategy to national policy and cultural change.
Broadens the scope of the discussion to include government action, reinforcing the earlier point about public‑private partnerships and setting up a narrative for global standards.
Speaker: Julie Sweet
Global standards must be harmonized for safety and for sectors where AI can make the greatest impact, such as pharma drug discovery; otherwise vulnerable populations suffer.
Links regulatory alignment directly to real‑world health outcomes, illustrating the tangible stakes of fragmented AI policies.
Creates a concrete call‑to‑action for international cooperation, steering the dialogue toward cross‑border governance and highlighting the urgency of coordinated standards.
Speaker: Julie Sweet
We must lead with excellence, confidence, and humility, recognizing we cannot do this alone.
Summarizes the leadership ethos required for responsible AI deployment, tying together earlier themes of collaboration, responsibility, and ambition.
Provides a concluding rallying point that reinforces the earlier insights and leaves the audience with a clear behavioral prescription, setting the tone for any subsequent Q&A or follow‑up actions.
Speaker: Julie Sweet
Overall Assessment

Julie Sweet’s address introduced a series of interlocking ideas—historical precedent for growth, inclusive access for SMEs, a leadership‑first philosophy, workforce reinvention, education reform, and the need for global standards—that collectively shifted the conversation from abstract AI hype to concrete, actionable strategies. Each pivotal comment acted as a catalyst, opening new sub‑topics, reframing existing concerns, and deepening the analysis of AI’s societal impact. The brief echo by Speaker 1 reinforced the keynote’s central thesis, ensuring that the summit’s subsequent dialogue would be anchored around the ambition of making the impossible possible.

Follow-up Questions
How can small and medium‑sized enterprises (SMEs) be provided with equitable access to AI technologies and talent?
SMEs represent 50 % of global GDP and 70 % of employment in the Global South; ensuring they can leverage AI is essential for inclusive growth.
Speaker: Julie Sweet
What public‑private partnership models are most effective for scaling AI adoption among SMEs?
Julie highlighted the need for collaborations, such as internships linking U.S. colleges with SMEs, to bridge talent gaps and foster AI uptake.
Speaker: Julie Sweet
How can global AI standards—covering safety and industry‑specific applications like pharma—be developed and harmonized across jurisdictions?
She warned that divergent national regulations could hinder scaling of AI‑driven innovations, especially in critical sectors affecting vulnerable populations.
Speaker: Julie Sweet
What strategies should governments and the private sector employ to embed lifelong learning and continuous upskilling into the workforce?
She emphasized that formal education is no longer a destination, making ongoing learning vital for maintaining AI‑native talent.
Speaker: Julie Sweet
In what ways should entry‑level job roles be redefined and training programs redesigned to cultivate AI‑native talent?
Entry‑level positions are crucial for developing future leaders, but AI changes the skill set required; intentional redesign is needed.
Speaker: Julie Sweet
What best practices can companies adopt to reinvent their processes and invest in AI‑driven growth?
She noted that many AI failures stem from a lack of reinvention; identifying effective transformation approaches is critical.
Speaker: Julie Sweet
How can AI accelerate drug discovery timelines, and what regulatory frameworks are needed to support faster approvals while ensuring safety?
She cited pharma as a sector where AI could reduce the average nine‑year development cycle, raising questions about appropriate oversight.
Speaker: Julie Sweet
What does the concept of LLMs becoming the ‘new mall’ imply for consumer retail business models, and how should companies prepare?
She introduced a novel commerce paradigm that warrants exploration of emerging retail strategies and customer engagement methods.
Speaker: Julie Sweet
What role should governments play in integrating AI education into primary schools to ensure early exposure and equity?
She praised India’s efforts and suggested that worldwide adoption could build a foundation for future AI literacy.
Speaker: Julie Sweet
How can the impact of AI on global prosperity be measured to ensure that growth is inclusive and benefits all stakeholders?
She linked AI to global prosperity but highlighted the need for metrics and accountability to track equitable outcomes.
Speaker: Julie Sweet

Disclaimer: This is not an official session record. DiploAI generates these resources from audiovisual recordings, and they are presented as-is, including potential errors. Due to logistical challenges, such as discrepancies in audio/video or transcripts, names may be misspelled. We strive for accuracy to the best of our ability.

Keynote-Demis Hassabis

Session at a glanceSummary, keypoints, and speakers overview

Summary

At the India AI summit, DeepMind co-founder and CEO Demis Hassabis addressed leaders from industry, academia and government about the rapid evolution of artificial intelligence and its coming challenges [8-9]. He highlighted DeepMind’s breakthrough AlphaFold, which solved the 50-year-old protein-folding problem, as a concrete example of AI accelerating scientific discovery [20]. Hassabis noted that similar AI tools are now being applied to material science, fusion, physics and mathematics, suggesting that virtually every branch of science and medicine can be transformed [22-23].


Looking ahead, he warned that artificial general intelligence (AGI) may appear within five years, with foundational models improving dramatically week by week [24-26]. He stressed that this opportunity must be pursued with humility, acknowledging that we still lack full understanding of how the technology will develop or be deployed [27-28]. To manage the risks, Hassabis advocated a scientific-method approach to build guardrails, monitoring systems and clear purpose-aligned objectives [43]. He compared the forthcoming impact of AGI to the advent of fire or electricity, estimating it could be ten times the effect of the Industrial Revolution but compressed into a single decade [40-42].


During his recent visit to Bangalore, Hassabis was impressed by the enthusiasm of students and faculty at the Indian Institute of Science and by DeepMind’s research office focused on efficient models, continual learning and multilingual capabilities [30-33]. He announced a deep partnership with Mukesh Ambani’s Reliance group to deliver Gemini Foundation models across India, building on broader Google collaborations announced by Sundar Pichai [36-38]. Hassabis expressed confidence that India will become a global AI powerhouse, contributing both to economic growth and to scientific breakthroughs [34-35].


He argued that international summits like this are essential for bringing together technologists, scientists, policymakers, artists and philosophers to shape responsible AI governance [44-45]. By fostering such interdisciplinary dialogue, the community can create the safeguards and bold initiatives needed to turn AI’s disruptive potential into a “golden era” of discovery and improved human health [46]. Overall, Hassabis’s remarks framed the current AI moment as a historic crossroads where coordinated global effort can ensure that the transformative power of AGI benefits all of humanity [43].


Keypoints

AI as a catalyst for scientific discovery across all domains – DeepMind’s AlphaFold solved the 50-year protein-folding problem and the company is extending AI tools to material science, fusion, physics, mathematics and medicine, arguing that “almost every branch of science and medicine can be impacted by AI” and that the coming era of AGI will amplify productivity and research breakthroughs [20-23][24-26].


India’s emerging position as a global AI hub and strategic partner – Hassabis highlights his recent visit to Bangalore, the strong research presence of Google/DeepMind there, the enthusiasm of Indian academia, and the deep partnership with Reliance (via the Gemini foundation models) that aims to “bring intelligence to everyone in India” [30-38].


The unprecedented societal impact of AGI and the need for careful stewardship – He likens the arrival of AGI to “the advent of fire or electricity,” estimating an impact ten times that of the Industrial Revolution within a decade, and stresses the importance of scientific-method-based guardrails, monitoring, and responsible deployment [40-42][43-45].


Call for inclusive, international dialogue and interdisciplinary collaboration – Hassabis argues that the challenges of AGI cannot be left to technologists alone; governments, artists, social scientists, philosophers and others must join the conversation to shape a “golden era of scientific discovery” that benefits all humanity [44-46].


Overall purpose/goal of the discussion


The talk serves to celebrate the AI summit, showcase India’s growing AI ecosystem, illustrate the transformative promise of AGI, and rally a broad coalition of stakeholders-industry, academia, governments, and the wider society-to collaborate on responsible development and deployment of AI for global scientific and economic benefit.


Overall tone


The tone begins with enthusiastic admiration for AI’s achievements and India’s energy, shifts to awe-inspired optimism about the coming AGI era, adopts a sober, cautionary note regarding risks and the need for guardrails, and concludes with a collaborative, inclusive appeal for worldwide dialogue and partnership. The progression moves from celebratory to visionary, then to prudent, and finally to unifying.


Speakers

Speaker 1


– Role/Title: Event moderator/host (introducing speakers) [S1][S3]


– Areas of expertise:


Demis Hassabis


– Role/Title: Co-founder and CEO of Google DeepMind; Sir; Nobel laureate (as described) [S4]


– Areas of expertise: Artificial intelligence research, neuroscience, game design, chess, scientific discovery


Additional speakers:


Mukesh Dhirubhai Ambani


– Role/Title: Business leader, Chairman & Managing Director of Reliance Industries (inferred)


– Areas of expertise: Business, energy, telecommunications


Narendra Modi


– Role/Title: Prime Minister of India


– Areas of expertise: Politics, governance


Rishi Sunak


– Role/Title: Prime Minister of the United Kingdom


– Areas of expertise: Politics, governance


Sundar Pichai


– Role/Title: CEO of Alphabet Inc. and Google


– Areas of expertise: Technology, business leadership


Full session reportComprehensive analysis and detailed insights

Speaker 1 opened the India AI Summit by applauding Mukesh Dhirubhai Ambani for his confidence in India’s capabilities and describing him as a leader of the nation’s AI revolution, then introduced the co-founder and CEO of Google DeepMind, Sir Demis Hassabis[1-4].


Hassabis said thank you to the audience and congratulated Prime Minister Modi [5-6] and the Indian government for convening this pivotal AI summit. He noted that the inaugural meeting was convened by Prime Minister Sunak at Bletchley Park in the UK [7-9].


He said AI is a powerful catalyst for scientific discovery and one of the most important technologies ever invented. He cited DeepMind’s AlphaFold breakthrough, which solved the fifty-year-old protein-folding problem, as concrete evidence of AI’s transformative potential [20-21].


He explained that DeepMind is now extending AI tools to material science, fusion research, physics, mathematics, and medicine, arguing that “almost every branch of science and medicine can be impacted by AI” [22-23].


He recounted a recent visit to Bangalore where he toured DeepMind’s large research office focused on efficient models, continual learning and multilingual capabilities, and delivered a talk at the Indian Institute of Science, noting the enthusiasm of students and faculty [30-33]. He highlighted the deep partnership with Mukesh Ambani’s Reliance Geo to deploy the Gemini foundation models across the country [34-38].


He described the present as a “threshold moment” in which artificial general intelligence may appear within the next five years [24-26]. He added that foundational (or general-purpose) models are becoming increasingly capable almost week by week, offering unprecedented opportunities for economic productivity and scientific advancement [24-26].


He compared the upcoming impact of AGI to the advent of fire or electricity and estimated that it could be roughly ten-fold the impact of the Industrial Revolution, unfolding at ten times the speed-i.e., within a decade rather than a century [40-42].


He advocated a scientific-method approach to harness this change responsibly: building robust guardrails, continuous monitoring systems, and clear purpose-aligned objectives while being bold enough to seize the health- and science-advancing opportunities that AGI presents [27-28][43].


He called for inclusive governance, urging governments, artists, social scientists, philosophers, and the broader public to join technologists and scientists in shaping AI policy [44-46].


He concluded that if the community navigates this historic crossroads with scientific rigor, humility, and interdisciplinary collaboration, the world can usher in a new “golden era” of discovery and improved human health, turning AGI’s disruptive potential into a force for the betterment of all humanity [43][46].


Session transcriptComplete transcript of the session
Speaker 1

much, Mr. Mukesh Dhirubhai Ambani for your strong belief in India’s capabilities and also for being at the forefront of India’s AI revolution. Ladies and gentlemen, let’s have a big round of applause for Mr. Ambani. And now I would like to invite Sir Damis Hassabis. He is the founder and CEO of Google DeepMind, the co -founder of Google DeepMind, and well, he is also the Nobel laureate, a chess prodigy, neuroscientist, and game designer before he became one of the world’s leading artificial intelligence researchers. Sir Damis Hassabis brings an almost uniquely cross -disciplinary mind to the challenge of building artificial general intelligence. DeepMind’s AlphaFold solved a 50 -year -old problem in biology, and he’s, well, just getting started.

Please welcome the co -founder and CEO of Google DeepMind, Sir Hesibus.

Demis Hassabis

Thank you. It’s a huge honor to be here today with so many of my esteemed colleagues from industry, academia, and government. And congratulations to Prime Minister Modi and the Indian government on convening such an impressive summit at this very pivotal moment for AI. It’s fantastic to see how the summit has evolved over the years, with the first meeting convened by Prime Minister Sunak at Bletchley Park in the UK. It’s become an incredibly important convening point for international dialogue and hopefully cooperation in the future over the future of AI. Thank you. When we started DeepMind in 2010, almost nobody was working on AI in industry. It was just a dream. And it’s been incredible to see how in the last 15 years, where we’ve come from the beginnings, those humble beginnings, to now where we are today, where the whole world is talking about AI.

The reason I’ve spent my whole career working on AI is I always believed it would be one of the most important and beneficial technologies ever invented. And for me, my passion is to advance scientific discovery. And I always felt that AI would be the ultimate tool for accelerating scientific discovery and being a force multiplier for human ingenuity. From a very young age, I’ve been obsessed with the Greek questions of science, the nature of reality, the nature of consciousness, these deep mysteries that are in the universe. And I think AI can help us find answers to these questions that we’ve pondered over for thousands of years. We’re already starting to see the beginnings of this, with systems like AlphaFold.

that we built to solve the 50 -year grand challenge of protein folding. And we hope that this will just be the first example of amazing advances in science and medicine that have been enabled by AI. And we and others are working on many other branches of science now to bring AI tools to help advance material science, fusion, physics, and mathematics. In fact, almost every branch of science and medicine can be impacted by AI. And now in 2026, we’re another threshold moment where AGI, artificial general intelligence, is on the horizon, maybe within the next five years. And we’re seeing these general purpose systems, foundational model systems, becoming increasingly capable almost week by week. So this is obviously an amazing opportunity that we should all grasp, economic and productivity, as well as in the sciences.

But it’s also something that we have to approach with humility and understanding. And understanding that we don’t have all the answers yet as to how this technology is going to develop and to be deployed into the world. I’ve been really impressed with what I’ve seen in India. I spent some time prior to the summit in Bangalore. We have a really big office there, Google and DeepMind, a great research office, where we do some really critical research that we then bring to our products and our technologies around the world in areas like efficient models, continual learning and multilingual capabilities. I also had time to give a talk at the Indian Institute of Science at Bangalore, and I was incredibly impressed by the students and the faculty there, their enthusiasm and their energy for AI and their ideas for how to use AI to improve India standing in the world and to also seize all the new economic opportunities and scientific opportunities that AI opens up.

It was incredibly impressive to see the energy around. And I think that India will indeed be a powerhouse for AI across the globe. We have many partnerships. Sundar, you heard Sundar announce many Google partnerships that we have with India. I’m especially proud of our deep partnership with Mr. Ambani and the Reliance Geo group to bring intelligence to everyone in India with our Gemini Foundation models in partnership with Reliance. So we hope to build on that in the next few years. So if I was to try and quantify what’s coming down the line with the advent of AGI, I think it’s going to be one of the most momentous periods in human history. Probably something more like the advent of fire or electricity.

One way maybe we can quantify that is I think it’s going to be something like 10 times the impact of the Industrial Revolution, but happening at 10 times the speed, probably unfolding in a matter of a decade rather than a century. So really, this enormous amount of change is going to come. And it’s still to be written how we can make that beneficial for the whole world. and I think the main way we should try to do this is by taking a scientific approach using the scientific method to understand what the capabilities of these systems are to build good guardrails and monitoring systems to understand more deeply what these systems are capable of and how we can make sure that they serve the purposes that we want and then of course simultaneously we also have to be bold to grasp these new opportunities to advance science and medicine and to improve human health and the human condition that society and the globe badly needs so I think we’ve got to try and navigate this moment very carefully very thoughtfully and if we do so I’m very optimistic that will usher in a great new era and but of course this can’t just be left to technologists and that’s why summits like this are really important to bring together all parts of society the technologists that are building it the scientists that are building it the scientists that are building it the scientists that are building it governments and how to deploy it for the best use of their citizens, but also artists, social scientists, and philosophers.

We need to bring all of these debates into the tent to understand how we should navigate this next period in human history. And I think summits like this, international summits like this, are critical to encourage this kind of international dialogue and cooperation. And if we get this right, these next steps right, I think we can usher in a new golden era of scientific discovery and improve the lives and health of everyone in the world. Thank you.

Related ResourcesKnowledge base sources related to the discussion topics (7)
Factual NotesClaims verified against the Diplo knowledge base (10)
Confirmedhigh

“Demis Hassabis is the co‑founder and CEO of Google DeepMind”

The knowledge base identifies Hassabis as the co-founder and chief executive of DeepMind, confirming his role [S6] and [S4].

Confirmedhigh

“Hassabis thanked the audience and congratulated Prime Minister Modi and the Indian government for convening the summit”

A source records Hassabis extending congratulations to Prime Minister Modi during the summit [S12].

Confirmedhigh

“The inaugural meeting was convened by Prime Minister Sunak at Bletchley Park in the UK”

British Prime Minister Rishi Sunak is noted as the force behind hosting the AI summit at Bletchley Park [S35] and [S36].

Confirmedhigh

“DeepMind’s AlphaFold breakthrough solved the fifty‑year‑old protein‑folding problem”

AlphaFold is described as having solved the decades-old protein-folding challenge, confirming the claim [S7].

Confirmedmedium

“AI is a powerful catalyst for scientific discovery and one of the most important technologies ever invented”

The knowledge base highlights AI’s role as a catalyst for scientific progress and a transformative technology [S7] and [S37].

Additional Contextmedium

“DeepMind is extending AI tools to material science, fusion research, physics, mathematics, and medicine”

DeepMind’s partnership with the U.S. Department of Energy on AI-driven science demonstrates its expansion into multiple scientific domains, providing context for the claim [S41].

!
Correctionhigh

“The impact of AGI could be roughly ten‑fold the Industrial Revolution, unfolding at ten times the speed (within a decade)”

Hassabis is reported to have said AGI could be about 100 times more impactful than the Industrial Revolution, not ten-fold, indicating a discrepancy [S25].

Confirmedmedium

“Hassabis advocated a scientific‑method approach with robust guardrails, continuous monitoring, and purpose‑aligned objectives”

The source notes Hassabis’s call for thoughtful, collaborative, and safety-focused development of AI, aligning with the described approach [S6].

Additional Contextlow

“He called for inclusive governance, urging governments, artists, social scientists, philosophers, and the public to join technologists in shaping AI policy”

While not stated verbatim, the knowledge base emphasizes the need for broad, collaborative governance of AI, adding nuance to the claim [S6].

Confirmedmedium

“If managed correctly, AGI could usher in a new “golden era” of scientific discovery and improved human health”

Hassabis’s remarks about a potential golden era of discovery if AGI is handled responsibly are reflected in the source [S25].

External Sources (42)
S1
Keynote-Martin Schroeter — -Speaker 1: Role/Title: Not specified, Area of expertise: Not specified (appears to be an event moderator or host introd…
S2
Responsible AI for Children Safe Playful and Empowering Learning — -Speaker 1: Role/title not specified – appears to be a student or child participant in educational videos/demonstrations…
S3
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Vijay Shekar Sharma Paytm — -Speaker 1: Role/Title: Not mentioned, Area of expertise: Not mentioned (appears to be an event host or moderator introd…
S4
Keynote-Demis Hassabis — -Demis Hassabis: Role – Co-founder and CEO of Google DeepMind; Titles – Sir, Nobel laureate; Areas of expertise – Artifi…
S5
Folding Science / DAVOS 2025 — This discussion focused on the intersection of artificial intelligence (AI) and biology, particularly in the context of …
S6
Keynote-Demis Hassabis — Evidence:And I always felt that AI would be the ultimate tool for accelerating scientific discovery and being a force mu…
S7
Breakthroughs in human-centric bioscience with AI — This breakthrough is not happening in isolation; it forms part of a rapidly expanding constellation of AI-driven advance…
S8
Revisiting 10 AI and digital forecasts for 2025: Predictions and Reality — Demis Hassabis on AGI Development:Demis Hassabis, CEO of Google DeepMind, predicts that Artificial General Intelligence …
S9
AI Impact Summit 2026: Global Ministerial Discussions on Inclusive AI Development — And this requires proactive and coherent policy responses. First, people must be at the center of AI strategy, as we hea…
S10
Panel Discussion AI in Healthcare India AI Impact Summit — Yes, and I had a friend call me up and seeing the stock market very heavily when you guys launched your recent version, …
S11
Building Trusted AI at Scale – Keynote Anne Bouverot — Setting the Global Context and India’s Strategic Position
S12
Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session — “Our program, AlphaFold, that solved the 50‑year grand challenge of protein folding, I think is just the first example o…
S13
WS #362 Incorporating Human Rights in AI Risk Management — Multilateral Organizations and Global Cooperation Stadelmann highlights the importance of global AI summits in fosterin…
S14
Comprehensive Report: European Approaches to AI Regulation and Governance — The conversation highlighted the importance of international cooperation and alignment, even when pursuing different imp…
S15
WS #288 An AI Policy Research Roadmap for Evidence-Based AI Policy — The level of disagreement is moderate but significant for implementation. While speakers share fundamental goals of resp…
S16
Open Forum #30 High Level Review of AI Governance Including the Discussion — High level of consensus with significant implications for AI governance development. The alignment suggests that despite…
S17
Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session — The foundation for this optimism lies in India’s remarkable digital transformation over the past decade. As Mukesh Amban…
S18
Partnering on American AI Exports Powering the Future India AI Impact Summit 2026 — This comment demonstrates sophisticated understanding that ‘AI sovereignty’ isn’t a monolithic concept but represents di…
S19
Keynote-Mukesh Dhirubhai Ambani — Ambani framed artificial intelligence as the cornerstone of India’s transformation into a fully developed nation by 2047…
S20
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Jeetu Patel President and Chief Product Officer Cisco Inc — Explanation:Both speakers offer unqualified praise for India’s AI initiatives and leadership without presenting any chal…
S21
Agents of Change AI for Government Services & Climate Resilience — Explanation:Unexpectedly, there was strong consensus that industry should take a proactive role in developing AI standar…
S22
Agentic AI in Focus Opportunities Risks and Governance — This discussion represents a significant evolution in AI governance thinking, moving from abstract principles to practic…
S23
AI-Powered Chips and Skills Shaping Indias Next-Gen Workforce — The discussion revealed strong alignment between industry needs, academic capabilities, and government policy. David Fre…
S24
Keynote-Demis Hassabis — Ladies and gentlemen, let’s have a big round of applause for Mr. Ambani. And now I would like to invite Sir Damis Hassab…
S25
Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session — Evidence:AlphaFold solving 50-year grand challenge of protein folding as first example, impact happening over a decade i…
S26
Folding Science / DAVOS 2025 — Alison Snyder: Thank you all for being here this morning. Thank you to those of you watching online. In industry the b…
S27
Keynote-Demis Hassabis — Evidence:We’re already starting to see the beginnings of this, with systems like AlphaFold. that we built to solve the 5…
S28
AI Governance Dialogue: Steering the future of AI — Development | Sociocultural Last year, the Nobel Prize for Chemistry was awarded to the developers of AlphaFold, an AI …
S29
Building Trusted AI at Scale – Keynote Anne Bouverot — Setting the Global Context and India’s Strategic Position
S30
Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session — “Our program, AlphaFold, that solved the 50‑year grand challenge of protein folding, I think is just the first example o…
S31
Keynote-Mukesh Dhirubhai Ambani — Ambani makes a bold prediction about India’s future position in global AI leadership. He expresses confidence that India…
S32
Keynote_ 2030 – The Rise of an AI Storytelling Civilization _ India AI Impact Summit — Speaker 1’s presentation represents a masterful progression from current state analysis to future vision, punctuated by …
S33
Leaders’ Plenary | Global Vision for AI Impact and Governance Morning Session Part 1 — Artificial intelligence requires enormous competition. Artificial capacity, which in turn requires unprecedented amounts…
S34
UK seeks alignment with EU on AI policy framework and copyright issues — As part of a warmup of relations on science and technology, Jonathan Berry, the UK’s AI minister, used positive language…
S35
UK PM Sunak urges for government action on AI risks — British Prime Minister Rishi Sunak has emphasised the need for governments to addressthe risks associated with AI. Sunak…
S36
Keynote-Rishi Sunak — Evidence:Sunak was the force behind hosting the landmark AI Safety Summit at Bletchley Park, described as the point wher…
S37
Artificial intelligence: a catalyst for scientific discovery and advancement — While concerns about AI’s dangers abound, experts believe that it can greatly accelerate scientific progress and lead to…
S38
9821st meeting — France:Secretary of State, ministers, ladies and gentlemen, I must first and foremost thank the Secretary General for hi…
S39
The rise of tech giants in healthcare: How AI is reshaping life sciences — The intersection of technology and healthcareis rapidly evolving, fuelled by advancements in ΑΙ and driven by major tech…
S40
Competing visions of AGI emerge at Google DeepMind and Microsoft — Two former DeepMind co-founders now leading rival AI labs haveoutlined sharply different visionsfor how artificial gener…
S41
Google DeepMind partners with DOE for AI-driven science — Google DeepMind ispartnering with the US Department of Energy(DOE) to support the White House’s Genesis Mission, a natio…
S42
JANUARY 14 TH , 2019 — 14 C4 – Competence Centre in Cloud Computing: http://c4.ubi.pt a complete and true interdisciplinary way with c…
Speakers Analysis
Detailed breakdown of each speaker’s arguments and positions
D
Demis Hassabis
10 arguments161 words per minute1057 words393 seconds
Argument 1
AI accelerates scientific discovery across all fields (Demis Hassabis)
EXPLANATION
Hassabis states that AI serves as a universal accelerator for research, enabling faster breakthroughs in every scientific discipline. He emphasizes that the technology multiplies human ingenuity and can transform both science and medicine.
EVIDENCE
He explains that his passion is to advance scientific discovery and that AI is the ultimate tool for accelerating it, noting that “almost every branch of science and medicine can be impacted by AI” and citing examples such as material science, fusion, physics, and mathematics [16-23].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Hassabis describes AI as a powerful accelerator and force-multiplier for scientific research across disciplines, emphasizing its transformative potential [S4][S6].
MAJOR DISCUSSION POINT
AI as a catalyst for scientific discovery
Argument 2
AI can help answer fundamental questions about reality and consciousness (Demis Hassabis)
EXPLANATION
Hassabis argues that AI can be employed to explore deep philosophical questions concerning the nature of reality and consciousness. He believes that machine intelligence may provide new insights into mysteries that have puzzled humanity for millennia.
EVIDENCE
He recounts his lifelong obsession with the Greek questions of science, reality, and consciousness, and asserts that “AI can help us find answers to these questions that we’ve pondered over for thousands of years” [18-19].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
He recounts his lifelong obsession with the Greek questions of reality and consciousness and argues that AI can provide new insights into these mysteries [S6].
MAJOR DISCUSSION POINT
AI addressing fundamental scientific and philosophical questions
Argument 3
AlphaFold solved the 50‑year protein‑folding problem, demonstrating AI’s breakthrough potential (Demis Hassabis)
EXPLANATION
Hassabis highlights AlphaFold as a concrete example of AI achieving a historic scientific milestone, solving a problem that had resisted solution for five decades. This achievement showcases the transformative power of AI in biology and beyond.
EVIDENCE
He points to the development of AlphaFold, which “solved the 50-year grand challenge of protein folding” and frames it as the first of many AI-enabled advances in science and medicine [20-21].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
AlphaFold is highlighted as having solved the decades-old protein-folding challenge and is now used by thousands of scientists worldwide [S7][S5].
MAJOR DISCUSSION POINT
Proof of AI’s breakthrough capability
Argument 4
Foundational models are improving weekly; AGI may arrive within five years (Demis Hassabis)
EXPLANATION
Hassabis notes the rapid, continuous improvement of large foundational models, suggesting that artificial general intelligence is approaching within a short horizon. He predicts that AGI could emerge in the next five years.
EVIDENCE
He states that we are at a “threshold moment where AGI, artificial general intelligence, is on the horizon, maybe within the next five years” and observes that “general purpose systems, foundational model systems, becoming increasingly capable almost week by week” [24-26].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
He states that AGI is on the horizon, possibly within five years, and notes that foundational model systems become increasingly capable week by week [S4][S6].
MAJOR DISCUSSION POINT
Progress toward AGI
Argument 5
India’s students, faculty, and research environment impressed DeepMind visitors (Demis Hassabis)
EXPLANATION
During his visit to Bangalore, Hassabis was struck by the enthusiasm and talent of Indian academia. He believes this vibrant ecosystem positions India as a future global AI powerhouse.
EVIDENCE
He describes his time at the Indian Institute of Science, noting the “students and the faculty there, their enthusiasm and their energy for AI” and calls the experience “incredibly impressive” [31-34].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Hassabis praises the enthusiasm and talent of Indian academia and cites DeepMind’s research office in Bangalore as evidence of India’s AI potential [S6][S4].
MAJOR DISCUSSION POINT
India’s AI ecosystem
AGREED WITH
Speaker 1
Argument 6
Deep partnership with Reliance and Gemini foundation models aims to bring intelligence to every Indian (Demis Hassabis)
EXPLANATION
Hassabis announces a strategic collaboration with Reliance’s Geo group to deploy Gemini foundation models across India, aiming for widespread AI access. The partnership is presented as a cornerstone for national AI adoption.
EVIDENCE
He mentions being “especially proud of our deep partnership with Mr. Ambani and the Reliance Geo group to bring intelligence to everyone in India with our Gemini Foundation models” and expresses intent to expand the effort in the coming years [36-38].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
He announces a deep partnership with Mr. Ambani and Reliance Geo to deploy Gemini foundation models across India, aiming for widespread AI access [S4][S6].
MAJOR DISCUSSION POINT
DeepMind/Google partnership with Indian industry
AGREED WITH
Speaker 1
Argument 7
AGI could have an impact ten times greater than the Industrial Revolution, unfolding within a decade (Demis Hassabis)
EXPLANATION
Hassabis compares the forthcoming AGI era to historic revolutions, estimating its economic and societal impact to be an order of magnitude larger than the Industrial Revolution, but compressed into a much shorter timespan. He warns that the world must prepare for rapid change.
EVIDENCE
He quantifies the potential change as “something like 10 times the impact of the Industrial Revolution, but happening at 10 times the speed, probably unfolding in a matter of a decade rather than a century” [40-42].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
He quantifies AGI’s potential impact as roughly tenfold that of the Industrial Revolution and notes it could unfold at ten times the speed, within a decade [S4].
MAJOR DISCUSSION POINT
Anticipated societal and economic impact of AGI
Argument 8
Apply the scientific method to build guardrails, monitoring, and understanding of AI capabilities (Demis Hassabis)
EXPLANATION
Hassabis advocates for a rigorous, evidence‑based approach to AI governance, using the scientific method to assess capabilities and design safety mechanisms. He stresses the need for systematic monitoring and robust guardrails.
EVIDENCE
He calls for “taking a scientific approach using the scientific method to understand what the capabilities of these systems are to build good guardrails and monitoring systems” [43].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Hassabis calls for a scientific, evidence-based approach to assess AI capabilities and design robust guardrails and monitoring systems [S6][S4].
MAJOR DISCUSSION POINT
Responsible AI governance through scientific methodology
Argument 9
Involve governments, artists, social scientists, and philosophers in shaping AI’s future (Demis Hassabis)
EXPLANATION
Hassabis stresses that AI’s trajectory cannot be decided by technologists alone; a broad coalition of policymakers, cultural creators, and scholars must contribute. This multidisciplinary input is essential for aligning AI with societal values.
EVIDENCE
He notes that “this can’t just be left to technologists and that’s why summits like this are really important to bring together … governments … artists, social scientists, and philosophers” [43].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
He stresses that AI’s trajectory must involve technologists, governments, artists, social scientists, and philosophers, emphasizing multidisciplinary dialogue at summits [S6][S4].
MAJOR DISCUSSION POINT
Multidisciplinary participation in AI governance
Argument 10
International summits are essential for dialogue, cooperation, and aligning AI development with global benefit (Demis Hassabis)
EXPLANATION
Hassabis argues that global forums are crucial for fostering cooperation, sharing perspectives, and ensuring AI advances serve humanity as a whole. He views such summits as a platform for building consensus on responsible AI deployment.
EVIDENCE
He states that “summits like this, international summits like this, are critical to encourage this kind of international dialogue and cooperation” and that getting the next steps right can “usher in a new golden era of scientific discovery and improve the lives and health of everyone in the world” [44-46].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
He argues that international summits are critical for fostering cooperation, sharing perspectives, and ensuring AI serves humanity as a whole [S6][S4].
MAJOR DISCUSSION POINT
Need for international cooperation on AI
AGREED WITH
Speaker 1
S
Speaker 1
1 argument118 words per minute133 words67 seconds
Argument 1
Recognition of Ambani’s leadership and India’s AI capabilities (Speaker 1)
EXPLANATION
The opening speaker publicly acknowledges Mukesh Dhirubhai Ambani’s belief in India’s potential and his role in driving the nation’s AI revolution. This praise sets a tone of national pride and highlights leadership in the AI sector.
EVIDENCE
He thanks Mr. Ambani for his “strong belief in India’s capabilities” and applauds him, calling for a round of applause and positioning him as a key figure at the forefront of India’s AI revolution [1-3].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
The opening remarks thank Mr. Ambani for his strong belief in India’s AI capabilities and call for applause, highlighting his leadership role [S4].
MAJOR DISCUSSION POINT
Acknowledgment of Indian AI leadership
AGREED WITH
Demis Hassabis
Agreements
Agreement Points
Recognition of India’s AI leadership and Ambani’s role in driving AI adoption
Speakers: Speaker 1, Demis Hassabis
Recognition of Ambani’s leadership and India’s AI capabilities (Speaker 1) India’s students, faculty, and research environment impressed DeepMind visitors (Demis Hassabis) Deep partnership with Reliance and Gemini foundation models aims to bring intelligence to every Indian (Demis Hassabis)
Both speakers highlighted India’s growing AI ecosystem, praising Mukesh Ambani’s leadership and the enthusiasm of Indian academia, and noted the partnership between DeepMind/Google and Reliance to deploy Gemini models across India [1-3][31-34][36-38].
POLICY CONTEXT (KNOWLEDGE BASE)
This recognition aligns with India’s National AI Strategy and was prominently highlighted at the India AI Impact Summit 2026, where Mukesh Ambani described AI as the cornerstone of India’s transformation and noted the country’s rapid rise from a low connectivity ranking to global digital leadership [S17][S19][S20].
Importance of international summits and cooperation for AI governance
Speakers: Speaker 1, Demis Hassabis
Recognition of Ambani’s leadership and India’s AI capabilities (Speaker 1) (implies support for summit) International summits are essential for dialogue, cooperation, and aligning AI development with global benefit (Demis Hassabis)
Both speakers underscored the value of the summit as a platform for international dialogue and cooperation on AI, with the opening remarks framing the event as impressive and Demis explicitly stating that such summits are critical for global cooperation [1][44-46].
POLICY CONTEXT (KNOWLEDGE BASE)
The emphasis on AI summits reflects broader multilateral calls for coordinated governance; Stadelmann stresses the upcoming India summit as a venue to foreground Global South perspectives [S13], while the European report underscores the need for cross-regional cooperation despite differing implementation paths [S14]; high-level reviews report strong consensus on governance direction [S16].
Similar Viewpoints
Hassabis consistently presents AI as a powerful catalyst for scientific breakthroughs, using AlphaFold as concrete evidence of its transformative capacity [16-23][20-21].
Speakers: Demis Hassabis
AI accelerates scientific discovery across all fields (Demis Hassabis) AlphaFold solved the 50‑year protein‑folding problem, demonstrating AI’s breakthrough potential (Demis Hassabis)
He links rapid progress of foundational models to an imminent AGI era with unprecedented societal impact [24-26][40-42].
Speakers: Demis Hassabis
Foundational models are improving weekly; AGI may arrive within five years (Demis Hassabis) AGI could have an impact ten times greater than the Industrial Revolution, unfolding within a decade (Demis Hassabis)
He advocates a multidisciplinary, evidence‑based governance approach that combines scientific assessment with broad societal participation [43][43].
Speakers: Demis Hassabis
Apply the scientific method to build guardrails, monitoring, and understanding of AI capabilities (Demis Hassabis) Involve governments, artists, social scientists, and philosophers in shaping AI’s future (Demis Hassabis)
Unexpected Consensus
Industry‑government partnership as a vehicle for nationwide AI deployment
Speakers: Speaker 1, Demis Hassabis
Recognition of Ambani’s leadership and India’s AI capabilities (Speaker 1) Deep partnership with Reliance and Gemini foundation models aims to bring intelligence to every Indian (Demis Hassabis)
While Speaker 1’s remarks focus on national pride, Demis’s detailed partnership announcement shows a concrete industry-government collaboration, revealing an unexpected alignment on the strategic role of private-public alliances for AI diffusion [1-3][36-38].
Overall Assessment

The two speakers converge on celebrating India’s AI potential, the significance of the summit, and the need for collaborative frameworks that involve both industry and broader societal actors.

High consensus on the strategic importance of India’s AI ecosystem and international cooperation, suggesting strong momentum for coordinated policy and investment initiatives.

Differences
Different Viewpoints
Unexpected Differences
Overall Assessment

The exchange shows strong convergence rather than conflict. Speaker 1’s opening remarks praise Indian leadership in AI, and Demis Hassabis reinforces that view by describing India’s vibrant research community and a strategic partnership with Reliance. No substantive points of contention emerge between the speakers.

Minimal – the dialogue is largely complementary, indicating a unified stance on promoting India’s AI development and the need for collaborative, multidisciplinary engagement.

Partial Agreements
Both speakers commend India’s AI ecosystem and highlight Mukesh Ambani’s pivotal role in driving AI progress in the country. Speaker 1 publicly thanks Ambani for his strong belief in India’s capabilities and applauds him as a front‑runner of India’s AI revolution [1-3], while Hassabis describes his visit to Bangalore as “incredibly impressive” and stresses that India will be a global AI powerhouse, also noting a deep partnership with Reliance (Ambani’s group) to bring Gemini models to every Indian [31-34][36-38].
Speakers: Speaker 1, Demis Hassabis
Recognition of Ambani’s leadership and India’s AI capabilities (Speaker 1) India’s students, faculty, and research environment impressed DeepMind visitors (Demis Hassabis)
Takeaways
Key takeaways
AI is a powerful catalyst that can accelerate scientific discovery across all fields and help address fundamental questions about reality and consciousness. Foundational models are improving rapidly; breakthroughs like AlphaFold demonstrate AI’s potential, and AGI may emerge within the next five years. India’s AI ecosystem—students, faculty, research institutions, and industry partners—is vibrant and positioned to become a global AI powerhouse. DeepMind and Google have a strategic partnership with Reliance (via Gemini foundation models) to bring advanced AI capabilities to a broad Indian audience. The societal and economic impact of AGI could be ten times that of the Industrial Revolution, unfolding within a decade, requiring careful stewardship. Responsible development demands a scientific‑method approach to guardrails and monitoring, and must involve governments, artists, social scientists, philosophers, and other stakeholders through international cooperation.
Resolutions and action items
DeepMind will deepen its collaboration with Reliance’s Geo group to deploy Gemini foundation models across India. DeepMind will continue its Bangalore research on efficient models, continual learning, and multilingual capabilities, feeding advances into global products. Commit to applying the scientific method to design guardrails, monitoring systems, and capability assessments for emerging AI systems. Organize and participate in future international summits that bring together technologists, scientists, policymakers, artists, and ethicists to shape AI governance.
Unresolved issues
Specific governance frameworks and regulatory mechanisms needed to ensure AGI benefits all of humanity. Concrete safety and alignment strategies for AGI expected within the next five years. How the economic gains from rapid AI adoption will be distributed equitably, especially in emerging economies. Detailed policies for cross‑border AI collaboration, data sharing, and intellectual‑property rights. Mechanisms for ongoing monitoring of AI impacts on society, health, and the environment.
Suggested compromises
Pursue aggressive AI research and deployment while simultaneously instituting scientifically‑driven guardrails and oversight. Balance the drive for economic and scientific opportunities with humility and caution regarding unknown risks.
Thought Provoking Comments
AI will be the ultimate tool for accelerating scientific discovery and being a force multiplier for human ingenuity.
Frames AI not merely as a commercial technology but as a fundamental catalyst for all branches of science, expanding the conversation from industry applications to global knowledge creation.
Sets the thematic foundation of the talk, prompting the audience to view AI through a scientific lens and opening space for later discussion of specific breakthroughs like AlphaFold.
Speaker: Demis Hassabis
AlphaFold solved a 50‑year‑old problem in biology; this is just the first example of amazing advances in science and medicine that AI can enable.
Provides a concrete, high‑impact case study that validates the earlier claim about AI as a scientific accelerator, turning abstract optimism into tangible evidence.
Strengthens credibility of the argument, leading the audience to anticipate further AI‑driven breakthroughs and reinforcing the narrative that AI’s potential is already being realized.
Speaker: Demis Hassabis
We are at another threshold moment where AGI is on the horizon, maybe within the next five years, and general‑purpose models are becoming increasingly capable almost week by week.
Introduces a time‑bound prediction that shifts the conversation from past achievements to imminent, transformative change, raising the stakes for policy and governance discussions.
Creates a sense of urgency, prompting listeners to consider immediate actions and preparing the ground for later remarks on safety, guardrails, and interdisciplinary involvement.
Speaker: Demis Hassabis
The impact of AGI could be something like ten times the impact of the Industrial Revolution, but happening at ten times the speed, likely unfolding within a decade.
Offers a bold, quantifiable analogy that helps the audience grasp the magnitude and rapidity of the upcoming transformation, challenging any complacent views of gradual progress.
Acts as a turning point that intensifies the tone of the talk, moving from hopeful optimism to a call for proactive preparation, and justifies the need for robust governance frameworks.
Speaker: Demis Hassabis
We must approach this with humility, using the scientific method to understand capabilities, build guardrails and monitoring systems, while also being bold enough to seize opportunities for health and scientific advancement.
Balances caution with ambition, introducing a nuanced stance that acknowledges unknowns while urging decisive action—a departure from purely celebratory or alarmist narratives.
Steers the discussion toward concrete next steps (research, safety, policy) and invites collaboration across sectors, setting the stage for the later call to include non‑technical voices.
Speaker: Demis Hassabis
This cannot just be left to technologists; we need artists, social scientists, philosophers, governments, and the broader public to join the debate and shape how we navigate this period.
Expands the conversation beyond the technical community, emphasizing interdisciplinary governance and democratic participation, which challenges any siloed approach to AI development.
Broadens the scope of the summit, encouraging a more inclusive dialogue and reinforcing the importance of international cooperation, thereby aligning the audience’s expectations for collaborative outcomes.
Speaker: Demis Hassabis
Overall Assessment

The identified comments collectively transformed the speech from a simple showcase of AI achievements into a multidimensional discourse on imminent AGI, its historic scale, and the societal responsibilities it entails. Early remarks established AI as a scientific catalyst, while later predictions about AGI’s timeline and impact injected urgency. The balanced call for humility, safety, and bold exploitation, followed by an explicit invitation to non‑technical stakeholders, shifted the tone toward inclusive, proactive governance. These pivotal statements guided the audience’s focus, introduced new thematic layers, and set the agenda for the summit’s broader collaborative objectives.

Follow-up Questions
How can we develop effective guardrails and monitoring systems to ensure AGI behaves safely and aligns with human values?
Critical for preventing unintended harmful outcomes as AGI becomes more capable and for building public trust.
Speaker: Demis Hassabis
What scientific methods can be employed to systematically assess the capabilities of emerging AI systems?
A rigorous, evidence‑based approach is needed to understand system limits and inform safety and policy decisions.
Speaker: Demis Hassabis
How can AI be leveraged to accelerate discovery in material science, fusion energy, physics, and mathematics?
These scientific domains could see transformative breakthroughs, but require focused research on integrating AI tools effectively.
Speaker: Demis Hassabis
What are the most promising approaches for efficient models, continual learning, and multilingual capabilities, especially in the Indian context?
Improving model efficiency and adaptability is essential for scaling AI access across India’s diverse languages and resource constraints.
Speaker: Demis Hassabis
How should societies deploy AI technologies to maximize economic productivity while ensuring equitable benefits for all citizens?
Balancing economic gains with social fairness is vital to avoid widening inequality as AI reshapes labor markets.
Speaker: Demis Hassabis
How can international cooperation be structured to manage the rapid development of AGI and share its benefits globally?
Coordinated global governance can mitigate risks, align standards, and distribute the advantages of AGI worldwide.
Speaker: Demis Hassabis
What role should artists, social scientists, and philosophers play in shaping AI governance and ethical frameworks?
Interdisciplinary input is needed to address cultural, ethical, and societal dimensions that technologists alone may overlook.
Speaker: Demis Hassabis
How can we quantify the impact of AGI relative to historic technological revolutions, such as the Industrial Revolution?
Understanding the magnitude and speed of AI’s impact helps policymakers plan for infrastructure, education, and regulatory needs.
Speaker: Demis Hassabis
What specific strategies can India adopt to become a global AI powerhouse while addressing local challenges?
Leveraging India’s talent pool and market potential requires tailored policies, partnerships, and research investments to lead responsibly.
Speaker: Demis Hassabis

Disclaimer: This is not an official session record. DiploAI generates these resources from audiovisual recordings, and they are presented as-is, including potential errors. Due to logistical challenges, such as discrepancies in audio/video or transcripts, names may be misspelled. We strive for accuracy to the best of our ability.

Keynote-Brad Smith

Session at a glanceSummary, keypoints, and speakers overview

Summary

Brad Smith opened the AI summit for the Global South by highlighting the persistent economic gap between the North and South, which he attributes primarily to a technology divide rooted in unequal access to electricity and, now, AI [9-11]. He framed the central question of the summit as whether AI will close or widen this divide [12].


To address the gap, Smith outlined three priorities. First, he called for massive physical AI infrastructure-data centers, compute, connectivity, and electricity-backed by large-scale investment, noting Microsoft’s pledge to spend $50 billion by the end of the year [17-22]. He added that private capital, government funding, and demand generation must complement this effort to keep market momentum alive [26-28].


Second, he stressed the need for “skilling for people,” citing Microsoft’s Elevate for Educators program that equips teachers to teach AI and urging employers to open doors to AI tools and train employees across generations [33-39][41-45]. Third, he argued AI must be effective for the Global South by improving multilingual capabilities and applying the technology to local challenges such as Indian agriculture and African food security [48-57].


Smith warned that without these steps, AI could exacerbate inequities and raised public concerns about AI’s impact on jobs, families, and societies [59-68]. He urged continuous, coordinated effort across annual summits, proposing clear goals, common measurement systems, and yearly progress reviews to hold the global community accountable [90-96].


By linking infrastructure, skills, and context-specific AI applications, Smith concluded that AI can become a catalyst for human curiosity, health, and economic opportunity [74-79]. He emphasized that technology amplifies human potential only when governments, companies, NGOs, and individuals work together, turning AI into a tool for broader societal benefit [86-89]. The discussion closed with a call to build bridges between summits, define measurable outcomes, and use AI to build a better world for all [90-101][102-104].


Overall, Smith presented a roadmap that combines massive investment, widespread skill development, and culturally relevant AI deployment to ensure the technology benefits the Global South and addresses broader societal concerns [5-8][22][38][48].


Keypoints

Building foundational AI infrastructure in the Global South – Smith stresses that closing the economic gap requires massive investment in data centers, compute power, connectivity, and electricity, noting Microsoft’s pledge to spend $50 billion by year-end to support these efforts [17-22].


Investing in people through skilling and education – He argues that hardware alone is insufficient; widespread AI adoption hinges on “skilling for people,” highlighting Microsoft Elevate initiatives such as “Microsoft Elevate for Educators” that aim to equip teachers and workers with AI competencies [33-38].


Ensuring AI works for local languages and real-world challenges – Smith points out that AI performance is currently English-centric and must be improved for all languages. He cites upcoming investments in multilingual data, tools, and measurement, and gives concrete examples like agricultural improvements in India and food-security projects across Africa [48-57].


Addressing the societal impact of AI and establishing ongoing accountability – The speaker shifts to broader concerns about AI’s effect on jobs, families, and the future of work, urging continuous dialogue, measurable goals, and “bridges” between successive AI summits to track progress year-over-year [59-68][90-100].


Overall purpose/goal


The discussion is a call to action for governments, private sector, and civil society to collaboratively deploy AI in the Global South. It outlines three pillars-infrastructure, skills, and locally relevant AI applications-and urges the establishment of measurable, recurring commitments so that AI can narrow, rather than widen, existing economic divides.


Overall tone


The tone begins optimistic and visionary, emphasizing opportunity and the scale of investment. It then becomes inspirational and inclusive, stressing education and collective responsibility. Mid-speech, a cautious, reflective tone emerges as Smith acknowledges public anxieties about jobs and the future. The closing segment adopts a rallying, accountable tone, urging concrete follow-through, regular measurement, and sustained collaboration across global summits. Throughout, the tone remains forward-looking but grows increasingly urgent and action-oriented.


Speakers

Brad Smith – Role/Title: Vice Chair and President of Microsoft [S1][S2][S3]; Areas of expertise: Technology policy, privacy, cybersecurity, AI regulation, AI strategy [S1][S2][S3]


Speaker 1 – Role/Title: Event moderator / host [S4][S6]; Areas of expertise: 


Additional speakers:


– None


Full session reportComprehensive analysis and detailed insights

Speaker 1 opened the session by welcoming Microsoft’s Vice-Chair and President, Brad Smith, describing him as the company’s “conscience” and chief diplomat on technology policy and noting his book Tools and Weapons as a clear statement of the responsibilities that tech firms bear today [1-4].


Smith began by stressing the significance of the gathering, calling it the first AI summit for the Global South and arguing that any discussion of artificial intelligence must start with a view of the broader world in which we live [5-8]. He identified the deepest and most enduring disparity as the economic divide between the Global North and South, which he said is fundamentally a technology divide rooted in unequal access to electricity – a historic precedent that mirrors today’s AI gap [9-11]. He framed the summit’s central dilemma as whether AI will close this divide or widen it further, describing this as “the single most important question for us” [12-13].


Acknowledging the urgency of the issue, Smith asked how the community can “do better” and what will be required to achieve that [14-16]. He then outlined three inter-linked priorities.


The first priority is the construction of foundational AI infrastructure. Smith argued that the Global South needs data-centres, compute capacity, reliable connectivity and, crucially, electricity [17-20]. He noted that delivering this will demand not only cutting-edge technology but also massive investment, citing Microsoft’s commitment to spend US$50 billion by the end of the year on such infrastructure [21-24][S44][S45]. While India is highlighted as a major investment destination, he stressed that private capital, additional tech-company funding and government resources must be mobilised, and that governments need to generate demand for AI to keep market dynamics active [25-28].


The second priority is “skilling for people”. Smith asserted that infrastructure is incomplete without a workforce capable of using it [31-34]. He pointed to the launch of Microsoft Elevate initiatives, including the newly announced “Microsoft Elevate for Educators”, which equips teachers with tools to teach AI [35-39][S46]. He broadened the responsibility to all employers, urging them to open doors to AI tools and invest in upskilling employees across generations, not just the next cohort of graduates [40-45].


The third priority concerns making AI effective for the Global South. He emphasized that AI must be deployed in the Global South to address the region’s most pressing challenges [46-48]. He detailed two special initiatives: first, an upstream investment in multilingual data, tools and measurement systems to achieve linguistic parity [48-52]; second, the deployment of AI on locally-relevant problems such as agricultural productivity pilots in India and a collaborative food-security initiative across Africa [55-57].


Turning to broader societal implications, Smith raised concerns about the future of work and the anxieties of parents worldwide regarding AI’s impact on jobs, families and children’s futures [60-63]. He highlighted that human capability is neither fixed nor finite [84-85] and noted, “Compared to people of the Bronze Age we are already geniuses” [86-87]. He urged that we view AI as “the next great generator for human curiosity” [78-80] and emphasized its potential to improve health by helping cure more diseases [81-83]. To illustrate the transformative power of technology, he invoked the washing-machine analogy, explaining how AI can free time for higher-value activities just as the washing machine did [92-95].


To ensure progress is sustained, Smith argued that AI summits should not be isolated “islands” but linked by “bridges”. He called for clear, shared goals, common measurement frameworks and an annual review to assess whether the previous twelve months have delivered measurable advances [90-95][96-100].


He concluded by stressing collective responsibility-governments, companies, NGOs and employers must work together, be held accountable, and use AI to build a better world for all [101-104][S1][S7]. He closed by urging the global community to hold itself accountable and to use AI to build a better world for everyone [101-104].


Session transcriptComplete transcript of the session
Speaker 1

Ladies and gentlemen, I would like to now welcome Mr. Brad Smith, Vice Chair and President Microsoft. Mr. Brad Smith has been Microsoft’s conscience and its chief diplomat through some of the most consequential debates in technology policy, from privacy and cybersecurity to AI regulation. His book, Tools and Weapons, remains one of the most lucid accounts of the responsibilities tech companies carry in the modern world. So please welcome the Vice Chair and President of Microsoft, Mr. Brad Smith.

Brad Smith

Good afternoon. It’s always an important day when the world comes together under one roof, as we have today. It gives us an opportunity together to ask important questions, even hard questions, and think about how we want to answer them. As we think about this summit, the first AI summit in the global south, it’s only right that we start, I believe, by focusing on AI and what it means for the global south. In some ways, I think the best way to start thinking about AI is to look more broadly and think about the state of the world in which we live. We live in a tumultuous time and in a fragmented world, but I think in so many ways, the deepest and most enduring divide has been the economic divide between the global north and south.

And what I believe we need to recognize is that this economic divide is a result, more than anything else, of a technology divide. The technology divide created by unequal access to electricity. electricity became one of humanity’s most important general purpose technologies meaning it spread across economies it was applied in every industry it boosted productivity where electricity went economic development and prosperity followed but as we all know electricity did not spread everywhere at the same pace it was literally 144 years ago that the first electrical power plant started operating in lower manhattan and yet we come together today and we still live in a world where 700 million people lack access to electricity now comes ai ai perhaps perhaps more than any other technology this century will play a bigger role either in closing this economic divide or in exacerbating it and making it even wider?

That is perhaps the single most important question for us, I would suggest today, as we think about the role of AI in the global South. How can we do better? Because we need to do better. What will it take? I think it’s going to take a few things that will require that we all come together and work together. First, the obvious. We need to bring infrastructure to the global South. That means data centers and compute. It also means more connectivity. It means more electricity. That is going to take not only the world’s best technology, it’s going to require an enormous amount of investment. That’s why we at Microsoft announced yesterday morning that we’re on pace to spend $50 billion by the end of this year.

We’re going to be able to do that. We’re going to be able to do that. to bring AI to the global South. And of all the countries in which we are investing, India, not surprisingly, is one of the largest. But we’ll need to harness private capital, investments from tech companies, other sources of private capital, government funding. We’ll need governments and others to generate demand for the use of AI in the global south, because that is the only way to get the wheels of the market spinning and to do what we need to do together. That’s the first thing we need to do together. There’s a second thing we need to do. We’ve talked about it here already today, and it is so clear when you study the history of technology.

Infrastructure is not only hardware. It’s not only wires and grids. It’s skilling for people. Because the key to enabling a country and a population to use a general purpose technology at scale is to give people across the country access to the skills they need to put it to work. And that’s why it’s such good news that we see so many governments supported by the United Nations and supported by private companies investing in more skilling. It’s actually something that should speak to all of us. Certainly as a tech company, we’re committed. We’ve launched through Microsoft Elevate new initiatives, including one we’re announcing this week, Microsoft Elevate for Educators, to equip teachers with access to help their students learn how to use AI.

But in the truth, it doesn’t matter where you work. You have a role to play. Because the lesson of digital technology, I was an example of this, was that it took employers to open their doors. It took employers to computing. It takes employers today to open their doors to new AI tools. It will take employers to invest in the skilling of their employees. it’s not just for the next generation. It’s for every generation that this fully matters. Then there’s a third challenge for the Global South. We need to make AI work effectively for the Global South, and that requires some special initiatives, at least two. First, we need to make AI as effective in every language as it is in English, and today it is not.

Performance tests show that’s the case. That’s why one of the good things to come out of this week is new announcements to invest upstream in better data in other languages, to provide better tools and measurement systems for AI that is built in other languages, to build out data providence with a view to linguistic diversity, diversity that we need to advance around the world. And we need to use AI in the Global South. We need to use AI in the Global South. We need to use AI in the Global South. We need to use AI in the Global South to solve the problems that matter to the Global South. Oftentimes, as we’re doing here in India, that’s about improvements in agriculture.

Or as a number of partners, including Microsoft, are doing this week, launching a new initiative to address food security across Africa. These are just two of the myriad of opportunities we have to put AI to work in ways that will bring faster benefits to countries in most of the world. If we do those three things well, build infrastructure, invest in skilling, address real -world problems, then I think it may create the foundation to think more and do more about the question I am willing to bet will be a bigger part of the conversation in the next few AI summits. What will AI mean for the future of work and jobs? Within these walls, I think we’re going to have a lot of questions.

We’re mostly enthusiastic about the future of AI. But outside these walls, I think we need to recognize that increasingly around the world, and especially in some countries, many parents are asking a common question. What will AI mean for my kids? What will AI mean for my family? What will AI mean for our future? I get it that some people are excited and they’ll do well, but what about us? Us, meaning most of the people who live on this planet. I think we have something to prove. I think we have something to prove not only to communities and countries and our customers, but to ourselves, that we can not only embrace but pursue a brighter future for people.

It’s great to come to conferences like this and hear people talk about all of the advances in technology, but let’s remember one other thing as well. Human capability is neither fixed nor finite. It’s great to think about what it would mean to have computers in a data center that would be like a country of geniuses, but let’s also recognize this. Compared to the people who lived in the Bronze Age, all of you, all of us, are already geniuses. Whenever technology advances, it creates a new platform, a new foundation that enables people to stand taller and reach higher if, and only if, we’re committed to using that technology well. As AI makes it possible to cure more diseases, then it is right that we expect that it will improve human health.

As we use AI already every day to find faster solutions. And when we find faster answers, it gives us the opportunity to ask more questions. The fundamental fuel of human capability has always been the same. curiosity. We need to look at AI as the next great generator for human curiosity. And we need to take some inspiration because we all know the world could use a little more inspiration. I often think about the following. Before the invention of the washing machine, it took someone, almost always a woman, between six and eight hours to wash a load of laundry. But as the washing machine improved, that was compressed to 30 minutes. But do you know what happened? One thing happened more than anything else.

Everyone wanted to wear cleaner clothes. Everyone expected to wear cleaner clothes. People did their laundry a lot more often. They had better clothes and they had more time and they put that time to work to do more with their lives. ultimately that is the question for us and this too is not just a question for companies this is not what tech will do to people it is people will use technology to do for people and it’s not people who create products it’s every government it’s every company it’s every non -profit it’s every employer because we all have the opportunity to work with our people to manage through the change that is coming to show people how with the right ai skills they can create jobs and careers that will be brighter for their future that will not be easy but if that’s not our goal then we’re missing the big picture as we come away from this ai summit it reflects so much progress but i would say one thing as well each of the these ai summits is a proud moment for a great nation But we have an opportunity.

Rather than have summits that are islands that are disconnected from the summits before or that follow, we need to build bridges. We need to build bridges between these summits. We need to define clear goals. We need to have common measurement systems. And every year, we need to ask the same question. Did we make 12 months of progress in the year that just preceded our meeting? So how can we build on that progress to do more and move faster in the year ahead? For those of us who come to these meetings in different countries every year, I hope we will take that away. Let’s aim higher, not just for technology, but for what technology can do for people.

Let’s be clearer in defining what we want to accomplish. Let’s put in place the ability for us to measure our progress and all hold ourselves accountable as a global community. If we can do these things, then we can use these summits and we can use this next generation of technology to build a better world. I know that’s what we within these walls want us ourselves to do. But even more than that, I know that the people outside these walls are hoping and expecting us to do just that. Thank you very much.

Related ResourcesKnowledge base sources related to the discussion topics (13)
Factual NotesClaims verified against the Diplo knowledge base (5)
Confirmedhigh

“Brad Smith was introduced as Microsoft’s Vice‑Chair and President, described as the company’s “conscience” and chief diplomat on technology policy.”

The moderator’s introduction explicitly calls Brad Smith Microsoft’s Vice-Chair and President and highlights his role as the company’s conscience and chief diplomat in technology-policy debates [S7].

Confirmedhigh

“Brad Smith holds the titles Vice‑Chair and President of Microsoft.”

Multiple sources list Brad Smith as Microsoft’s Vice-Chair and President, including the keynote introduction and Davos 2025 appearance [S7] and [S8].

Additional Contextmedium

“The first priority is the construction of foundational AI infrastructure – data‑centres, compute capacity, reliable connectivity and electricity – and this will require massive investment.”

The AI-impact summit agenda also emphasizes building infrastructure (data-centres, grid capacity, connectivity) as a key priority, providing detail on the types of assets needed and the scale of investment required [S64] and [S68].

Additional Contextlow

“India is highlighted as a major investment destination for AI infrastructure.”

The India AI Impact Summit 2026 discussion underscores India’s central role in partnership and infrastructure initiatives, adding nuance to the claim that India is a primary focus for investment [S64].

Additional Contextmedium

“Smith outlined three inter‑linked priorities: infrastructure, skilling, and making AI effective for the Global South.”

The summit’s partner briefing also structures its agenda around three pillars-building infrastructure, fostering innovation/skilling, and partnership-which aligns with Smith’s three-part framework [S64].

External Sources (68)
S1
Keynote-Brad Smith — -Brad Smith: Role/Title: Vice Chair and President of Microsoft; Areas of expertise: Technology policy, privacy, cybersec…
S2
Brad Smith — As Microsoft’s vice chair and president, Brad Smith leads a team of more than 1,900 business, legal and corporate affair…
S3
A Digital Future for All (morning sessions) — – Brad Smith – Vice Chair and President, Microsoft Brad Smith: Well, I know the time is running out. Let me be brief,…
S4
Keynote-Martin Schroeter — -Speaker 1: Role/Title: Not specified, Area of expertise: Not specified (appears to be an event moderator or host introd…
S5
Responsible AI for Children Safe Playful and Empowering Learning — -Speaker 1: Role/title not specified – appears to be a student or child participant in educational videos/demonstrations…
S6
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Vijay Shekar Sharma Paytm — -Speaker 1: Role/Title: Not mentioned, Area of expertise: Not mentioned (appears to be an event host or moderator introd…
S7
Keynote-Brad Smith — Smith argues that the economic divide between global north and south stems fundamentally from a technology divide, parti…
S8
AI: Lifting All Boats / DAVOS 2025 — – Brad Smith: Vice Chair and President of Microsoft Brad Smith: And nonprofits are playing a key role as well. Brad S…
S9
Towards a Safer South Launching the Global South AI Safety Research Network — Ms. Crampton commits Microsoft to fulfilling the New Delhi Frontier AI commitments regarding multilingual and multicultu…
S10
Welfare for All Ensuring Equitable AI in the Worlds Democracies — Thank you. Yes, skills gap is really important. We see it as part of the sort of foundational infrastructure for what we…
S11
Global Digital Compact: AI solutions for a digital economy inclusive and beneficial for all — Jean‑Francois Saint‑Pierre: Thank you very much. And thank you for having us. I’m happy today to introduce Microsoft Ele…
S12
AI for Social Empowerment_ Driving Change and Inclusion — Education and Skills System Overhaul:Investment requires fundamental reimagining rather than incremental improvement. Cu…
S13
Discussion Report: AI Implementation and Global Accessibility — Chadha articulated the challenge of preparing current students for unknown future jobs, noting that “today’s education i…
S14
Open Forum #64 Local AI Policy Pathways for Sustainable Digital Economies — Abhishek Singh: Thank you for convening this and bringing this very, very important subject at FORC, like how do we bala…
S15
How can AI improve multilingualism — ChatGPT-4 performances in languages other than English are of lower qulity. In a recent test, ChatGPT-4 scored 85% on a …
S16
Driving Indias AI Future Growth Innovation and Impact — These key comments fundamentally shaped the discussion by expanding it beyond technical infrastructure to encompass trus…
S17
Artificial intelligence — Content policy Future of work Sustainable development
S18
Comprehensive Report: Preventing Jobless Growth in the Age of AI — Moderate disagreement with significant implications. While speakers agreed on broad goals, their different assessments o…
S19
Open Forum #64 Local AI Policy Pathways for Sustainable Digital Economies — Clear metrics are needed to move beyond vague principles and create accountability mechanisms
S20
Keynote-Brad Smith — Smith argues that AI summits should not be disconnected events but should build bridges between meetings with clear goal…
S21
How Multilingual AI Bridges the Gap to Inclusive Access — The discussion produced several concrete commitments, including a collaboration between Current AI and Bhashini announce…
S22
Panel Discussion Data Sovereignty India AI Impact Summit — “One, of course, is basically the policies need to evolve along with the infrastructure.”[37]. “As far as governments ar…
S23
Interdisciplinary approaches — AI-related issues are being discussed in various international spaces. In addition to the EU, OECD, and UNESCO, organisa…
S25
A Digital Future for All (afternoon sessions) — AI governance requires a multi-stakeholder approach due to the diverse nature of opportunities, risks, and inclusivity c…
S26
WS #288 An AI Policy Research Roadmap for Evidence-Based AI Policy — Multi-stakeholder partnerships between policy researchers and private sector are essential for surfacing potential harms…
S27
From Technical Safety to Societal Impact Rethinking AI Governanc — ing used as well. So it might perform really good in English, but we know that these systems are not safe or secure or p…
S28
Welfare for All Ensuring Equitable AI in the Worlds Democracies — of seven. So other than the incentives that we’re giving you to learn these technologies, which of course is to the comp…
S29
How AI Is Transforming Indias Workforce for Global Competitivene — In diverse countries like India, AI systems must understand cultural context, multiple languages, dialects, and informal…
S30
AI and Global Power Dynamics: A Comprehensive Analysis of Economic Transformation and Geopolitical Implications — I think we should, let’s not talk about Saudi Arabia or India for a moment, but let’s just talk about the global north a…
S31
Open Forum #30 Harnessing GenAI to transform Education for All — Mohamed Shareef argues that generative AI is creating a new front in the digital divide, as developed countries invest h…
S32
High Level Session 3: AI & the Future of Work — Nthati Moorosi: Thank you very much. I think the biggest words for me in closing is inclusivity is collaboration. AI-dri…
S33
WS #362 Incorporating Human Rights in AI Risk Management — – Nathalie Stadelmann- Min thu Aung Global South Perspectives and Context-Specific Challenges Human rights | Developme…
S34
A Global Human Rights Approach to Responsible AI Governance | IGF 2023 WS #288 — Different governments and countries are adopting varied approaches to AI governance. The transition from policy to pract…
S35
Transforming Agriculture_ AI for Resilient and Inclusive Food Systems — 1 ,000 hectares in some big island of Indonesia in order to get the safe efficiency in the next five years. And then we …
S36
Lightning Talk #173 Artificial Intelligence in Agrotech and Foodtech — Comprehensive support systems are needed for agricultural innovation Development | Legal and regulatory Supporting Agr…
S37
Transforming Agriculture_ AI for Resilient and Inclusive Food Systems — Thank you, Sarah. Is this working? Yeah. Thank you all for sharing this wonderful moment for me because we’re here with …
S38
Artificial intelligence (AI) – UN Security Council — The discussions on structuring capacity-building initiatives in AI to maximize their impact, especially in regions with …
S39
Keynote-Brad Smith — That is perhaps the single most important question for us, I would suggest today, as we think about the role of AI in th…
S40
Unleashing Digital Trade and Investment for Sustainable Development (UN ESCAP) — Skill development especially amongst the marginalized is critical
S41
Youth-Driven Tech: Empowering Next-Gen Innovators | IGF 2023 WS #417 — The importance of accessible knowledge, a suitable environment, and mentors is emphasized for youth to make a significan…
S42
WS #462 Bridging the Compute Divide a Global Alliance for AI — Development | Sociocultural Ivy argues that having chips and compute power isn’t sufficient without entrepreneurs and p…
S43
Big Ideas from Small Economies / Davos 2025 — Education and skills development are crucial
S44
Keynote-Brad Smith — Infrastructure investment requirements: The need for massive investment in data centers, compute power, connectivity, an…
S45
Keynote-Brad Smith — -Infrastructure investment requirements: The need for massive investment in data centers, compute power, connectivity, a…
S46
Global Digital Compact: AI solutions for a digital economy inclusive and beneficial for all — Jean‑Francois Saint‑Pierre: Thank you very much. And thank you for having us. I’m happy today to introduce Microsoft Ele…
S47
AI for Social Empowerment_ Driving Change and Inclusion — This insight suggests that the future of work may increasingly centre on fundamentally human capabilities rather than te…
S48
Skilling and Education in AI — This discussion focused on leveraging artificial intelligence as a tool for development and equality in India, examining…
S49
Open Forum #64 Local AI Policy Pathways for Sustainable Digital Economies — Abhishek Singh: Thank you for convening this and bringing this very, very important subject at FORC, like how do we bala…
S50
How can AI improve multilingualism — ChatGPT-4 performances in languages other than English are of lower qulity. In a recent test, ChatGPT-4 scored 85% on a …
S51
How Small AI Solutions Are Creating Big Social Change — It has been announced just today in the AI Summit. And we will be allocating 5 .5 million to support data collection for…
S52
Open Internet Inclusive AI Unlocking Innovation for All — “I think, firstly, it’s important that India is not trying to get to AGI”[30]. “We need to uplift 100 million farmers, a…
S53
Artificial intelligence — Future of work
S54
Discussion Report: AI Implementation and Global Accessibility — These key comments transformed what could have been a superficial discussion about AI benefits into a nuanced exploratio…
S55
Open Forum #27 Make Your AI Greener a Workshop on Sustainable AI Solutions — Mario Nobile emphasized that sustainability, energy consumption, and job displacement from AI cannot be addressed in iso…
S56
Comprehensive Report: AI’s Impact on the Future of Work – Davos 2026 Panel Discussion — Bhan highlights the widespread concern and uncertainty about how AI will affect employment at different time horizons. S…
S57
Why science metters in global AI governance — -Brad Smith- Vice Chair and President of Microsoft Corporation
S58
Press Briefing by HMIT Ashwani Vaishnav on AI Impact Summit 2026 l Day 5 — Congratulations on the declaration, sir. I just wanted to know, could you give us names of some of the countries that ha…
S59
Scaling Enterprise-Grade Responsible AI Across the Global South — I think this is a very important question. We do see the scale of innovation and the pace. This is going at so high rate…
S60
Shaping the Future AI Strategies for Jobs and Economic Development — Hi, good morning everybody. I’ve got an incredible panel here this morning. The topic that we have is, I think, the most…
S61
Comprehensive Discussion Report: AI’s Existential Challenge to Human Identity and Society — Harari explained that if thinking primarily involves organizing words and language tokens, then AI already surpasses man…
S63
Building a Digital Society, from Vision to Implementation — Christopher Reckord opened by emphasizing that small island states have unique opportunities to leapfrog technological b…
S64
Partnering on American AI Exports Powering the Future India AI Impact Summit 2026 — Secretary S. Krishnan outlined three key priorities: building infrastructure, focusing on innovation, and maintaining a …
S65
Building Public Interest AI Catalytic Funding for Equitable Compute Access — Dr. Garg also referenced observations about the contrast between current AI systems requiring gigawatts of power and hum…
S66
Sam Altman officially returns to OpenAI — Sam Altman is officially returning as the CEO of OpenAI, with Mira Murati resuming her Chief Technology Officer (CTO) ro…
S67
Partnering on American AI Exports Powering the Future India AI Impact Summit 2026 — That’s true. I think the energy has been amazing. And so we’re going to talk about exports, but all of this comes from w…
S68
Conversation: 01 — Krishnan outlined the Trump administration’s three-pillar strategy developed over 13 months. The first pillar focuses on…
Speakers Analysis
Detailed breakdown of each speaker’s arguments and positions
B
Brad Smith
17 arguments143 words per minute1863 words779 seconds
Argument 1
AI as a pivotal factor that can either close or widen the economic gap (Brad Smith)
EXPLANATION
Brad Smith argues that artificial intelligence has the power to either narrow or broaden the existing economic disparity between the Global North and South. He positions AI as the most consequential technology of this century for shaping that outcome.
EVIDENCE
He explains that the economic divide is rooted in a technology divide and states that AI, perhaps more than any other technology this century, will play a bigger role either in closing the economic divide or in exacerbating it [10-12].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Smith emphasizes that AI can either close or widen the economic divide, framing it as the most consequential technology of the century [S7] and [S1].
MAJOR DISCUSSION POINT
Economic and technology divide between Global North and South
Argument 2
The technology divide stems largely from unequal access to electricity, a historic precedent for current AI disparities (Brad Smith)
EXPLANATION
Smith highlights that unequal access to electricity historically created a technology divide, which now mirrors the disparities in AI adoption. He uses the spread of electricity as an analogy for today’s AI gap.
EVIDENCE
He describes electricity as a general-purpose technology that boosted productivity where it reached, notes that the first power plant opened 144 years ago while 700 million people still lack electricity today, and links this historic inequality to the present AI divide [10-11].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
He draws a parallel between today’s AI gap and the historic technology divide created by unequal electricity access, noting that 700 million people still lack electricity [S1] and [S7].
MAJOR DISCUSSION POINT
Economic and technology divide between Global North and South
Argument 3
Massive investment in data centers, compute, connectivity, and electricity is essential (Brad Smith)
EXPLANATION
Smith states that building the physical infrastructure—data centers, computing power, network connectivity, and reliable electricity—is a prerequisite for AI deployment in the Global South. He calls this the first step that requires massive investment.
EVIDENCE
He lists the need for data centers, compute, more connectivity, and more electricity as essential infrastructure, emphasizing that delivering this will require “the world’s best technology” and huge investment [17-21].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Smith lists data centers, compute, connectivity and reliable electricity as prerequisite infrastructure, calling for massive investment [S1] and [S7].
MAJOR DISCUSSION POINT
Infrastructure investment to enable AI in the Global South
Argument 4
Microsoft’s pledge to spend $50 billion by year‑end to bring AI to the Global South (Brad Smith)
EXPLANATION
Smith announces Microsoft’s commitment to allocate $50 billion before the end of the year to support AI initiatives in the Global South. This pledge is presented as a concrete financial commitment to the infrastructure agenda.
EVIDENCE
He says, “That’s why we at Microsoft announced yesterday morning that we’re on pace to spend $50 billion by the end of this year… to bring AI to the global South” [22-24].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
He announces Microsoft’s commitment to spend $50 billion by year-end on AI infrastructure in the Global South [S1] and [S9].
MAJOR DISCUSSION POINT
Infrastructure investment to enable AI in the Global South
Argument 5
Collaboration among private capital, tech firms, and governments is required to generate demand and scale AI (Brad Smith)
EXPLANATION
Smith argues that private investment, corporate resources, and government funding must be coordinated to create market demand for AI in the Global South. He stresses that only a joint effort can spin the market wheels needed for scale.
EVIDENCE
He mentions the need to harness private capital, investments from tech companies, other private sources, and government funding, and that governments must generate demand to get the market spinning [26-28].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Smith stresses that private capital, tech firms and government funding must be coordinated to generate demand for AI in the Global South [S1] and [S7].
MAJOR DISCUSSION POINT
Infrastructure investment to enable AI in the Global South
Argument 6
Providing AI‑related skills to people across societies is as important as hardware infrastructure (Brad Smith)
EXPLANATION
Smith emphasizes that skill development is equally critical to physical infrastructure for AI adoption. He argues that without a skilled workforce, the hardware alone cannot deliver impact.
EVIDENCE
He notes that infrastructure is not only wires and grids but also “skilling for people,” and that giving people access to the skills they need is key to using a general-purpose technology at scale [31-35].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
He highlights that skilling people is as critical as hardware, describing it as “skilling for people” and part of foundational infrastructure [S10] and [S1].
MAJOR DISCUSSION POINT
Skills development and education as core to AI adoption
Argument 7
Launch of Microsoft Elevate for Educators to equip teachers and students with AI tools (Brad Smith)
EXPLANATION
Smith announces a new initiative, Microsoft Elevate for Educators, aimed at providing teachers with resources to teach AI to their students. This program is presented as part of Microsoft’s broader skilling effort.
EVIDENCE
He says Microsoft has launched new initiatives through Microsoft Elevate, including “Microsoft Elevate for Educators, to equip teachers with access to help their students learn how to use AI” [38-39].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Smith unveils Microsoft Elevate for Educators, a program to give teachers AI tools for students [S1] and [S11].
MAJOR DISCUSSION POINT
Skills development and education as core to AI adoption
Argument 8
Employers must open doors to AI tools and invest in upskilling their employees across generations (Brad Smith)
EXPLANATION
Smith calls on employers to provide access to AI tools and to invest in continuous upskilling of their workforce, noting that this responsibility spans all generations, not just new entrants.
EVIDENCE
He explains that it took employers to open doors to computing and now they must open doors to new AI tools, investing in the skilling of employees for every generation that matters [41-45].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
He calls on employers to open doors to AI tools and continuously upskill workers across generations, likening it to the opening of computing doors [S1].
MAJOR DISCUSSION POINT
Skills development and education as core to AI adoption
Argument 9
AI must perform equally well in all languages, not just English (Brad Smith)
EXPLANATION
Smith points out that current AI systems perform best in English and stresses the need for parity across languages to ensure inclusive benefits. He frames linguistic equity as essential for global AI relevance.
EVIDENCE
He states that AI is not as effective in every language as it is in English and that performance tests confirm this gap [48-49].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Smith points out that current AI systems perform best in English and calls for parity across languages [S1].
MAJOR DISCUSSION POINT
Making AI effective and relevant for the Global South
Argument 10
Investment in multilingual data, tools, and measurement systems to ensure linguistic diversity (Brad Smith)
EXPLANATION
Smith outlines a plan to invest upstream in multilingual data, develop tools and measurement systems, and build data provenance that supports linguistic diversity. This is presented as a concrete step to close the language gap.
EVIDENCE
He describes new announcements to invest in better data in other languages, provide better tools and measurement systems for AI built in other languages, and build out data provenance for linguistic diversity [50].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
He outlines investments in multilingual data, tools, measurement systems and data provenance to support linguistic diversity [S1].
MAJOR DISCUSSION POINT
Making AI effective and relevant for the Global South
Argument 11
Deploy AI to solve region‑specific challenges such as agriculture improvement and food security in Africa (Brad Smith)
EXPLANATION
Smith gives examples of AI applications tailored to local needs, such as improving agriculture in India and addressing food security across Africa. He suggests these use‑cases illustrate AI’s potential for tangible impact.
EVIDENCE
He cites work being done in India on agricultural improvements and a partnership launching a new initiative to address food security across Africa as examples of AI solving real-world problems [55-57].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Smith cites AI projects improving agriculture in India and addressing food security across Africa as examples of region-specific impact [S1].
MAJOR DISCUSSION POINT
Making AI effective and relevant for the Global South
Argument 12
Growing public concern about how AI will affect children, families, and overall livelihoods (Brad Smith)
EXPLANATION
Smith notes that outside the conference hall, many parents are asking what AI means for their children, families, and future, indicating widespread societal anxiety about AI’s impact on livelihoods.
EVIDENCE
He mentions that many parents are asking “What will AI mean for my kids? What will AI mean for my family? What will AI mean for our future?” reflecting growing public concern [62-66].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
He notes growing public anxiety, with parents asking what AI means for their children and families [S1].
MAJOR DISCUSSION POINT
Future of work, jobs, and societal impact of AI
Argument 13
AI should be viewed as a catalyst for human curiosity and capability, not a replacement for people (Brad Smith)
EXPLANATION
Smith frames AI as a tool that amplifies human curiosity and capability rather than displacing humans. He argues that technology creates new platforms that enable people to achieve more, provided it is used responsibly.
EVIDENCE
He states that human capability is not fixed, that AI can cure diseases, find faster solutions, and that curiosity remains the fundamental fuel, positioning AI as a generator for human curiosity [71-79].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Smith frames AI as a catalyst that amplifies human curiosity and capability rather than replacing people [S1].
MAJOR DISCUSSION POINT
Future of work, jobs, and societal impact of AI
Argument 14
Historical analogy of the washing machine illustrates how technology can free time and improve quality of life when paired with skill development (Brad Smith)
EXPLANATION
Smith uses the washing‑machine example to show how a technology can dramatically reduce labor time, leading to better living standards and more time for productive activities, provided people acquire the skills to leverage it.
EVIDENCE
He recounts that before the washing machine, laundry took six to eight hours, which dropped to 30 minutes after the invention, leading people to wash more often, have cleaner clothes, and use the saved time for other pursuits [81-89].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
He uses the washing-machine analogy to illustrate how technology can free time and raise living standards when paired with skill development [S1].
MAJOR DISCUSSION POINT
Future of work, jobs, and societal impact of AI
Argument 15
AI summits must be linked, building bridges rather than operating as isolated events (Brad Smith)
EXPLANATION
Smith calls for AI summits to be interconnected, creating continuity rather than operating as stand‑alone events. He emphasizes the need for “building bridges” between summits.
EVIDENCE
He says “Rather than have summits that are islands… we need to build bridges… between these summits” [90-92].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Smith argues AI summits should be interconnected, building bridges rather than being isolated islands [S1] and [S7].
MAJOR DISCUSSION POINT
Need for continuous, measurable progress across AI summits
Argument 16
Establish clear goals, common measurement systems, and annual accountability to track progress (Brad Smith)
EXPLANATION
Smith proposes setting explicit objectives, developing shared metrics, and reviewing progress each year to ensure accountability for AI initiatives. He frames this as essential for sustained advancement.
EVIDENCE
He outlines the need for clear goals, common measurement systems, and a yearly question on whether progress was made, urging annual accountability [93-100].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
He proposes clear goals, common metrics and annual accountability to track progress on AI initiatives [S1] and [S7].
MAJOR DISCUSSION POINT
Need for continuous, measurable progress across AI summits
Argument 17
Collective global responsibility of governments, companies, NGOs, and employers to ensure AI benefits people worldwide (Brad Smith)
EXPLANATION
Smith stresses that a broad coalition—including governments, corporations, NGOs, and employers—must work together to ensure AI delivers benefits globally and that the world holds itself accountable.
EVIDENCE
He calls for the global community to define goals, measure progress, and hold ourselves accountable, noting that people outside the walls are expecting us to deliver [100-104].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Smith calls for a collective global responsibility involving governments, companies, NGOs and employers to ensure AI benefits all [S1] and [S7].
MAJOR DISCUSSION POINT
Need for continuous, measurable progress across AI summits
Agreements
Agreement Points
Technology companies have a responsibility to address societal challenges and shape the future of AI responsibly
Speakers: Speaker 1, Brad Smith
Speaker 1 highlights Brad Smith as Microsoft’s conscience and chief diplomat, emphasizing corporate responsibility Brad Smith stresses the need to do better and to prove that tech can pursue a brighter future for people
Both speakers acknowledge that technology firms must act responsibly in guiding AI development and its societal impact [2][14-15]
POLICY CONTEXT (KNOWLEDGE BASE)
This view echoes calls for corporate responsibility in AI governance and co-accountability frameworks discussed in policy panels, where private sector partners with governments and civil society to address societal challenges [S22][S39].
AI can either close or widen the economic divide between the Global North and South
Speakers: Brad Smith
AI as a pivotal factor that can either close or widen the economic gap (Brad Smith) The technology divide stems largely from unequal access to electricity, a historic precedent for current AI disparities (Brad Smith)
Brad repeatedly frames AI as the key technology that will determine whether the existing economic gap narrows or expands, drawing a parallel with the historic electricity divide [10-12]
POLICY CONTEXT (KNOWLEDGE BASE)
Analyses highlight AI’s potential to both narrow and widen the North-South economic gap, noting the risk of a new digital divide driven by uneven generative AI adoption [S30][S31].
Massive investment in infrastructure (data centers, compute, connectivity, electricity) is essential for AI deployment in the Global South
Speakers: Brad Smith
Massive investment in data centers, compute, connectivity, and electricity is essential (Brad Smith) Microsoft’s pledge to spend $50 billion by year‑end to bring AI to the Global South (Brad Smith)
Brad outlines the need for physical infrastructure and backs it with a $50 billion investment commitment [17-21][22-24]
POLICY CONTEXT (KNOWLEDGE BASE)
Panel discussions stress that AI policies must evolve alongside physical infrastructure, emphasizing the need for data centers, compute and reliable electricity in the Global South [S22][S42][S38].
Skill development for people is as critical as hardware infrastructure
Speakers: Brad Smith
Providing AI‑related skills to people across societies is as important as hardware infrastructure (Brad Smith) Launch of Microsoft Elevate for Educators to equip teachers and students with AI tools (Brad Smith) Employers must open doors to AI tools and invest in upskilling across generations (Brad Smith)
Brad stresses that skilling is a core component of AI adoption, announcing Microsoft Elevate for Educators and urging employer-led upskilling [31-35][38-45]
POLICY CONTEXT (KNOWLEDGE BASE)
Multiple sources underline that skills training and education are as vital as hardware, with initiatives targeting marginalized groups and youth to build AI competencies [S28][S40][S41][S43].
AI must work effectively in all languages, not just English
Speakers: Brad Smith
AI must perform equally well in all languages, not just English (Brad Smith) Investment in multilingual data, tools, and measurement systems to ensure linguistic diversity (Brad Smith)
Brad points out the current English bias in AI performance and announces investments in multilingual data and tools to address it [48-50]
POLICY CONTEXT (KNOWLEDGE BASE)
Research points out current AI systems underperform in non-English languages and calls for multilingual development, exemplified by the Current AI-Bhashini partnership announced at the summit [S27][S29][S21].
AI should be applied to solve region‑specific challenges such as agriculture and food security
Speakers: Brad Smith
Deploy AI to solve region‑specific challenges such as agriculture improvement and food security in Africa (Brad Smith)
Brad cites examples of AI projects in Indian agriculture and African food-security initiatives as concrete use-cases [55-57]
POLICY CONTEXT (KNOWLEDGE BASE)
Several sessions presented AI applications for agriculture and food security, including AI-driven crop prediction and island-scale pilots, underscoring the need for region-specific solutions [S35][S36][S37].
AI summits need continuity, clear goals, common metrics, and annual accountability
Speakers: Brad Smith
AI summits must be linked, building bridges rather than operating as isolated events (Brad Smith) Establish clear goals, common measurement systems, and annual accountability to track progress (Brad Smith)
Brad calls for interconnected summits, shared objectives, and yearly progress reviews to ensure measurable outcomes [90-95][96-100]
POLICY CONTEXT (KNOWLEDGE BASE)
Summit organizers stress continuity and measurable outcomes, recommending annual progress metrics and clear goal-setting to ensure accountability across meetings [S19][S20][S21].
Collective global responsibility of governments, companies, NGOs, and employers to ensure AI benefits people worldwide
Speakers: Brad Smith
Collective global responsibility of governments, companies, NGOs, and employers to ensure AI benefits people worldwide (Brad Smith)
Brad emphasizes a broad coalition-including governments, private sector, NGOs, and employers-to deliver AI benefits and be held accountable [100-104]
POLICY CONTEXT (KNOWLEDGE BASE)
Discussions repeatedly invoke collective responsibility and co-accountability among governments, firms, NGOs and employers to guarantee equitable AI benefits worldwide [S22][S32][S33][S39].
Similar Viewpoints
Infrastructure investment is the foundational step for AI deployment in the Global South, reinforced by a concrete $50 billion commitment [17-21][22-24]
Speakers: Brad Smith
Massive investment in data centers, compute, connectivity, and electricity is essential (Brad Smith) Microsoft’s pledge to spend $50 billion by year‑end to bring AI to the Global South (Brad Smith)
Skill development, education, and employer‑led upskilling are essential complements to hardware for effective AI adoption [31-35][38-45]
Speakers: Brad Smith
Providing AI‑related skills to people across societies is as important as hardware infrastructure (Brad Smith) Launch of Microsoft Elevate for Educators to equip teachers and students with AI tools (Brad Smith) Employers must open doors to AI tools and invest in upskilling across generations (Brad Smith)
Linguistic equity requires upstream investment in multilingual data and tools to make AI effective beyond English [48-50]
Speakers: Brad Smith
AI must perform equally well in all languages, not just English (Brad Smith) Investment in multilingual data, tools, and measurement systems to ensure linguistic diversity (Brad Smith)
Continuity, shared metrics, and yearly review are necessary for sustained progress across AI summits [90-95][96-100]
Speakers: Brad Smith
AI summits must be linked, building bridges rather than operating as isolated events (Brad Smith) Establish clear goals, common measurement systems, and annual accountability to track progress (Brad Smith)
Unexpected Consensus
Multi‑stakeholder collaboration (private sector, governments, NGOs) as essential for AI impact
Speakers: Speaker 1, Brad Smith
Speaker 1 describes Brad Smith as Microsoft’s chief diplomat, implying a role in cross‑sector dialogue Brad Smith calls for harnessing private capital, tech firms, and government funding to generate demand and scale AI (Brad Smith)
While Speaker 1’s introduction focuses on Brad’s diplomatic stature, it unexpectedly aligns with Brad’s explicit call for coordinated private-public-NGO effort, showing early consensus on the need for broad collaboration [2][26-28]
POLICY CONTEXT (KNOWLEDGE BASE)
Policy literature and workshop reports consistently advocate multi-stakeholder collaboration as essential for inclusive AI governance and risk management [S24][S25][S26][S33][S34].
Overall Assessment

The discussion shows strong internal coherence in Brad Smith’s vision: AI can shape the North‑South economic gap, but only through massive infrastructure investment, widespread skill development, linguistic inclusivity, region‑specific applications, and a coordinated multi‑stakeholder approach, all measured by clear metrics and annual review. The only cross‑speaker agreement is the shared view that technology firms bear responsibility and must collaborate with governments and other actors.

High consensus on the core strategic pillars within Brad Smith’s remarks; limited multi‑speaker consensus (only between Speaker 1 and Brad Smith on corporate responsibility and collaboration). The consensus suggests a unified roadmap for AI deployment in the Global South, but broader stakeholder buy‑in beyond Microsoft remains to be demonstrated.

Differences
Different Viewpoints
Unexpected Differences
Overall Assessment

The transcript contains only an introductory remark by Speaker 1 and a single, uninterrupted presentation by Brad Smith. No opposing speaker offers a contrasting viewpoint, so there are no explicit disagreements or partial agreements identified in the discussion.

Minimal – the dialogue is essentially a monologue, implying consensus rather than conflict, which suggests that the primary challenges discussed (infrastructure, skills, multilingual AI, and governance) are presented without internal contestation.

Takeaways
Key takeaways
The economic gap between the Global North and South is fundamentally a technology gap, historically rooted in unequal access to electricity and now mirrored in AI access. AI has the potential to either narrow or widen this gap; intentional action is required to ensure it becomes a closing force. Infrastructure—data centers, compute power, connectivity, and reliable electricity—is a prerequisite for AI adoption in the Global South. Microsoft has committed to invest $50 billion by year‑end to build the necessary AI infrastructure and to partner with governments, private capital, and NGOs to generate demand. Skills development is as critical as hardware; initiatives like Microsoft Elevate for Educators aim to equip teachers, students, and workers with AI competencies across generations. AI must be linguistically inclusive; investments are needed in multilingual data, tools, and measurement systems to ensure AI works effectively in non‑English languages. Deploying AI to solve region‑specific challenges—such as agricultural productivity and food security in Africa—is essential for tangible benefits. Public concern about AI’s impact on jobs, families, and future generations must be addressed by positioning AI as a catalyst for human curiosity and capability, not a replacement. Future AI summits should be interconnected, with clear goals, common metrics, and annual accountability to track progress and avoid isolated “island” events.
Resolutions and action items
Microsoft will allocate $50 billion by the end of the year toward AI infrastructure (data centers, compute, connectivity, electricity) in the Global South. Launch and expand Microsoft Elevate for Educators to provide AI tools and training for teachers and students. Invest upstream in multilingual data, tools, and measurement systems to improve AI performance across languages. Collaborate with governments, private investors, and NGOs to create demand for AI solutions in the Global South. Initiate AI projects targeting agriculture improvement and food‑security challenges in Africa and other regions. Encourage employers to open access to AI tools and fund upskilling programs for employees of all generations. Develop common measurement frameworks and set annual progress targets for AI summits, ensuring continuity and accountability.
Unresolved issues
Specific mechanisms for mobilizing the required private‑sector capital and aligning government funding with AI infrastructure projects. Detailed plans for expanding reliable electricity access to the 700 million people currently without it. Timeline and concrete milestones for achieving parity of AI performance in non‑English languages. Exact metrics and governance structures for the proposed common measurement system across AI summits. How to systematically address widespread public anxiety about AI’s impact on employment and societal well‑being.
Suggested compromises
A shared‑responsibility model where governments, tech companies, NGOs, and employers jointly fund and implement AI infrastructure and skill‑building programs. Balancing investment between hardware infrastructure and human‑capital development to ensure technology adoption is effective and inclusive. Linking AI summits together rather than treating each as an isolated event, creating “bridges” that allow for cumulative progress tracking and coordinated goals.
Thought Provoking Comments
The deepest and most enduring divide has been the economic divide between the global north and south, and this economic divide is a result, more than anything else, of a technology divide.
Links the long‑standing economic disparity directly to unequal access to foundational technologies, reframing the problem from a purely economic issue to a technological one.
Sets the analytical framework for the entire speech, prompting listeners to view AI as a lever that can either widen or narrow that divide and steering the conversation toward infrastructure and access solutions.
Speaker: Brad Smith
AI, perhaps more than any other technology this century, will play a bigger role either in closing this economic divide or in exacerbating it – that is the single most important question for us.
Poses a stark, binary framing that forces stakeholders to consider both optimistic and pessimistic outcomes, highlighting the urgency of responsible AI deployment.
Creates a turning point that shifts the tone from descriptive to prescriptive, leading to discussions about concrete actions (investment, skilling, language support) needed to ensure the positive outcome.
Speaker: Brad Smith
Microsoft is on pace to spend $50 billion by the end of this year to bring AI to the global South, including data centers, compute, connectivity and electricity.
Introduces a concrete, large‑scale financial commitment, moving the conversation from abstract challenges to tangible resources and timelines.
Triggers a focus on the feasibility of private‑sector investment, invites questions about partnership models, and underscores the scale of effort required.
Speaker: Brad Smith
Infrastructure is not only hardware; it’s also skilling for people. We must give people across the country access to the skills they need to put technology to work.
Expands the definition of infrastructure to include human capital, emphasizing that technology alone cannot drive development without education and training.
Broadens the discussion to include education initiatives (e.g., Microsoft Elevate for Educators), prompting consideration of long‑term capacity building alongside physical assets.
Speaker: Brad Smith
We need to make AI as effective in every language as it is in English; performance tests show it is not today.
Highlights linguistic bias as a critical barrier to equitable AI adoption, introducing the often‑overlooked dimension of language diversity.
Leads to a new sub‑topic about data provenance, multilingual datasets, and measurement systems, and signals future investment priorities beyond hardware.
Speaker: Brad Smith
AI must be used to solve the problems that matter to the Global South—like improving agriculture in India or addressing food security across Africa.
Shifts the narrative from generic AI benefits to concrete, region‑specific use cases, grounding the discussion in real‑world impact.
Encourages participants to think about sector‑specific pilots, partnerships with local stakeholders, and outcome‑based metrics, deepening the conversation about application versus theory.
Speaker: Brad Smith
What will AI mean for my kids, my family, my future? Parents are asking this question worldwide.
Moves the focus from macro‑economic and technical considerations to personal, human concerns, challenging the audience to consider societal and intergenerational effects.
Changes the tone from corporate optimism to a more empathetic, people‑centered dialogue, prompting reflections on job displacement, education, and ethical responsibilities.
Speaker: Brad Smith
We need to build bridges between AI summits, define clear goals, common measurement systems, and assess progress each year.
Calls for systematic continuity and accountability across global AI initiatives, introducing governance and measurement as essential components of progress.
Sets a forward‑looking agenda, encouraging the audience to think about institutional frameworks, longitudinal tracking, and collaborative standards rather than isolated events.
Speaker: Brad Smith
Overall Assessment

Brad Smith’s remarks repeatedly reframed the AI conversation—from viewing the north‑south divide as a technology issue, to emphasizing massive private investment, to expanding infrastructure to include skills and linguistic inclusivity, and finally to grounding AI in everyday human concerns and institutional accountability. Each of these pivot points introduced new dimensions, shifted the tone from abstract optimism to concrete responsibility, and deepened the dialogue by linking technical solutions to societal outcomes. Collectively, the comments steered the discussion toward a holistic, action‑oriented roadmap for AI in the Global South, shaping the summit’s narrative around equity, capacity building, real‑world impact, and sustained, measurable progress.

Follow-up Questions
How can we do better in deploying AI to the Global South?
Seeks actionable strategies to bridge the technology and economic divide between North and South.
Speaker: Brad Smith
What will it take to bring AI infrastructure, compute, connectivity, and electricity to the Global South?
Identifies the scale of investment, private capital, and government involvement required for effective AI rollout.
Speaker: Brad Smith
What will AI mean for the future of work and jobs?
Addresses a major societal concern about AI’s impact on employment and economic opportunities worldwide.
Speaker: Brad Smith
What will AI mean for my kids, my family, and our future?
Reflects public anxiety and the need to communicate AI’s long‑term implications for everyday life.
Speaker: Brad Smith
How can we make AI as effective in every language as it is in English?
Highlights the necessity of linguistic equity for inclusive AI adoption across diverse populations.
Speaker: Brad Smith
What better data and measurement systems are needed for AI in non‑English languages?
Calls for research into data provenance, multilingual datasets, and evaluation metrics to improve model performance.
Speaker: Brad Smith
How can AI be applied to solve real‑world problems in the Global South such as agriculture and food security?
Points to the need for sector‑specific studies to quantify AI’s impact on critical challenges like farming and nutrition.
Speaker: Brad Smith
How can we establish common measurement systems and clear goals across AI summits to track yearly progress?
Proposes a framework for accountability and continuous improvement in global AI policy and implementation.
Speaker: Brad Smith
What is the potential of AI to improve human health and accelerate disease cures?
Identifies a research avenue to evaluate AI’s contributions to medical breakthroughs and public health outcomes.
Speaker: Brad Smith

Disclaimer: This is not an official session record. DiploAI generates these resources from audiovisual recordings, and they are presented as-is, including potential errors. Due to logistical challenges, such as discrepancies in audio/video or transcripts, names may be misspelled. We strive for accuracy to the best of our ability.

Keynote-Sam Altman

Session at a glanceSummary, keypoints, and speakers overview

Summary

The event opened with Speaker 1 highlighting Sam Altman’s role in moving artificial general intelligence from science-fiction into mainstream institutions and introducing him as CEO of OpenAI [2-4]. Altman thanked the audience and noted the significance of speaking in India, praising the country’s rapid advancement in advanced AI [5-7]. He described how AI systems have progressed from struggling with high-school-level mathematics to performing research-level mathematics and generating novel theoretical-physics results [10]. Highlighting India’s impact, he said more than 100 million Indians use ChatGPT weekly, a third of whom are students, and that India is the fastest-growing market for OpenAI’s Codex coding assistant [13-15]. Altman argued that India’s democratic framework positions it to not only build AI but also shape its future direction [16]. He warned that, on the current trajectory, early versions of true superintelligence could appear within a few years, and by the end of 2028 a majority of the world’s intellectual capacity might reside in data centres [19-21]. Such superintelligence, he suggested, could outperform human CEOs and scientists, underscoring the need for careful preparation [23]. OpenAI’s strategy rests on three core beliefs: first, that democratizing AI is the only safe and fair path, while centralising power risks ruin [25-28]; second, that AI resilience-broad societal safety measures beyond technical alignment-is essential [33-41]; and third, that many stakeholders must help shape AI’s uncertain future [42-45]. He emphasized iterative deployment, allowing society to integrate each new capability before moving on, a practice he says has worked well so far [53-55]. Altman projected that AI will dramatically lower costs and spur economic growth, especially through automation of manufacturing and supply chains, but also warned that many current jobs will be disrupted as machines out-perform human labor in many tasks [56-61]. He noted that humans remain more motivated by caring for other people than by machines, suggesting a cultural shift rather than total displacement [63-65]. To safeguard a democratic AI future, Altman called for both decentralized agency and strong regulatory frameworks, proposing an international body akin to the IAEA to coordinate rapid responses to AI risks [76-78]. He concluded that the coming years will test global society, and the world can choose either to empower people with AI or to concentrate power in a few hands [78].


Keypoints


Major discussion points


Rapid advances toward super-intelligent AI – Altman notes that AI has moved from “high-school level math” to “research-level mathematics” and predicts “early versions of true superintelligence” within a few years, possibly by the end of 2028 when “more of the world’s intellectual capacity could reside inside of data centers” [19-21].


Democratization and shared governance of AI – He argues that “democratization of AI is the only fair and safe path forward” and warns that “centralization… could lead to ruin.” He stresses the need for “society-wide debate,” “stake-holding,” and an international coordination body “like the IAEA for AI” to prevent concentration of power and to protect democratic values [25-33][76-78].


Safety through resilience and iterative deployment – Altman outlines a three-part safety strategy: (1) broaden safety to include societal resilience, (2) continue technical alignment work, and (3) use “iterative deployment” so each new capability can be integrated, understood, and regulated before the next leap [33-41][53-55].


Economic transformation and job disruption – He highlights that AI will make many goods and services cheaper, boost growth, and improve access to health-care and education, but also acknowledges that “current jobs are going to get disrupted” as AI “can do more and more of the things that drive our economy today” [55-61][62-68].


India’s growing AI ecosystem – Altman points to India’s massive adoption (“more than 100 million people… every week,” with a third being students) and its leadership in “sovereign AI, building on infrastructure, SLMs,” as well as the rapid growth of the Codex coding agent market there [11-16].


Overall purpose / goal of the discussion


The talk is intended to inform and inspire the Indian audience about the unprecedented speed of AI progress, to underline the strategic importance of democratic, inclusive governance, and to call for proactive, society-wide preparation-both in policy and in economic adaptation-as AI reshapes global capability and power structures.


Overall tone


The tone begins with enthusiasm and pride (“it’s incredible to see the country’s leadership in advanced AI”) and quickly shifts to a sober, cautionary stance when addressing superintelligence and governance risks. Throughout, Altman balances optimism about AI’s benefits (economic growth, better health and education) with a serious, responsible call to action, ending on a hopeful yet urgent note urging empowerment over concentration of power. The progression moves from celebratory to reflective to rally-cry.


Speakers

Sam Altman – Role/Title: CEO of OpenAI [S1] – Area of Expertise: Artificial intelligence, artificial general intelligence development, technology leadership [S1]


Speaker 1 – Role/Title: Event moderator / host (introducing the main speaker) [S4] – Area of Expertise: (not specified)


Additional speakers:


Full session reportComprehensive analysis and detailed insights

The moderator opened the event by praising Sam Altman’s pivotal role in moving artificial general intelligence from science-fiction into boardrooms, parliaments and living rooms, and formally introduced him as the CEO of OpenAI [2-4].


Altman thanked the audience and expressed his enthusiasm for speaking in India, noting the country’s rapid ascent in advanced AI and recalling that his previous visit was just over a year earlier [5-8]. He framed the talk around the extraordinary progress since that visit, highlighting the leap from AI systems that could barely solve high-school-level mathematics to models now capable of research-level mathematics and even deriving novel results in theoretical physics [9-10].


Turning to India’s ecosystem, Altman pointed out that more than 100 million Indians use ChatGPT each week, with roughly one-third of those users being students [13-14]. He added that India is the fastest-growing market for Codex, OpenAI’s coding assistant, and praised the nation’s leadership in sovereign AI, infrastructure and large-scale models [15-16].


Discussing the future trajectory of AI, Altman warned that, on the current path, early versions of true superintelligence could appear within a couple of years [17-19]. He projected that by the end of 2028 a majority of the world’s intellectual capacity might reside inside data centres rather than outside them [20-21]. If this forecast proves accurate, superintelligent systems could outperform human CEOs and conduct research beyond the capabilities of today’s top scientists [22-24].


Altman then outlined OpenAI’s three core beliefs that guide its approach to this emerging landscape.


Democratization of AI – the company holds that spreading AI widely is the only fair and safe path forward; concentrating the technology in a single company or nation could undermine liberty, democracy and human flourishing [25-28].


AI resilience & societal safety – beyond technical alignment, society must develop broad-based mechanisms to defend against threats such as open-source biomodels that could be weaponised [33-41]. Altman noted that while many tasks will become harder for humans to outwork a GPU, other tasks will remain easier for people than machines [45-46].


Broad stakeholder involvement & humility – many parties must have a seat at the table because the future of AI will not unfold exactly as anyone predicts; humility about unknowns and a commitment to wide-scale debate are essential [42-48].


A practical expression of the resilience principle is the policy of iterative deployment. Altman argued that releasing each new capability in stages gives societies time to integrate, understand and decide how to move forward with each successive level of AI [53-55].


On the economic front, Altman predicts that AI will dramatically lower costs and accelerate growth, improving access to high-quality healthcare, education and other services [56-58]. He foresees robots and automated supply chains making physical goods cheaper, while noting that the ultimate limit to cost reduction may be set by government policy [59-60]. At the same time, he acknowledges that many current jobs will be disrupted as AI increasingly performs tasks that drive today’s economy [61-62].


Despite the disruption, Altman maintains a hopeful view of human purpose. He observes that people are hard-wired to care more about other people than machines, suggesting that future work will be reshaped rather than eliminated [63-65]. He likens today’s occupations to “silly” past jobs and envisions future generations finding richer, more fulfilling ways to express creativity, compete and collaborate [66-70][71-74]. He frames this as a moral imperative: protecting the collective external lattice-the set of tools we have built around ourselves-for the benefit of great-great-grandchildren [68-70].


To safeguard a democratic AI future, Altman calls for both decentralized agency and robust regulation. He proposes the creation of an international coordination body-akin to the International Atomic Energy Agency-to oversee AI safety and enable rapid responses to emerging risks [76-78]. He concludes that the coming years will test global society, and humanity can choose either to empower people with AI or to allow power to concentrate in the hands of a few [78].


The moderator closed the session by thanking Altman for his compelling remarks [79].


Session transcriptComplete transcript of the session
Speaker 1

level to change the lives of human beings. Ladies and gentlemen, few individuals have done more to bring artificial general intelligence from the realm of science fiction into boardrooms, into parliaments and living rooms than our next speaker, Sam Altman, CEO, OpenAI. Under his leadership, OpenAI launched ChatGPT and forced the world to re -evaluate his relationship with artificial intelligence. So ladies and gentlemen, please welcome CEO of OpenAI, Mr. Sam Altman.

Sam Altman

Thank you so much. It’s really a treat to be here in India, and it’s incredible to see the country’s leadership in advanced AI. I was last here a little over a year ago, and it’s striking that I’m here today. I’m here to talk to you about the future of AI. I’m here to talk to you about how much progress has happened since then. We’ve gone from AI systems that struggled with high school level math to systems that can do research level mathematics now and derive novel results in theoretical physics. It’s also striking how much progress India has made in its mission to put AI to work for more people in more parts of the country.

And India’s leadership in sovereign AI, building on infrastructure, SLMs, and much more has been great to watch. More than 100 million people in India use ChatGPT every week. More than a third of them are students. India is also the fastest growing market now for Codex, our coding agent that works to help people develop software faster and better. India, the world’s largest democracy, is well positioned to lead in AI, not just to build it, but to shape it and decide what our future is going to look like. And it’s important to move quickly. On our current trajectory. We believe we may be only a couple of years away from early versions of true superintelligence. If we are right, by the end of 2028, more of the world’s intellectual capacity could reside inside of data centers than outside of them.

This is an extraordinary statement to make, and of course we could be wrong. But I think it really bears serious consideration. A superintelligence, at some point on its development curve, would be capable of doing a better job being the CEO of a major company than any executive, certainly me, or doing better research than our best scientists. As we prepare for this possibility, we are guided by three core beliefs. Number one, we believe that democratization of AI is the only fair and safe path forward. Democratization of AI is the best way to ensure that humanity flourishes. On the other hand, centralization of this technology in one company or country could lead to ruin. The desirable future a couple of decades from now has got to look like a world of liberty, democracy, widespread flourishing, and an increase in human agency.

Some people want effective totalitarianism in exchange for a cure for cancer. I don’t think we should accept that trade -off, nor do I think we need to. AI should extend individual human will. We’ll probably need superintelligence to help us figure out the new governance mechanisms to ensure that this happens fairly at scale, and to avoid problems like extremely unbalanced compute, access, or something else. Second, we believe that AI resilience is a core safety strategy. We don’t mean that this is the only safety strategy. We will continue to need to build safe systems and solve difficult technical alignment challenges. But increasingly, we need to start broadening how we think about safety to include societal resilience. No AI lives in a world where we don’t have to worry about safety.

We need to build a system where we can do that. No AI system can deliver a good future on their own. For an obvious example, there’ll be extremely capable biomodels available open source that could help people create new pathogens. We need a society -wide approach about how we’re going to defend against this. And third, the future of AI is not going to unfold exactly like anyone predicts. And we believe that many people need to have a stake in shaping the outcome. The development of AI has already held many surprises, and I assume there are bigger ones to come. We understand that with technology this powerful, people want answers. But it’s important to be humble about what we don’t know, and always remember that sometimes our best guesses are wrong.

Most of the important discoveries happen when technology and society meet, sometimes have some friction, and co -evolve. For example, we don’t yet know how to think about some superhuman problem. We don’t know how to think about superintelligence being aligned with dictators in totalitarian countries. We don’t know how to think about countries using AI to fight new kinds of war with each other. We don’t know how to think about when and whether countries are going to have to think about new forms of social contracts. But we think it’s important to have more understanding and society -wide debate before we’re all surprised. Of special note, and related to all three points, we continue to believe that iterative deployment is a key strategic insight, and that society needs to contend with and use each successive new level of AI capability, have time to integrate it, understand it, and decide how to move forward.

This has been working surprisingly well so far. If we are right, and systems continue to improve at this pace, it’s going to change the economics of a lot of things. A really great thing about AI progress is that it looks like many things are going to get much cheaper and have much faster economic growth. We’re already seeing what AI is doing, for access to high -quality healthcare, education, and more. In the coming years, we expect to see robots make many products and physical goods cheaper as supply chains get automated. The limit to how far this cost reduction can go may only be government policy. But the other side of this coin is that current jobs are going to get disrupted, as AI can do more and more of the things that drive our economy today.

It’ll be very hard to outwork a GPU in many ways. It’ll be easy in some other ways. For example, we really seem hardwired to care about other people much more than we care about machines. We’re somewhat less concerned about the long -term future. Technology always disrupts jobs. We always find new and better things to do. The people of 500 years ago would have thought that our current jobs often look silly, like ways to entertain ourselves, create stress. And the people 500 years from now hopefully will look at us, hopefully look to us, like impossibly rich people playing games, trying to find ways to pass their times. But we should all hope that they feel much more fulfilled.

We should all hope that they feel much more fulfilled. than we do today. I’m confident we will keep being driven to be useful to each other, to express our creativity, to gain status, to compete, and much more. But the specifics of what we do day to day will probably look very different. Each generation has built on the work of the generations before, and with new tools, the scaffolding gets a little taller. This collective external lattice, the set of tools that we have built up around ourselves, is remarkable, and we are capable of doing things that our great -great -grandparents couldn’t have dreamed possible. It is a moral imperative to make sure that our great -great -grandchildren can stay the same, and technology, and especially AI, is how we’re going to get there.

For a democratic AI future, it is not enough to just give people tools and wealth. We also need to give them agency and power. The vision that AI companies lay out fundamentally reduced to either unilateral control or decentralized power. sharing control means accepting that some things are going to go wrong in exchange for not having one thing go mega wrong cemented totalitarian control this is a fundamental trade -off of democracy and it is one that we believe in very strongly as the way to give everyone collective agency over the future of course this is not to suggest that we won’t need any regulation or safeguards we obviously do urgently like we have for other powerful technologies in particular we expect the world may need something like the IAEA for international coordination of AI and especially for it to have the ability to rapidly respond to change in circumstances the next few years will test global society as this technology continues to improve at a rapid pace we can choose to either empower people or concentrate power thank you very much

Speaker 1

thank you mr. Sam Altman for your very interesting and compelling remarks

Related ResourcesKnowledge base sources related to the discussion topics (17)
Factual NotesClaims verified against the Diplo knowledge base (6)
Confirmedhigh

“The moderator opened the event by praising Sam Altman’s pivotal role in moving artificial general intelligence from science‑fiction into boardrooms, parliaments and living rooms, and formally introduced him as the CEO of OpenAI.”

The knowledge base explicitly notes that the moderator highlighted Altman’s pivotal role in bringing AGI from sci-fiction concepts into practical, policy-relevant discussions [S1] and [S7].

Confirmedhigh

“Altman pointed out that more than 100 million Indians use ChatGPT each week, with roughly one‑third of those users being students.”

Source [S57] reports 100 million regular users in India, roughly one-third of the population, corroborating the scale and student proportion mentioned in the claim.

Additional Contextmedium

“India is the fastest‑growing market for Codex, OpenAI’s coding assistant, and praised the nation’s leadership in sovereign AI, infrastructure and large‑scale models.”

While Codex is not named, [S58] describes India as the world’s strongest growth market for AI, and [S51] outlines India’s three-pillar sovereignty strategy (data, infrastructure, talent), providing contextual support for the claim about rapid market growth and sovereign AI leadership.

Additional Contextmedium

“Altman warned that, on the current path, early versions of true superintelligence could appear within a couple of years.”

[S55] states that AI systems are rapidly approaching a threshold where they could exceed human cognitive abilities across many tasks, aligning with the notion of imminent superintelligence, though it does not specify a two-year horizon.

Confirmedhigh

“If this forecast proves accurate, superintelligent systems could outperform human CEOs and conduct research beyond the capabilities of today’s top scientists.”

[S63] explicitly mentions that an AGI could emulate a human CEO and make decisions, confirming the possibility of AI outperforming CEOs.

Additional Contextmedium

“Altman outlined three core beliefs: democratization of AI, AI resilience & societal safety, and broad stakeholder involvement & humility.”

The knowledge base emphasizes the need for broad stakeholder participation and societal mechanisms to manage AI risks ([S53], [S55]), adding nuance to Altman’s stated principles.

External Sources (63)
S1
Keynote-Sam Altman — -Moderator: Role/Title: Event moderator; Area of expertise: Not mentioned -Sam Altman: Role/Title: CEO of OpenAI; Area …
S2
Oversight of AI: Hearing of the US Senate Judiciary Subcommitee — 10“GPT-4 Is OpenAI’s Most Advanced System, Producing Safer and More Useful Responses.” OpenAI, https://openai.com/produc…
S3
Driving U.S. Innovation in Artificial Intelligence — 26. Sam Altman – CEO, OpenAI
S4
Keynote-Martin Schroeter — -Speaker 1: Role/Title: Not specified, Area of expertise: Not specified (appears to be an event moderator or host introd…
S5
Responsible AI for Children Safe Playful and Empowering Learning — -Speaker 1: Role/title not specified – appears to be a student or child participant in educational videos/demonstrations…
S6
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Vijay Shekar Sharma Paytm — -Speaker 1: Role/Title: Not mentioned, Area of expertise: Not mentioned (appears to be an event host or moderator introd…
S7
Keynote-Sam Altman — It’ll be very hard to outwork a GPU in many ways. It’ll be easy in some other ways. For example, we really seem hardwire…
S8
Keynote interview with Sam Altman (remote) and Nick Thompson (in-person) — Sam Altman:Great question. We don’t expect that we’re near an asymptote. But, you know, this is like a debate in the wor…
S9
GPT-5 launches with ‘PhD-level performance’ — OpenAIhas unveiledGPT-5, the latest generation of its widely used ChatGPT tool, offering what CEO Sam Altman described a…
S10
AI in education: Leveraging technology for human potential — Kevin Mills: Hello. It’s an incredible honor to be here with you today. The last UN gathering I attended was almost exac…
S11
OpenAI CEO emphasises democratic control in the future of AI — Sam Altman, co-founder and CEO of OpenAI,raisesa critical question: ‘Who will control the future of AI?’. He frames it a…
S12
Sam Altman: AI regulations should evolve in step with tech-society co-evolution | AI For Good Global Summit 2024 — In ariveting conversationduring theAI for Good Global Summit 2024, Nicholas Thompson, the CEO of The Atlantic and Sam Al…
S13
Altman urges urgent AI regulation — OpenAI chief Sam Altman hascalledfor urgent global regulation of AI, speaking at the AI Impact Summit in New Delhi. Addr…
S14
The CEO of OpenAI advocated for worldwide regulation of AI — Sam Altman, the CEO of OpenAI, called for global regulation of AI during his visit to India. While corporations typicall…
S15
How AI Drives Innovation and Economic Growth — Arguments:clearly health care and education, I think. It’s a no -brainer. I’m personally very excited, especially what h…
S16
How AI Drives Innovation and Economic Growth — I agree that there is huge potential in health. and education. I think we’ll see big improvements in that, but the risk …
S17
Son warns of vast AI leap as SoftBank outlines future risks — SoftBank chief Masayoshi Son told South Korean President Lee Jae Myung that advanced AI couldsurpass humans by an extrem…
S18
Open Forum #75 Shaping Global AI Governance Through Multistakeholder Action — Noorman warns that having AI development and deployment controlled by a small number of private companies creates risks …
S19
Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session — This comment introduced a crucial tension between the massive scale of change and the need for distributed, democratic a…
S20
Informal Stakeholder Consultation Session — Framing Security through Resilience:Argued that confidence and security should be understood through the lens of technic…
S21
Comprehensive Summary: AI Governance and Societal Transformation – A Keynote Discussion — These technological disparities will coincide with massive job displacement and economic disruption across all sectors s…
S22
How AI Drives Innovation and Economic Growth — <strong>Jeanette Rodrigues:</strong> all around the Bharat Mandapam. So once again, thank you very much for your time th…
S23
HETEROGENEOUS COMPUTE FOR DEMOCRATIZING ACCESS TO AI — This comment provides crucial context about India’s position in the global AI ecosystem, distinguishing between applicat…
S24
The Global Power Shift India’s Rise in AI & Semiconductors — Building India’s AI and Semiconductor Ecosystem: The panel discussed India’s positioning in the global AI and semiconduc…
S25
India’s AI Future Sovereign Infrastructure and Innovation at Scale — Absolutely, Ankit, just trying to, this is something which I know two years back when we said that I’m putting 8000 GPUs…
S26
Searching for Standards: The Global Competition to Govern AI | IGF 2023 — In conclusion, the UNESCO recommendation on AI ethics provides crucial guidance for global AI governance. By grounding A…
S27
Global AI Policy Framework: International Cooperation and Historical Perspectives — So global coordination will always require an inclusive participation from all stakeholders across all regions. especial…
S28
Artificial intelligence (AI) – UN Security Council — During the9821st meetingof the Artificial Intelligence Security Council, a key discussion centered around whether existi…
S29
Comprehensive Report: 18th Meeting of the Disarmament and International Security Committee — Among these, cyber security has been consulted as a priority area of action and cooperation. In this context, Guatemala …
S30
AI Governance Dialogue: Steering the future of AI — Legal and regulatory | Development governance is a shared multi stakeholder responsibility. Everyone in the generation …
S31
Multistakeholder Partnerships for Thriving AI Ecosystems — Artificial intelligence’s future depends on multi-stakeholder engagement including government, private sector, civil soc…
S32
What is it about AI that we need to regulate? — Multi-stakeholder Approaches:Despite some resistance, multi-stakeholder governance remains central. TheWS #288discussion…
S33
Keynote-Sam Altman — On our current trajectory. We believe we may be only a couple of years away from early versions of true superintelligenc…
S34
Keynote-Sam Altman — Altman projects that by the end of 2028 most of humanity’s intellectual output could be hosted inside data centres rathe…
S35
Fireside Conversation: 02 — So, usually in technological shifts of this type, we are overestimating. And the changes in the short term and overestim…
S36
Open Forum #75 Shaping Global AI Governance Through Multistakeholder Action — Human rights | Economic Noorman warns that having AI development and deployment controlled by a small number of private…
S37
Open Forum #33 Building an International AI Cooperation Ecosystem — Klauweiter argues that since AI governance is a global problem affecting all countries, only the United Nations can prov…
S38
Democratizing AI Building Trustworthy Systems for Everyone — Carsten acknowledges that coordinating international efforts in AI development presents major challenges, particularly a…
S39
AI for Democracy_ Reimagining Governance in the Age of Intelligence — “Because what we essentially need is four types of governance.”[22]. “We need a technological governance because whose v…
S40
High Level Dialogue: Strengthening the Resilience of Telecommunication Submarine Cables — The advisory body has established three working groups addressing distinct aspects of resilience: “resilience by design,…
S41
Exploring Digital Transformation for Economic Empowerment in Africa: Opportunities, Challenges, and Policy Priorities (International Trade and Research Centre, Nigeria) — In anticipation of the next industrial revolution, young people are encouraged to equip themselves with future skills to…
S42
Comprehensive Discussion Report: AI’s Transformative Potential for Global Economic Growth — Fink acknowledged that while some jobs may be displaced, new opportunities are simultaneously created. Both speakers agr…
S43
Governments, Rewired / Davos 2025 — Blair suggests that artificial intelligence and digital technologies have the potential to revolutionize various aspects…
S44
How AI Drives Innovation and Economic Growth — Arguments:Labor market disruption is the biggest concern, especially for entry-level jobs that drive economic developmen…
S45
How AI Drives Innovation and Economic Growth — <strong>Jeanette Rodrigues:</strong> all around the Bharat Mandapam. So once again, thank you very much for your time th…
S46
The Global Power Shift India’s Rise in AI & Semiconductors — Building India’s AI and Semiconductor Ecosystem: The panel discussed India’s positioning in the global AI and semiconduc…
S47
India’s AI Future Sovereign Infrastructure and Innovation at Scale — Absolutely, Ankit, just trying to, this is something which I know two years back when we said that I’m putting 8000 GPUs…
S48
AI 2.0 Reimagining Indian education system — Thank you, sir. Thank you so much for giving me the opportunity. I would like to ask a few of the… I think I’m seeing …
S49
Open Forum #58 Collaborating for Trustworthy AI an Oecd Toolkit and Spotlight on AI in Government — Jibu Elias: Thank you very much, Yoichi-san. It’s an honor to be here to share my experience building a responsible AI e…
S50
Driving Indias AI Future Growth Innovation and Impact — The discussion maintained an optimistic and forward-looking tone throughout, characterized by enthusiasm for India’s AI …
S51
Keynote ‘I’ to the Power of AI An 8-Year-Old on Aspiring India Impacting the World — India’s approach, according to the speaker, centers on three pillars of sovereignty: data sovereignty, infrastructure so…
S52
Steering the future of AI — Describes the conversation as amazing and mind-bending
S53
Thinking through Augmentation — While Ucuzoglu is optimistic about the long-term impact of transformative technology, he acknowledges that it is not an …
S54
Comprehensive Discussion Report: AI Agents and Fiduciary Standards — Pentland presented a future where AI agents would handle virtually every business and government process, essentially ad…
S55
Keynote-Dario Amodei — Amodei contends that AI systems are rapidly approaching a threshold where they will exceed human cognitive abilities acr…
S56
AI 2.0 The Future of Learning in India — Evidence:Statistics showing ChatGPT took 40 days to reach 5 crore people globally while it took telephone 75 years to re…
S57
Powering AI Global Leaders Session AI Impact Summit India — Accessforms the foundation, requiring AI tools to be financially and practically available. India exemplifies this with …
S58
Building Trusted AI at Scale Cities Startups &amp; Digital Sovereignty – Keynote Hemant Taneja General Catalyst — Taneja argued that India is uniquely positioned to lead in AI deployment due to its status as the world’s strongest grow…
S59
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Giordano Albertazzi — Albertazzi positioned India as central to the AI evolution, citing several key advantages that make the country particul…
S60
Most transformative decade begins as Kurzweil’s AI vision unfolds — AI no longer belongs to speculative fiction or distant possibility. In many ways, it has arrived. From machine translati…
S61
AI in Africa: Beyond the algorithm — **The Compute Infrastructure Divide**: 90% of global data centre capacity is held by the United States and China despite…
S62
Keynote interview with Geoffrey Hinton (remote) and Nicholas Thompson (in-person) — Machines could potentially outperform humans in cognitive tasks
S63
Artificial General Intelligence and the Future of Responsible Governance — “the same AI that can generate, can pose more sophisticated attacks and when we get to AGI right, the biggest thing is I…
Speakers Analysis
Detailed breakdown of each speaker’s arguments and positions
S
Sam Altman
12 arguments176 words per minute1413 words480 seconds
Argument 1
Rapid capability leap from high‑school to research‑level math and physics (Sam Altman)
EXPLANATION
Altman highlights the dramatic improvement in AI abilities, noting that systems have moved from struggling with basic high‑school mathematics to performing research‑level mathematics and even generating novel theoretical physics results.
EVIDENCE
He states that AI has progressed from being unable to handle high-school level math to now being capable of research-level mathematics and deriving new results in theoretical physics, illustrating the rapid leap in capability [10].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Altman’s description of the jump from high-school to research-level mathematics is recorded in the keynote and corroborated by the launch of GPT-5 achieving PhD-level performance [S1][S9].
MAJOR DISCUSSION POINT
AI capability advancement
Argument 2
Forecast that by 2028 most intellectual capacity could reside in data centres, enabling superintelligent CEOs and researchers (Sam Altman)
EXPLANATION
Altman predicts that within a few years, the majority of the world’s intellectual work may be performed by AI housed in data centres, potentially allowing AI to act as CEOs and lead scientific research better than humans.
EVIDENCE
He explains that by the end of 2028, more of the world’s intellectual capacity could be inside data centres than outside, and that a superintelligence could outperform human CEOs and researchers, while acknowledging the uncertainty of this claim [20-22].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
He projects that by the end of 2028 most of humanity’s intellectual output will be hosted inside data centres, a claim made in the keynote and reiterated in later remarks about early superintelligence timelines [S1][S7].
MAJOR DISCUSSION POINT
Future superintelligence projection
Argument 3
Massive Indian adoption: 100 M weekly ChatGPT users and fast‑growing Codex market (Sam Altman)
EXPLANATION
Altman points to India’s rapid uptake of AI tools, citing over 100 million weekly ChatGPT users and a quickly expanding market for Codex, the coding assistant, indicating strong regional demand and impact.
EVIDENCE
He reports that more than 100 million people in India use ChatGPT each week, over a third are students, and that India is the fastest-growing market for Codex, the coding agent that helps develop software faster and better [13-15].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
The keynote notes India as the fastest-growing market for Codex, citing over 100 million weekly ChatGPT users with a large student share [S7].
MAJOR DISCUSSION POINT
AI adoption in India
Argument 4
Democratization is the only safe path; centralising AI power risks ruin (Sam Altman)
EXPLANATION
Altman argues that spreading AI access widely is essential for safety, whereas concentrating AI capabilities within a single entity or nation could lead to catastrophic outcomes.
EVIDENCE
He states that democratization of AI is the only fair and safe path forward and warns that centralization of the technology in one company or country could lead to ruin [25-27].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Altman’s warning about centralising AI and his call for democratization are echoed in the keynote and reinforced by commentary on democratic AI control [S7][S11].
MAJOR DISCUSSION POINT
Need for AI democratization
AGREED WITH
Speaker 1
Argument 5
AI should amplify individual human will, not enable totalitarian trade‑offs (Sam Altman)
EXPLANATION
Altman emphasizes that AI must empower personal agency rather than being used to justify authoritarian measures, rejecting the notion of sacrificing freedom for technological benefits.
EVIDENCE
He critiques the idea of accepting totalitarianism in exchange for cures, stating that AI should extend individual human will and that such trade-offs should not be accepted [29-31].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
His emphasis on extending individual agency and rejecting totalitarian trade-offs aligns with statements about liberty, democracy, and agency in the keynote and related analyses [S7][S11].
MAJOR DISCUSSION POINT
AI and individual agency
Argument 6
Call for an international coordination body (akin to the IAEA) to oversee AI safety and rapid response (Sam Altman)
EXPLANATION
Altman proposes establishing a global institution similar to the International Atomic Energy Agency to coordinate AI governance, ensuring swift collective action when circumstances change.
EVIDENCE
He mentions that the world may need something like the IAEA for international coordination of AI and for rapid response to evolving risks [78].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Altman’s proposal for an IAEA-like AI coordination body is reflected in his call for urgent global regulation and the creation of a similar institution [S13][S14].
MAJOR DISCUSSION POINT
Global AI governance
Argument 7
Societal resilience is a core safety strategy, e.g., defending against open‑source biomodels that could create pathogens (Sam Altman)
EXPLANATION
Altman stresses that beyond technical alignment, building societal capacity to respond to misuse—such as open‑source biomodels that could be weaponized—is essential for AI safety.
EVIDENCE
He gives the example of extremely capable open-source biomodels that could help create new pathogens and calls for a society-wide approach to defend against such threats [40-42].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
The keynote explicitly frames societal resilience as a core safety strategy and cites open-source biomodels as a concrete threat [S1].
MAJOR DISCUSSION POINT
AI safety through societal resilience
Argument 8
Iterative deployment lets society integrate, understand, and steer each new AI capability (Sam Altman)
EXPLANATION
Altman advocates for a step‑by‑step rollout of AI systems, allowing time for integration, comprehension, and policy decisions before moving to the next level of capability.
EVIDENCE
He describes iterative deployment as a key strategic insight that gives society time to integrate, understand, and decide how to move forward with each new AI level, noting its success so far [53-55].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Altman’s advocacy for iterative deployment, giving society time to adapt and decide on each capability level, is outlined in the keynote’s iterative deployment section [S1].
MAJOR DISCUSSION POINT
Iterative AI rollout
Argument 9
Humility about unknowns; importance of broad, pre‑emptive debate to avoid surprise risks (Sam Altman)
EXPLANATION
Altman calls for humility regarding AI’s unknown future, urging extensive societal debate and preparation to mitigate unforeseen challenges.
EVIDENCE
He acknowledges that many predictions may be wrong, stresses humility about unknowns, and argues for broader, society-wide debate before surprises arise [46-52].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
His call for humility and broad societal debate is reflected in the keynote’s discussion of intellectual humility and the need for collective engagement [S7][S8].
MAJOR DISCUSSION POINT
Need for cautious, inclusive AI discourse
AGREED WITH
Speaker 1
Argument 10
AI drives cost reductions, faster economic growth, and improves access to healthcare and education (Sam Altman)
EXPLANATION
Altman outlines how AI is lowering costs across sectors, accelerating economic growth, and expanding high‑quality healthcare and educational services to broader populations.
EVIDENCE
He notes that AI makes many things cheaper, spurs faster economic growth, and is already enhancing access to high-quality healthcare and education, with future automation of physical goods further reducing costs [56-58].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Analyses of AI-driven innovation highlight economic growth, lower costs, and expanded access to health care and education, supporting Altman’s claim [S15].
MAJOR DISCUSSION POINT
Economic benefits of AI
Argument 11
Automation will disrupt current jobs; future work will be different but humans will find new, fulfilling roles (Sam Altman)
EXPLANATION
Altman acknowledges that AI will displace many existing jobs but argues that history shows technology creates new opportunities, and future work will be more fulfilling as humans adapt.
EVIDENCE
He discusses job disruption due to AI, the historical pattern of technology creating new work, and envisions future generations finding more fulfilling activities compared to today [60-68].
MAJOR DISCUSSION POINT
Future of work and AI disruption
Argument 12
Moral imperative to ensure future generations retain agency and power; AI must empower rather than concentrate control (Sam Altman)
EXPLANATION
Altman frames it as an ethical duty to preserve agency for future generations, insisting AI should distribute power broadly rather than enable concentration in the hands of a few.
EVIDENCE
He declares a moral imperative to protect the agency of future generations, argues that a democratic AI future requires giving people both tools and power, and warns that society can choose to empower people or concentrate power [75-78].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Altman’s moral argument for preserving agency of future generations is reinforced by his remarks on liberty, agency, and a democratic AI future [S7][S11].
MAJOR DISCUSSION POINT
Ethical responsibility for AI governance
AGREED WITH
Speaker 1
S
Speaker 1
2 arguments104 words per minute84 words48 seconds
Argument 1
Recognition of Sam Altman’s pivotal role in bringing AGI to public and policy arenas (Speaker 1)
EXPLANATION
The moderator acknowledges Altman’s significant influence in moving artificial general intelligence from theory into mainstream business, governmental, and public discussions.
EVIDENCE
The moderator states that few individuals have done more to bring AGI from science fiction into boardrooms, parliaments, and living rooms than Sam Altman, highlighting his leadership at OpenAI and the launch of ChatGPT [2-4].
MAJOR DISCUSSION POINT
Altman’s impact on AGI awareness
Argument 2
Emphasis on the need for leadership and collective shaping of AI’s future (Speaker 1)
EXPLANATION
The moderator underscores the importance of strong leadership and collective effort to guide AI development responsibly, implying a shared responsibility among stakeholders.
EVIDENCE
In the introductory remarks, the moderator frames Altman’s talk as a discussion about the future of AI and the need for collective shaping, thereby emphasizing leadership and shared responsibility (implicit in [2-4] and the closing thank-you [79]).
MAJOR DISCUSSION POINT
Call for collective AI leadership
AGREED WITH
Sam Altman
Agreements
Agreement Points
AI’s future should be shaped collectively by many stakeholders and guided by shared leadership
Speakers: Speaker 1, Sam Altman
Recognition of Sam Altman’s pivotal role in bringing AGI to public and policy arenas (Speaker 1) Emphasis on the need for leadership and collective shaping of AI’s future (Speaker 1) Democratization is the only safe path; centralising AI power risks ruin (Sam Altman) Humility about unknowns; importance of broad, pre‑emptive debate to avoid surprise risks (Sam Altman) Moral imperative to ensure future generations retain agency and power; AI must empower rather than concentrate control (Sam Altman)
Both speakers stress that the development and governance of AI must involve broad, collective participation and strong leadership, rather than being left to a single entity or country. The moderator frames this as a need for collective shaping of the AI future [2-4], while Altman explicitly says many people need a stake in shaping the outcome and calls for society-wide debate [43-44][52-53].
POLICY CONTEXT (KNOWLEDGE BASE)
This view echoes the UNESCO Recommendation on AI Ethics, which calls for inclusive, multi-stakeholder governance grounded in fundamental values [S26], and aligns with calls for an inclusive global AI policy framework that stresses participation from all regions and sectors [S27]. The AI Governance Dialogue also stresses that governance is a shared multi-stakeholder responsibility requiring “all hands on deck” [S30], while the Multistakeholder Partnerships report highlights the necessity of government, private sector, civil society and academia working together for AI’s future [S31].
Similar Viewpoints
Both see leadership and a democratic, inclusive approach as essential to steer AI development safely and ethically, warning against concentration of power and emphasizing the need for widespread societal engagement [2-4][43-44][52-53].
Speakers: Speaker 1, Sam Altman
Emphasis on the need for leadership and collective shaping of AI’s future (Speaker 1) Democratization is the only safe path; centralising AI power risks ruin (Sam Altman) Humility about unknowns; importance of broad, pre‑emptive debate to avoid surprise risks (Sam Altman) Moral imperative to ensure future generations retain agency and power; AI must empower rather than concentrate control (Sam Altman)
Unexpected Consensus
Call for a global coordination mechanism similar to the IAEA
Speakers: Speaker 1, Sam Altman
Call for an international coordination body (akin to the IAEA) to oversee AI safety and rapid response (Sam Altman) Recognition of Sam Altman’s pivotal role in bringing AGI to public and policy arenas (Speaker 1)
While the moderator’s remarks do not explicitly mention an IAEA-like body, his acknowledgment of Altman’s influence on policy arenas implicitly aligns with the need for coordinated global governance, making the convergence on the importance of international coordination unexpected [2-4][78].
POLICY CONTEXT (KNOWLEDGE BASE)
The proposal mirrors discussions at the AI Security Council, where delegates examined whether existing bodies such as the IAEA, ICAO or IPCC could serve as models for a global AI governance mechanism [S28]. Similar calls for a new global mechanism have appeared in other security contexts, indicating a broader trend toward coordinated international structures [S29].
Overall Assessment

The discussion shows a clear convergence on the principle that AI governance must be democratic, inclusive, and led by shared leadership, with both speakers warning against centralisation of power. Apart from this core consensus, there is limited overlap on other topics such as economic impact or technical capabilities.

Moderate consensus on governance and democratic principles, suggesting that future policy discussions are likely to prioritize collective stakeholder involvement and global coordination.

Differences
Different Viewpoints
Unexpected Differences
Overall Assessment

The exchange was largely harmonious. The moderator praised Altman’s influence and framed the talk as a discussion about the future of AI, while Altman elaborated on capabilities, risks, and governance. No substantive conflict emerged between the two speakers.

Minimal – the dialogue showed alignment on the overarching goal of responsibly shaping AI’s future, implying smooth cooperation rather than contention.

Partial Agreements
Both speakers emphasize the importance of collective shaping of AI’s future and the need for broad societal engagement, even though Speaker 1 highlights Altman’s leadership role while Altman stresses humility and the necessity of wide debate [2-4][43-46].
Speakers: Speaker 1, Sam Altman
Recognition of Sam Altman’s pivotal role in bringing AGI to public and policy arenas (Speaker 1) Humility about unknowns; importance of broad, pre‑emptive debate to avoid surprise risks (Sam Altman)
Takeaways
Key takeaways
AI capabilities have advanced rapidly, moving from high‑school level tasks to research‑level mathematics and theoretical physics. OpenAI projects that by 2028 a large share of the world’s intellectual capacity could reside in data centres, enabling superintelligent systems to outperform human CEOs and researchers. India shows massive AI adoption: over 100 million weekly ChatGPT users and rapid growth of the Codex coding assistant. Democratization of AI is presented as the only safe path; centralising power in a single entity or nation is deemed risky. AI should extend individual human agency rather than enable totalitarian trade‑offs. A call for an international coordination body (similar to the IAEA) to oversee AI safety and enable rapid response to emerging risks. Societal resilience and iterative deployment are highlighted as essential safety strategies, allowing societies to adapt to each new AI capability. AI is expected to drive cost reductions, faster economic growth, and improved access to healthcare and education, while also disrupting many current jobs. There is a moral imperative to ensure future generations retain agency and power, requiring AI to empower rather than concentrate control.
Resolutions and action items
Proposed creation of an international AI coordination organization (IAEA‑like) for safety oversight and rapid response. Advocated for continued iterative deployment of AI systems to give society time to integrate, understand, and steer new capabilities. Encouraged policies that promote democratization of AI access and prevent concentration of power.
Unresolved issues
How to align superintelligent AI with democratic values and prevent misuse by authoritarian regimes. Mechanisms for global governance of AI, including the structure and authority of the suggested international coordination body. Specific strategies to mitigate risks from open‑source biomodels that could be used to create pathogens. Concrete plans for managing large‑scale job displacement and ensuring new, fulfilling roles for displaced workers. How to set government policies that balance cost reductions from AI with broader societal goals. Understanding and developing new social contracts in a world where AI can be weaponised or used for geopolitical conflict.
Suggested compromises
Accepting that some AI failures may occur in exchange for avoiding a single, catastrophic concentration of power (decentralised control vs totalitarian control). Balancing regulation and safeguards with the need to give people tools, wealth, agency, and power.
Thought Provoking Comments
We may be only a couple of years away from early versions of true superintelligence. If we are right, by the end of 2028, more of the world’s intellectual capacity could reside inside of data centers than outside of them.
Sets a bold, concrete timeline for a transformative technological milestone, forcing the audience to confront the proximity of superintelligence and its societal implications.
Shifts the discussion from past progress to an imminent future, prompting listeners to consider urgent governance, safety, and ethical questions. It frames the rest of the speech around the urgency of preparation.
Speaker: Sam Altman
Democratization of AI is the only fair and safe path forward. Centralization of this technology in one company or country could lead to ruin.
Introduces a normative stance on AI governance, contrasting two extreme outcomes and positioning openness as a moral and practical safeguard.
Creates a pivot from technical achievement to policy and societal values, laying groundwork for later points about shared agency, regulation, and the need for global coordination.
Speaker: Sam Altman
AI resilience is a core safety strategy… we need a society‑wide approach about how we’re going to defend against open‑source biomodels that could help people create new pathogens.
Expands the safety conversation beyond alignment research to include societal preparedness for misuse, highlighting concrete, near‑term risks.
Broadens the scope of the safety discussion, moving it from abstract technical challenges to tangible public‑health threats, thereby deepening the audience’s perception of AI’s dual‑use nature.
Speaker: Sam Altman
Iterative deployment is a key strategic insight… society needs to contend with each successive new level of AI capability, have time to integrate it, understand it, and decide how to move forward.
Proposes a pragmatic rollout philosophy that balances rapid innovation with societal learning, offering a concrete process to manage uncertainty.
Serves as a turning point that links earlier warnings (superintelligence, centralization) to a actionable roadmap, influencing the tone toward cautious optimism.
Speaker: Sam Altman
The limit to how far cost reduction can go may only be government policy.
Highlights the role of public policy as the ultimate brake on AI‑driven economic disruption, shifting responsibility from technologists to regulators.
Introduces a new dimension—policy‑level intervention—prompting listeners to think about legislative levers and the need for proactive governance.
Speaker: Sam Altman
Technology always disrupts jobs… we should hope that future generations feel much more fulfilled than we do today.
Acknowledges the socioeconomic upheaval AI will cause while framing it as an opportunity for higher‑order human fulfillment, adding a human‑centric perspective.
Balances the earlier technical and policy discussions with a philosophical reflection on purpose, softening the narrative and inviting broader societal contemplation.
Speaker: Sam Altman
We expect the world may need something like the IAEA for international coordination of AI, with the ability to rapidly respond to change in circumstances.
Draws a parallel to an established global governance model (IAEA) to propose a concrete institutional solution for AI oversight.
Concludes the speech with a forward‑looking institutional recommendation, reinforcing the earlier call for democratization and collective agency, and leaving the audience with a tangible next step.
Speaker: Sam Altman
Overall Assessment

Sam Altman’s remarks moved the conversation from a celebratory recount of AI progress to a nuanced, forward‑looking dialogue about governance, safety, economics, and human purpose. Each of the highlighted comments acted as a pivot point—introducing new ideas (superintelligence timeline, democratization, societal resilience, iterative deployment, policy limits, job fulfillment, and global coordination) that deepened the discussion and reshaped its tone from optimism to cautious, responsible optimism. By interweaving technical milestones with ethical and policy considerations, Altman set a multi‑layered agenda that compelled the audience to think beyond immediate capabilities toward the societal structures needed to steward AI’s transformative power.

Follow-up Questions
How can superintelligence be aligned with dictators in totalitarian countries?
Understanding alignment in authoritarian contexts is crucial to prevent misuse of powerful AI systems.
Speaker: Sam Altman
How might countries use AI to conduct new forms of warfare?
Identifying AI-driven conflict scenarios is essential for developing appropriate security and diplomatic responses.
Speaker: Sam Altman
What new forms of social contracts will be needed in an AI-driven society?
Societal agreements must evolve to address the distribution of power, responsibility, and benefits of AI.
Speaker: Sam Altman
How can society develop resilience against open‑source biomodels that could be used to create pathogens?
A coordinated bio‑security strategy is needed to mitigate risks from powerful, freely available biological AI tools.
Speaker: Sam Altman
What governance mechanisms are required to ensure AI extends individual human will and supports democratic control?
Designing institutions that prevent centralization of AI power while preserving individual agency is a key safety challenge.
Speaker: Sam Altman
What international coordination structure, similar to the IAEA, is needed for AI oversight and rapid response?
A global body could facilitate cooperation, monitoring, and emergency action on AI risks across nations.
Speaker: Sam Altman
How can regulation and safeguards be implemented without undermining the democratization of AI?
Balancing safety requirements with open access is vital to avoid concentrating AI capabilities in a few hands.
Speaker: Sam Altman
How should economies manage job disruption caused by AI and ensure future human fulfillment?
Addressing workforce transitions is necessary to maintain social stability and personal well‑being.
Speaker: Sam Altman
How can equitable access to compute resources be ensured to avoid unbalanced power in AI development?
Preventing compute concentration helps maintain a level playing field and reduces the risk of dominant actors.
Speaker: Sam Altman
What best practices are needed for iterative deployment of increasingly capable AI systems?
A structured rollout allows society to adapt, learn, and set policies before each new capability level is widely adopted.
Speaker: Sam Altman

Disclaimer: This is not an official session record. DiploAI generates these resources from audiovisual recordings, and they are presented as-is, including potential errors. Due to logistical challenges, such as discrepancies in audio/video or transcripts, names may be misspelled. We strive for accuracy to the best of our ability.

Keynote-Rishad Premji

Session at a glanceSummary, keypoints, and speakers overview

Summary

The discussion centered on how artificial intelligence is a generational technology that not only expands capabilities but also reshapes societal responsibilities [15-17]. Rishad Premji argued that India’s response in the coming years will determine both its economic trajectory and its ability to address problems affecting over a billion people [18]. He noted that the global conversation on AI has moved from speculative possibilities to practical adoption and scaled impact, emphasizing that value arises only when technology solves real-world problems responsibly [24-26]. According to Premji, this moment gives India the chance to become a leading environment for testing AI against complex, multilingual, urban-rural challenges, making any success especially meaningful [28-33]. He highlighted concrete sectors where AI is already making a difference, such as improving learning outcomes in local languages, enabling earlier disease screening in rural health, and enhancing public-service infrastructure and welfare delivery [35-38].


India’s strengths include the UPI system, which processes over 20 billion transactions monthly and demonstrates that inclusive, reliable technology can scale rapidly [41-44]. The country also boasts a rapidly growing AI talent pool of about 650 000 professionals, supported by government programs to train ten million youths and a vibrant deep-tech startup ecosystem of more than 4 000 companies [45-53]. Practical deployments are already visible in agriculture, where AI models using satellite imagery provide early pest alerts that have cut crop losses by up to 25 % for farmers in several states [58-59]. In small-commerce, AI-enabled platforms help artisans automatically catalogue products, translate descriptions, optimise pricing and coordinate logistics, opening markets previously out of reach [60-62].


Premji stressed that large enterprises face fragmented data and legacy systems, so AI must be aligned with specific workflows and supported by reskilling and change-management to become sustainable at scale [70-88]. He argued that India’s competitive advantage will stem not from model size but from the choices about responsible deployment, talent development, and the ability to translate capability into impact for governments, citizens and businesses [93-98]. As an illustration, the Azeem Premji Foundation is piloting a portable X-ray and AI solution for early tuberculosis detection in rural Tamil Nadu, showing how AI can multiply scarce medical expertise and potentially be replicated worldwide [109-112]. Premji concluded that with its inclusive technology base, skilled workforce and pragmatic governance, India is well positioned to lead the AI era responsibly and at scale [103][119].


Keypoints

AI is at a generational inflection point, moving from hype to practical, large-scale impact. Premji stresses that AI “doesn’t just change what we can do… it truly changes what we must do” and that the conversation has shifted “from possibility to practicality… from pilots to scaled impact”[15-18][23-26].


India possesses unique strengths that enable it to lead AI adoption. He cites the scalability of UPI (“processes over 20 billion transactions… demonstrates that technology can scale rapidly when it is accessible, reliable, and inclusive”)[42-44], a rapidly growing AI talent pool (“approximately 650,000 professionals… will double by 2027”)[44-48], a vibrant deep-tech startup ecosystem (“more than 4,000 startups in the deep-tech and AI space”)[52-53], and strong government-backed training initiatives (“train 10 million young people in AI… industry partnerships with universities”)[49-51].


Real-world AI applications are already delivering value across sectors. He highlights education (AI-supported learning in local languages), healthcare (earlier disease screening and rural care), and public services (smarter infrastructure, reduced welfare leakages)[35-38], as well as agriculture (satellite-imagery-based pest alerts reducing crop loss by ~25%) and small-commerce platforms that automate cataloguing, translation, pricing and logistics for artisans[58-60].


Enterprise-level AI success requires workflow-aligned models and people-centric change management. Premji explains that “models designed for specific processes… become more predictable, easier to govern”[75-82] and stresses that “organizations will have to invest in change… reskilling people… so they understand the outputs and exercise judgment”[84-88], making AI sustainable at scale.


Responsible governance and scalable impact are essential, illustrated by a TB-screening pilot. India is “putting early guardrails in place, balancing accountability with innovation”[64-66], and the Azeem Premji Foundation’s pilot uses portable X-ray devices and AI to detect tuberculosis in rural Tamil Nadu, demonstrating how AI can “multiply scarce expertise” and be replicated globally[104-112][115-118].


Overall purpose:


The discussion aims to persuade the audience that India is uniquely positioned to become a global leader in AI-not merely as a technology creator but as a testbed for responsible, large-scale applications that solve real-world problems, and to call on businesses, policymakers, and talent to act decisively to realize this potential.


Tone:


Premji begins with an enthusiastic, visionary tone, celebrating AI’s transformative promise. He then adopts a pragmatic, data-driven tone when outlining India’s capabilities and the practical steps needed for enterprise adoption. The closing shifts to an inspirational, hopeful tone, using a personal health-care story to illustrate tangible impact and a call to action. Throughout, the tone remains optimistic but moves from broad vision to concrete implementation and finally to motivational advocacy.


Speakers

Rishad Premji – Executive Chairman, Wipro; AI and technology leadership [S2]


Speaker 1 – Event host / moderator [S3][S5]


Additional speakers:


Mr. Nandan Nilekani – (role/title not specified)


Mr. Dario Amote – (role/title not specified); artificial intelligence (pioneer and thought leader)


Rahul Mattan – Moderator (role/title not specified)


Full session reportComprehensive analysis and detailed insights

Speaker 1 opened the session by thanking the distinguished guests - Mr Nandan Nilekani, Mr Dario Amote - and moderator Rahul Mattan, praising the panel as “pioneers and thought leaders of artificial intelligence” who offered “profound perspectives” and “candid voices on the responsibilities that business leaders carry in times of technological disruption” [1-6]. He then introduced Mr Rishad Premji, Executive Chairman of Wipro, describing him as a thoughtful steward of Wipro’s transformation into an AI-native technology-services company [7-12].


Premji began by declaring AI a “once-in-a-generation” technology that not only expands what we can do but fundamentally changes what we must do, and he warned that India’s response in the next few years will shape both its economic trajectory and its ability to solve problems for over a billion people [15-19]. He noted that the global conversation has moved from speculative possibilities to an inflection point where the focus is on practicality-shifting “from possibility to practicality, from experimentation to adoption, and from pilots to scaled impact” [23-26]. Technology shifts inevitably create uncertainty, but for countries that act decisively they also create opportunity[95-96].


He emphasized that technology creates value only when applied responsibly to real-world problems at scale. India, he argued, can become one of the world’s most consequential environments for testing AI against complex, multilingual, urban-rural challenges [27-33]. In this context he cited concrete sectors where AI is already delivering societal benefits:


Education – supporting learning outcomes in local languages and alleviating teacher shortages and skill mismatches;


Healthcare – enabling earlier disease screening and strengthening rural care;


Public services – building smarter, safer infrastructure and reducing leakages in welfare delivery [35-38].


Premji then outlined India’s key strengths. First, the country’s experience with digital-payments infrastructure (DPI); UPI now processes over 20 billion transactions each month, demonstrating that technology can scale rapidly when it is accessible, reliable, and inclusive [42-44]. Second, India boasts a large and fast-growing AI talent pool-approximately 650 000 professionals in AI-related roles today, projected to double by 2027-supported by government initiatives to train 10 million young people in AI, industry-university partnerships, evolving curricula, and apprenticeships that give job-ready exposure [45-51]. Third, a vibrant deep-tech ecosystem of more than 4 000 AI-focused startups translates capability into real-world applications [52-53].


He illustrated these capabilities with sector examples. In agriculture, AI models trained on satellite imagery and local crop data provide early pest alerts, reducing crop losses by up to 25 % for farmers across Karnataka, Maharashtra, Telangana, Andhra Pradesh and Punjab [58-60]. In small-commerce, AI-enabled platforms automatically catalogue products, translate descriptions across languages, optimise pricing and coordinate logistics, allowing artisans in Gujarat, Tamil Nadu and Uttar Pradesh to reach markets previously out of reach [60-62].


National initiatives are expanding compute infrastructure and building capacity across the AI stack, while a pragmatic governance approach puts early guardrails in place to balance accountability with innovation, ensuring AI can scale safely and with confidence [63-66].


Premji highlighted that large enterprises face fragmented data, siloed workflows, and legacy architectures. Making AI work in such environments requires modernising legacy systems, curating and labelling data to create context-aware models, orchestrating agents reliably and securely, and earning the trust of security teams, risk leaders, regulators, and everyday users [70-78]. He argued that the most effective enterprise AI solutions are those tightly aligned to specific processes or decisions, because such models become more predictable, easier to govern, and more effective over time [79-82].


However, technology alignment alone is insufficient. Organizations must invest in change-reskilling staff, redesigning roles, and building confidence so people can interpret AI outputs and exercise judgment where it matters most [84-88]. When models are well-aligned and people are supported, AI becomes sustainable at scale, a dynamic that plays to India’s decades of experience helping complex enterprises modernise systems, manage risk, and take people along through transformation [89-92].


Looking ahead, Premji warned that India’s advantage in the AI era will be defined not by the size of its models or the scale of its infrastructure, but by the choices it makes about where to apply AI, how responsibly it is deployed, and whether capability is translated into real impact for governments, citizens and enterprises [93-98]. He stressed that AI fluency must extend beyond engineers to teachers, nurses, administrators, supervisors and small-business owners, because competitive advantage will stem from talent that can apply AI with contextual judgment and adaptability [96-98]. The dividing line will not be human versus machine, but between those who adapt and those who hesitate[97-98].


Premji concluded with a personal story from the Azeem Premji Foundation. India records 2.7 million tuberculosis cases annually; early detection is essential, yet patients often must travel long distances for sputum or molecular tests. In a pilot in rural Tamil Nadu, community health workers bring portable X-ray devices to homes; AI analyses the images instantly, flagging signs of TB and enabling faster referral without travel. With a doctor-to-population ratio of roughly 1 to 800 (and deeper shortages in rural areas), AI does not replace care-it multiplies scarce expertise[112-115]. The same last-mile challenges exist across Asia, Africa and Latin America, home to more than 4 billion people; solutions that work in India at low cost, multilingual and resilient can be exported globally [104-112][113-115][116-118].


He closed by reaffirming that India has historically embraced technological shifts and is well-positioned to do so again, urging decisive action from businesses, policymakers and educators to harness AI responsibly, at scale and with inclusive impact, so the country can contribute not only to building AI but also to applying it to solve pressing problems for enterprises, the nation and the world [103][119][120].


Session transcriptComplete transcript of the session
Speaker 1

Thank you so much, Mr. Nandan Nilekani and Mr. Dario Amote. And thank you, Rahul Mattan, for moderating it. Well, it was quite an engaging conversation, I must say. And they are the pioneers and the thought leaders of artificial intelligence. And they shared their profound perspectives. I thank our panelists. Ladies and gentlemen, our next speaker is Mr. Rishad Premji. He is the executive chairman of Wipro, the son of one of India’s most beloved business leaders. Rishad Premji has carved out his own identity as a thoughtful steward of Wipro’s transformation into an artificial intelligence. artificial intelligence native technology services company. He’s also an unusually candid voice on the responsibilities that business leaders carry in times of technological disruption.

Ladies and gentlemen, please welcome the Executive Chairman of Wipro, Mr. Rishad Premji. A warm welcome once again. That’s Mr. Rishad Premji.

Rishad Premji

You know, thank you for those of you who are here for being here. Once in a generation, a technology emerges that doesn’t just change what we can do. It truly changes what we must do. AI for me is certainly that technology. And how we as a country, how India responds in the next few years, will shape not just our own economic trajectory, but our ability to solve problems that matter to over a billion people. For the past several years, the conversation around AI has focused just not on the possibility. What can it do? How powerful could it become? How quickly could it evolve? But we are now at an inflection point. The conversation has fundamentally shifted from possibility to practicality.

From experimentation to adoption and from pilots to scaled impact. This shift matters and it matters tremendously because technology creates value only when it is applied to solve real world problems responsibly and at scale. So what does this moment mean for India? It means India has the opportunity to become one of the world’s most consequential environments for the application of AI. Not just as a builder of the technology. But as a place where AI is tested against real world problems. complexity and made to work at scale. Our context as a country is demanding. Systems here must work across multiple languages, across urban and rural settings, and across populations with very different levels of access, need, data quality, and infrastructure.

That raises the bar, but it also makes success meaningful. We already see what this looks like in practice. In education, AI can support learning outcomes in local languages and help address teachers’ shortages and skill mismatches. In healthcare, it can enable earlier disease screening and strengthen rural care, especially where access is limited. And in public services, it can help build smarter, safer infrastructure and reduce leakages in welfare delivery. Which brings me to India’s strengths and why I believe India is a great country. India is well -placed today to take advantage of all of this. Let me just highlight a few that matter most and have been highlighted by many. One of the most significant is our experience with DPI.

UPI today processes over 20 billion transactions every month and has transformed how individuals and businesses participate in the digital economy. It has demonstrated that technology can scale rapidly when it is accessible, reliable, and most importantly, inclusive. India also has one of the largest and fastest growing pools of AI talent in the world. We are truly the AI and talent destination of the world. Approximately 650 ,000 professionals in India work in AI -related roles today, and this number will double by 2027. This talent brings not only technical capability, but importantly, practical experience in applying technology in complex real -world situations. and in real -world environments. Equally importantly, many of the foundations to build out this talent are already in place.

Government initiatives to train 10 million young people in AI, along with industry partnerships with universities, are expanding access to practical, job -ready training. Curricula is evolving and people are giving opportunities as apprenticeships to get exposure to real -world applications. This capability is reinforced by a vibrant innovation ecosystem. India is home, as many of us know, to the world’s third -largest technology startup base, including more than 4 ,000 startups in the deep tech and AI space. Together these companies are helping translate technological capability into practical, real -world applications. We are already seeing what this looks like on the ground. In agriculture, for example. We are already seeing what this looks like on the ground. We are already seeing what this looks like on the ground.

We are already seeing what this looks like on the ground. farmers across Karnataka, Maharashtra, Telangana, AP and Punjab are using AI systems trained on satellite imagery and local crop data. These systems provide early pest alerts and in some regions have reduced crop losses by nearly 25%. In small commerce, artisans in Gujarat, Tamil Nadu and UP are using AI -related platforms connected to open networks. Products are automatically catalogued, descriptions translated across languages, prices optimised and logistics coordinated, allowing small sellers to reach markets that were once out of reach. And we are also seeing deliberate investments in the future. National initiatives are expanding access to compute infrastructure and building capacity across the AI stack in our country. At the same time, an equally important we as a country are taking a very pragmatic approach to governance.

We are also seeing a rapid growth in the AI industry. putting early guardrails in place, but balancing accountability with innovation, so AI can scale safely and with confidence. All of this lays a strong foundation for India to lead in the AI era. But there is one more factor that I believe positions us as a country uniquely, our long engagement with global enterprises. Today, the constraint, as many have said, is not access to technology. The real world begins when AI is introduced into large, real -world organizations. In those environments, technology has evolved over many years. Application landscapes are complex. Data is fragmented. Workflows are siloed. Processes vary across geography, business units, and regulatory regimes. Decision -making is rarely uniform.

Making AI work in this environment means modernizing legacy architectures. It means curating and labeling data to create highly specialized context -aware models. It means orchestrating across agents in ways that are reliable and secure. And it means earning confidence of security teams, risk leaders, regulators, and critically, the people who are expected to use these systems every day. This is where a more practical pattern has emerged in enterprises. Models designed for specific processes or decisions tend to deliver the most reliable results. When AI is closely aligned to a defined workflow, it becomes more predictable, easier to govern, and more effective over time. What enterprises need today is not a model that does everything, but models that do the right thing consistently, inside how work actually happens.

But technology alignment alone is not enough. For these systems to deliver value in organizations, organizations themselves will have to invest in change. Taking people truly along, helping them adapt to new ways of working, redesigning roles and decision making, and building confidence in how AI is used. That includes reskilling people, reskilling teams to work effectively with AI tools so that they understand the outputs and exercise judgment where it matters most. When models are well aligned to workflows and people are supported through the transition, AI becomes just not deployable, but it becomes sustainable at scale. And that plays directly to India’s strengths. We have decades of experience. We have experience working inside complex enterprises, helping them modernize systems, manage risk, and take people along through this change.

And that works in environments like these is not just deployable, it is resilient, responsible, and truly scalable. The same dynamics that determine success inside organizations will also shape how countries navigate this moment. As we look ahead, India’s advantage in AI will not be defined only by the size of our models or the scale of our infrastructure. It will be defined by the choices we make. About where we apply AI, how we diffuse it, how responsibly is it deployed, and whether we can translate capability into real impact for governments, citizens, and enterprises. India’s advantage will come from developing talent at scale, not just people trained on AI, but people who can apply it with context, judgment, and an ability to adapt to change.

That is why AI fluency must extend beyond engineering. For more information, visit www .international .com to teachers, to nurses, to administrators, to supervisors, to small business owners, among everyone else. The dividing line will not be human versus machine. It will truly be between those who adapt and those who hesitate to adapt. Technology shifts inevitably create uncertainty. But for countries that act decisively, they also create opportunity. India has embraced such shifts in the past, and I believe we are really well positioned to do so again. I want to close by sharing a personal story from the work of the Azeem Premji Foundation, which is also the majority shareholder of our company, Wipro. India sees 2 .7 million tuberculosis cases every year, making it one of the country’s most serious public health challenges.

Early detection is essential. But confirming TB often requires patients to travel to distant public hospitals for sputum or molecular tests. For many, the real barrier is not medical capability, but it is access. To address this, our foundation is running a pilot in a rural community in Tamil Nadu. Community health workers carry portable x -ray devices directly to people’s homes. AI analyzes the x -rays instantly and identifies signs consistent with TB, enabling that early screening and faster referral without requiring patients to travel. If successful, this approach can help detect TB earlier and extend the reach of healthcare into communities that need it most. With a national doctor -to -population ratio of roughly 1 is to 800 and even deeper shortages in rural India, AI does not replace care.

It multiplies scarce expertise. AI is able to address these challenges infinitely. The same last -mile challenges exist across other countries in continents of Asia, Africa, and Latin America. home to more than 4 billion people. Solutions that work here in India at scale, low cost, multilingual and resilient can travel far beyond our own borders. If we can do that, India’s contribution can be vast, not just in building AI, but in applying it to solve problems for enterprises, for our own country and for the world at large. Thoughtfully, inclusively and with impact at scale. Thank you for listening to me.

Related ResourcesKnowledge base sources related to the discussion topics (13)
Factual NotesClaims verified against the Diplo knowledge base (5)
Confirmedhigh

“Speaker 1 thanked Mr Nandan Nilekani, Mr Dario Amote and moderator Rahul Mattan, describing them as “pioneers and thought leaders of artificial intelligence”.”

The transcript excerpt explicitly thanks Nilekani and Amote and calls them pioneers and thought leaders of AI, matching the report [S2].

Confirmedmedium

“The global conversation on AI has shifted from speculative possibilities to a focus on practicality, experimentation to adoption, and pilots to scaled impact.”

The knowledge base notes that the conversation has moved from distant existential risks to the practicalities of current regulation, confirming the shift toward practicality [S38].

Confirmedmedium

“India can become one of the world’s most consequential environments for testing AI against complex, multilingual, urban‑rural challenges.”

Analyses describe India as a micro-cosm with extreme linguistic and socio-economic diversity, making it an ideal testing ground for AI solutions that must work at scale [S40] and as a demanding test case for AI [S41].

Confirmedmedium

“AI is already delivering societal benefits in education (local‑language learning, teacher shortage relief) and healthcare (earlier disease screening, stronger rural care).”

Reports highlight AI’s potential to improve access to education and to enhance healthcare delivery, supporting the claim of sectoral benefits [S43].

Additional Contextlow

“India’s linguistic diversity (22 official languages with regional variations) creates unique datasets that can be leveraged for AI development.”

A keynote on India’s strategic advantages points out the country’s linguistic diversity as a source of unique AI-ready data, adding nuance to the claim about multilingual challenges [S42].

External Sources (44)
S1
Keynote-Rishad Premji — -Moderator: Role/Title: Event moderator; Area of expertise: Not specified -Mr. Dario Amote: Role/Title: Not specified; …
S2
Keynote-Rishad Premji — Evidence:Artisans in Gujarat, Tamil Nadu and UP are using AI-related platforms connected to open networks. Products are …
S3
Keynote-Martin Schroeter — -Speaker 1: Role/Title: Not specified, Area of expertise: Not specified (appears to be an event moderator or host introd…
S4
Responsible AI for Children Safe Playful and Empowering Learning — -Speaker 1: Role/title not specified – appears to be a student or child participant in educational videos/demonstrations…
S5
Building Trusted AI at Scale Cities Startups &amp; Digital Sovereignty – Keynote Vijay Shekar Sharma Paytm — -Speaker 1: Role/Title: Not mentioned, Area of expertise: Not mentioned (appears to be an event host or moderator introd…
S6
Nepal Engagement Session — India’s proven track record with large-scale digital infrastructure projects like Aadhaar, UPI, and GST, combined with i…
S7
The Global Power Shift India’s Rise in AI &amp; Semiconductors — -Building India’s AI and Semiconductor Ecosystem: The panel discussed India’s positioning in the global AI and semicondu…
S8
AI Innovation in India — The biggest delta multiplier of AI, the benefactor of this is India… 1.4 billion will be 1.6 by 2060. 1.6 billion peop…
S9
The Global Power Shift India’s Rise in AI & Semiconductors — Building India’s AI and Semiconductor Ecosystem: The panel discussed India’s positioning in the global AI and semiconduc…
S10
Sustainable development — AI-powered tools like remote sensing, drones, and predictive analytics can enhance precision agriculture practices. They…
S11
Building the Next Wave of AI_ Responsible Frameworks &amp; Standards — The panel demonstrated a maturing field where practitioners are converging on core principles while offering complementa…
S12
HETEROGENEOUS COMPUTE FOR DEMOCRATIZING ACCESS TO AI — This comment provides crucial context about India’s position in the global AI ecosystem, distinguishing between applicat…
S13
Building the Future STPI Global Partnerships & Startup Felicitation 2026 — India possesses several competitive advantages that position it well for AI innovation and startup growth. These include…
S14
Rethinking Africa’s digital trade: Entrepreneurship, innovation, &amp; value creation in the age of Generative AI (depHub) — Frontier technologies, including Artificial Intelligence (AI), have the power to bring about transformative changes in s…
S15
AI Transformation in Practice_ Insights from India’s Consulting Leaders — A kind of different angle and a question to you. You know, you talked about how AI has impacted some of your work intern…
S16
Keynote-Rishad Premji — India is taking a pragmatic approach to governance, balancing accountability with innovation
S17
Keynote-Rishad Premji — Implementation of early governance guardrails while maintaining innovation balance
S18
Building the Next Wave of AI_ Responsible Frameworks & Standards — These key comments collectively transformed the discussion from abstract principles to concrete, actionable approaches f…
S19
Large Language Models on the Web: Anticipating the challenge | IGF 2023 WS #217 — The analysis highlighted the delicate balance between freedom to innovate and responsible innovation principles. While i…
S20
Advancing Scientific AI with Safety Ethics and Responsibility — The speakers demonstrated strong consensus on several key areas: the need for context-specific governance frameworks tai…
S21
WS #123 Responsible AI in Security Governance Risks and Innovation — Michael Karimian: Thank you Yasmin, it’s a pleasure to join you all and thank you Yasmin not just for facilitating today…
S22
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Matthew Prince Cloudflare — Disagreement level:This transcript contains only a single speaker (Matthew Prince) presenting his vision for AI developm…
S23
Keynote-Rishad Premji — The conversation has shifted from possibility to practicality, from experimentation to adoption and scaled impact
S24
Keynote-Rishad Premji — The conversation has fundamentally shifted from possibility to practicality. From experimentation to adoption and from p…
S25
AI Innovation in India — India’s unique strength lies in its people’s ability to work in unstructured environments and get jobs done regardless o…
S26
HETEROGENEOUS COMPUTE FOR DEMOCRATIZING ACCESS TO AI — This comment provides crucial context about India’s position in the global AI ecosystem, distinguishing between applicat…
S27
Building the Future STPI Global Partnerships & Startup Felicitation 2026 — India possesses several competitive advantages that position it well for AI innovation and startup growth. These include…
S28
AI Innovation in India — This comment energized the discussion by providing a grand vision that contextualized all the individual innovations wit…
S29
HETEROGENEOUS COMPUTE FOR DEMOCRATIZING ACCESS TO AI — India’s unique position—combining technical talent, diverse datasets, a vibrant startup ecosystem, and supportive policy…
S30
Rethinking Africa’s digital trade: Entrepreneurship, innovation, &amp; value creation in the age of Generative AI (depHub) — Frontier technologies, including Artificial Intelligence (AI), have the power to bring about transformative changes in s…
S31
How Small AI Solutions Are Creating Big Social Change — He describes how pilots in various sectors are being scaled from villages to larger populations, focusing on offline, pl…
S32
AI Transformation in Practice_ Insights from India’s Consulting Leaders — A kind of different angle and a question to you. You know, you talked about how AI has impacted some of your work intern…
S33
AI Transformation in Practice_ Insights from India’s Consulting Leaders — The conversation concluded with optimism about AI’s potential to create abundance and societal impact, whilst acknowledg…
S34
Building Scalable AI Through Global South Partnerships — Sure. Hi, everyone. Thank you. Welcome. Thanks for being here. Thank you for having me. I suspect the way I got over her…
S35
Panel Discussion AI in Healthcare India AI Impact Summit — Aditya views India as a large‑scale pilot; success there can provide a replicable model for other low‑ and middle‑income…
S36
Fireside Conversation: 01 — Speakers:Dario Amodei, Nandan Nilekani Speakers:Rahul Matthan, Nandan Nilekani
S37
AI for Social Empowerment_ Driving Change and Inclusion — say, you know, it’s yet to unfold. We don’t know what the impact is and it’s yet to unfold. I believe that that contenti…
S38
Review of AI and digital developments in 2024 — The conversation has shifted—from distant existential risks to the practicalities of current regulation. The EU AI Act a…
S39
Biology as Consumer Technology — In conclusion, scientific understanding, particularly in biology, remains limited. However, AI offers unique capabilitie…
S40
From India to the Global South_ Advancing Social Impact with AI — Darren Farrant argues that India serves as a microcosm of the world, with all the diversity and challenges that exist gl…
S41
AI Without the Cost Rethinking Intelligence for a Constrained World — Bernie identifies India as presenting the most demanding test case for AI solutions due to its scale and constraints. He…
S42
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Vivek Raghavan Sarvam AI — India’s Strategic Advantages Rather than viewing India’s complexity as a challenge, Raghavan presented it as the countr…
S43
AI for Democracy_ Reimagining Governance in the Age of Intelligence — Chunggong acknowledges the significant positive potential of AI for social good, including improvements in healthcare de…
S44
Press Conference: Closing the AI Access Gap — Moreover, the speakers argue that AI can drive productivity, creativity, and overall economic growth. It has the capacit…
Speakers Analysis
Detailed breakdown of each speaker’s arguments and positions
R
Rishad Premji
14 arguments137 words per minute1643 words718 seconds
Argument 1
AI as a generational, practical technology – AI reshapes obligations, moving from possibility to scaled impact (Rishad Premji)
EXPLANATION
Premji describes AI as a once‑in‑a‑generation technology that not only expands what can be done, but also changes what societies must do. He notes that the debate has shifted from speculative possibilities to concrete, large‑scale adoption that delivers real‑world value.
EVIDENCE
He states that AI is a generational technology that changes what we must do, emphasizing its transformative power for the country’s economic trajectory and problem-solving capacity [15-18]. He then explains that the conversation has moved from possibility to practicality, from pilots to scaled impact, and that technology creates value only when applied responsibly at scale [23-26].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Premji emphasizes the shift from experimentation to scaled impact, noting that technology creates value only when applied responsibly at scale [S1][S2].
MAJOR DISCUSSION POINT
AI as a generational, practical technology
AGREED WITH
Speaker 1
Argument 2
India’s strengths for AI leadership – Scalable, inclusive digital infrastructure exemplified by UPI (Rishad Premji)
EXPLANATION
Premji highlights India’s existing digital payment infrastructure, UPI, as proof that technology can be rapidly scaled when it is accessible, reliable, and inclusive. This experience demonstrates India’s capacity to support large‑scale AI deployments.
EVIDENCE
He cites that UPI processes over 20 billion transactions each month and has transformed participation in the digital economy, showing that technology can scale rapidly when inclusive [41-44].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
He cites UPI processing over 20 billion transactions monthly and its inclusive, rapid scaling as proof of India’s digital infrastructure strength [S1][S6].
MAJOR DISCUSSION POINT
Scalable, inclusive digital infrastructure
Argument 3
India’s strengths for AI leadership – Large, fast‑growing AI talent pool and government training initiatives (Rishad Premji)
EXPLANATION
Premji points to India’s sizable and expanding AI workforce, estimating 650,000 AI professionals today with numbers set to double by 2027. He also notes government programmes to train ten million young people and industry‑university partnerships that build practical, job‑ready AI skills.
EVIDENCE
He mentions the current 650,000 AI-related professionals and projected doubling by 2027, as well as government initiatives to train 10 million youth, industry-university collaborations, evolving curricula, and apprenticeship opportunities that give real-world exposure [44-51].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Panel discussions highlight India’s vast engineering talent and government programmes aimed at upskilling millions, underscoring the talent pipeline [S7][S8].
MAJOR DISCUSSION POINT
Growing AI talent and training ecosystem
Argument 4
India’s strengths for AI leadership – Vibrant deep‑tech startup ecosystem translating capability into applications (Rishad Premji)
EXPLANATION
Premji emphasizes that India hosts the world’s third‑largest tech startup base, including over 4,000 deep‑tech and AI startups, which are turning technological capability into practical solutions for real‑world problems.
EVIDENCE
He notes India’s position as home to the third-largest technology startup ecosystem and more than 4,000 deep-tech/AI startups that help translate capability into real-world applications [51-54].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
The keynote points to a thriving innovation ecosystem with thousands of deep-tech startups turning capability into real-world solutions [S2][S7].
MAJOR DISCUSSION POINT
Vibrant deep‑tech startup ecosystem
Argument 5
Real‑world AI applications delivering societal value – AI supports multilingual education and mitigates teacher shortages (Rishad Premji)
EXPLANATION
Premji describes AI tools that can deliver learning outcomes in local languages, helping to address chronic teacher shortages and skill mismatches in India’s education system.
EVIDENCE
He gives the example that AI can support learning outcomes in local languages and help address teachers’ shortages and skill mismatches [35-36].
MAJOR DISCUSSION POINT
AI in education
Argument 6
Real‑world AI applications delivering societal value – AI enables earlier disease screening and strengthens rural healthcare (Rishad Premji)
EXPLANATION
Premji argues that AI can facilitate earlier disease detection and improve the quality of rural health services, where access to care is limited.
EVIDENCE
He states that AI can enable earlier disease screening and strengthen rural care, especially where access is limited [36-37].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Premji describes AI-driven early disease screening and a portable X-ray/TB detection pilot that extend scarce expertise in rural areas [S1][S2].
MAJOR DISCUSSION POINT
AI in healthcare
Argument 7
Real‑world AI applications delivering societal value – AI improves public services, infrastructure safety, and welfare delivery (Rishad Premji)
EXPLANATION
Premji notes that AI can make public services smarter and safer, improve infrastructure reliability, and reduce leakages in welfare programmes.
EVIDENCE
He mentions that AI can help build smarter, safer infrastructure and reduce leakages in welfare delivery [37].
MAJOR DISCUSSION POINT
AI in public services
Argument 8
Real‑world AI applications delivering societal value – AI in agriculture provides pest alerts, cutting crop losses by up to 25% (Rishad Premji)
EXPLANATION
Premji provides evidence that AI systems using satellite imagery and local crop data give farmers early pest warnings, leading to a measurable reduction in crop losses.
EVIDENCE
He describes farmers across several states using AI trained on satellite imagery and local data to receive early pest alerts, which have reduced crop losses by nearly 25% in some regions [58-60].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
He notes AI-based pest alerts that have reduced crop losses by nearly 25% and references broader AI tools for precision agriculture [S1][S10].
MAJOR DISCUSSION POINT
AI in agriculture
Argument 9
Real‑world AI applications delivering societal value – AI platforms help small merchants catalog, translate, price‑optimize, and reach new markets (Rishad Premji)
EXPLANATION
Premji explains that AI‑enabled platforms allow artisans and small sellers to automatically catalogue products, translate descriptions, optimise pricing, and coordinate logistics, thereby expanding market access.
EVIDENCE
He cites artisans in Gujarat, Tamil Nadu and UP using AI-related platforms that automatically catalogue products, translate across languages, optimise prices and coordinate logistics, enabling them to reach previously inaccessible markets [60-62].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Artisans in Gujarat, Tamil Nadu and UP use AI platforms to auto-catalogue, translate, price-optimize and coordinate logistics, expanding market access [S2].
MAJOR DISCUSSION POINT
AI for small commerce
Argument 10
Pragmatic governance and responsible deployment – Early guardrails balance accountability with innovation for safe scaling (Rishad Premji)
EXPLANATION
Premji says India is adopting a pragmatic approach to AI governance, putting early guardrails in place that balance accountability with the need for innovation, allowing AI to scale safely and confidently.
EVIDENCE
He notes that India is taking a pragmatic approach to governance, putting early guardrails in place while balancing accountability with innovation so AI can scale safely and with confidence [64-66].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
India is adopting early AI guardrails that balance accountability with innovation, a pragmatic approach echoed in responsible-AI frameworks [S1][S11].
MAJOR DISCUSSION POINT
Balanced AI governance
AGREED WITH
Speaker 1
Argument 11
Enterprise adoption: workflow alignment and people change – Process‑specific models deliver reliable, governable results (Rishad Premji)
EXPLANATION
Premji argues that AI models tailored to specific business processes or decisions are more predictable, easier to govern, and deliver consistent results, making them preferable for enterprise adoption.
EVIDENCE
He explains that models designed for specific processes tend to deliver the most reliable results, and when AI is closely aligned to a defined workflow it becomes more predictable, easier to govern and more effective over time [80-82].
MAJOR DISCUSSION POINT
Process‑specific AI models
Argument 12
Enterprise adoption: workflow alignment and people change – Reskilling, role redesign, and confidence building are essential for sustainable AI (Rishad Premji)
EXPLANATION
Premji stresses that beyond technical alignment, organizations must invest in people—through reskilling, redesigning roles, and building confidence—to ensure AI delivers sustainable value at scale.
EVIDENCE
He states that organizations need to take people along, help them adapt, redesign roles, reskill teams to work with AI tools, and build confidence, so that AI becomes sustainable at scale [84-87].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Premji stresses the need to take people along, reskill teams and redesign roles to ensure sustainable AI adoption [S1].
MAJOR DISCUSSION POINT
People‑centric AI adoption
Argument 13
AI fluency and adaptation as competitive advantage – Competitive edge comes from talent that applies AI with context, judgment, and adaptability (Rishad Premji)
EXPLANATION
Premji contends that a nation’s AI advantage will stem not just from model size or infrastructure, but from developing talent that can apply AI responsibly, with contextual judgment and the ability to adapt to change.
EVIDENCE
He says India’s advantage will be defined by the choices we make, by developing talent at scale who can apply AI with context, judgment and adaptability, and that AI fluency must extend beyond engineering [93-98].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
The competitive advantage is framed around developing AI-fluent talent capable of contextual judgment, aligning with observations on India’s engineering talent and execution strengths [S7][S8].
MAJOR DISCUSSION POINT
AI fluency as a strategic asset
Argument 14
Pilot example: AI‑driven TB detection at the last mile – Portable X‑ray + AI in rural Tamil Nadu enables early TB screening, extending scarce expertise without replacing clinicians (Rishad Premji)
EXPLANATION
Premji shares a pilot where community health workers bring portable X‑ray devices to homes, and AI instantly analyses the images to detect TB, allowing early screening and faster referrals without substituting doctors.
EVIDENCE
He describes a pilot in a rural Tamil Nadu community where health workers carry portable X-ray devices, AI analyses the images instantly to identify TB signs, enabling early screening and faster referral without requiring patients to travel, thereby multiplying scarce expertise rather than replacing care [109-115].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
A pilot in rural Tamil Nadu uses portable X-ray devices and AI to instantly detect TB, enabling early screening and referrals without replacing clinicians [S1][S2].
MAJOR DISCUSSION POINT
AI‑enabled TB screening pilot
Agreements
Agreement Points
AI as a generational, practical technology – AI reshapes obligations, moving from possibility to scaled impact (Rishad Premji)
Speakers: Speaker 1, Rishad Premji
AI as a generational, practical technology – AI reshapes obligations, moving from possibility to scaled impact (Rishad Premji)
Both speakers acknowledge AI as a once-in-a-generation technology that changes what societies must do and note the shift from speculative possibilities to concrete, large-scale adoption that delivers real-world value. Speaker 1 calls the panelists “pioneers and thought leaders of artificial intelligence” and highlights their “candid voice on the responsibilities that business leaders carry” [4-5], while Rishad Premji explicitly describes AI as a generational technology that changes what we must do and stresses the move from possibility to practicality and from pilots to scaled impact [15-18][23-26].
POLICY CONTEXT (KNOWLEDGE BASE)
Rishad Premji frames AI as a generational, practical technology whose impact is shifting from experimental to scaled, echoing India’s pragmatic governance stance that balances accountability with innovation as outlined in his keynote [S16][S17].
Pragmatic governance and responsible deployment – Early guardrails balance accountability with innovation for safe scaling (Rishad Premji)
Speakers: Speaker 1, Rishad Premji
Pragmatic governance and responsible deployment – Early guardrails balance accountability with innovation for safe scaling (Rishad Premji)
Speaker 1 emphasizes the need for responsible leadership in AI ([4-5]), and Rishad Premji later states that India is taking a pragmatic approach to AI governance by putting early guardrails in place while balancing accountability with innovation to allow safe scaling [64-66].
POLICY CONTEXT (KNOWLEDGE BASE)
The call for early governance guardrails that balance accountability and innovation aligns with Premji’s pragmatic approach to AI policy and mirrors broader discussions on responsible innovation and the need to protect freedom to experiment while mitigating harm [S16][S17][S19].
Similar Viewpoints
Both speakers stress that AI’s transformative power comes with a responsibility to govern it prudently, moving the discourse from hype to practical, responsible deployment. Speaker 1’s introductory remarks praise the panelists’ candidness on responsibility, while Rishad Premji provides concrete examples of shifting conversation and governance mechanisms [4-5][15-18][23-26][64-66].
Speakers: Speaker 1, Rishad Premji
AI as a generational, practical technology – AI reshapes obligations, moving from possibility to scaled impact (Rishad Premji) Pragmatic governance and responsible deployment – Early guardrails balance accountability with innovation for safe scaling (Rishad Premji)
Unexpected Consensus
Alignment on the need for responsible AI governance despite Speaker 1’s brief introductory role
Speakers: Speaker 1, Rishad Premji
Pragmatic governance and responsible deployment – Early guardrails balance accountability with innovation for safe scaling (Rishad Premji)
It is noteworthy that even though Speaker 1 only delivered a short introductory statement, they already echoed the same theme of responsible AI governance that Rishad Premji later elaborates on. This early alignment on governance and accountability was not explicitly anticipated given the limited content of the introduction [4-5][64-66].
POLICY CONTEXT (KNOWLEDGE BASE)
This consensus reflects the emerging responsible AI frameworks and standards highlighted in recent reports, which emphasize context-specific governance, pre-deployment assessment, and decentralized approaches for developing countries [S18][S20].
Overall Assessment

The discussion shows clear convergence between the moderator’s framing of AI leadership and responsibility and Rishad Premji’s detailed articulation of AI as a generational technology that must be deployed responsibly at scale. Both speakers stress the shift from possibility to practical impact and the importance of early governance guardrails.

High consensus on the transformative nature of AI and the necessity of responsible, pragmatic governance, reinforcing a shared vision for India’s AI future.

Differences
Different Viewpoints
Unexpected Differences
Overall Assessment

The transcript contains an introductory segment by Speaker 1 that solely offers acknowledgments and welcomes, followed by a single substantive contribution from Rishad Premji. No opposing viewpoints or contrasting policy prescriptions are presented, resulting in an absence of observable disagreement among the speakers.

Minimal – the discussion reflects consensus or at least a lack of conflicting positions, implying smooth alignment on the topics addressed.

Takeaways
Key takeaways
AI is a generational technology that has moved from a focus on possibility to practical, scalable impact, reshaping societal obligations. India possesses distinct strengths for AI leadership: inclusive digital infrastructure exemplified by UPI, a large and rapidly expanding AI talent pool, government‑driven training programs, and a vibrant deep‑tech startup ecosystem. Real‑world AI applications are already delivering societal value in education (multilingual learning, teacher support), healthcare (early disease screening, rural care), public services (infrastructure safety, welfare delivery), agriculture (pest alerts reducing crop loss up to 25%), and small commerce (automated cataloguing, translation, pricing, logistics). Pragmatic governance with early guardrails is essential to balance accountability and innovation, enabling safe scaling of AI. Enterprise adoption succeeds when AI models are tightly aligned to specific workflows and when organizations invest in people‑centric change—reskilling, role redesign, and confidence‑building. Competitive advantage comes from AI fluency that extends beyond engineering to include contextual judgment, adaptability, and the ability to apply AI responsibly at scale. A pilot AI‑driven TB detection program in rural Tamil Nadu illustrates how AI can multiply scarce expertise, improve last‑mile health access, and serve as a model for other low‑resource settings worldwide.
Resolutions and action items
Continue and expand government initiatives to train 10 million young Indians in AI skills. Scale national compute‑infrastructure programs to support AI development across the AI stack. Deploy and evaluate the AI‑enabled portable X‑ray TB screening pilot in Tamil Nadu, with a view to scaling if successful. Promote early AI guardrails that balance regulatory accountability with innovation to enable safe, rapid deployment.
Unresolved issues
How to create a comprehensive, nation‑wide governance framework that ensures responsible AI while keeping pace with rapid innovation. Methods for curating, labeling, and integrating fragmented enterprise data to build context‑aware AI models at scale. Strategies for ensuring AI solutions remain effective across India’s linguistic, geographic, and infrastructural diversity. Mechanisms to build sustained trust among security teams, risk leaders, regulators, and end‑users in AI‑driven workflows. Long‑term financing and business models for scaling AI pilots (e.g., TB detection) into nationwide programs.
Suggested compromises
Implement early, proportionate guardrails that provide accountability without stifling innovation. Focus on process‑specific AI models rather than attempting universal models, thereby simplifying governance and improving predictability. Align AI deployment with existing enterprise workflows to make regulation and risk management more manageable.
Thought Provoking Comments
AI is a technology that doesn’t just change what we can do, it truly changes what we must do.
Frames AI as a societal inflection point rather than a mere tool, shifting the conversation from technical capability to ethical and strategic responsibility.
Sets the tone for the entire address, prompting listeners to think about AI’s broader implications for policy, education, and public welfare, and prepares the audience for later points about responsible deployment.
Speaker: Rishad Premji
We are now at an inflection point where the conversation has shifted from possibility to practicality – from experimentation to adoption and from pilots to scaled impact.
Identifies a clear transition in the AI lifecycle, highlighting that the real challenge now lies in implementation at scale rather than proof‑of‑concept.
Redirects focus toward operational challenges, leading to discussion of real‑world use cases (education, healthcare, public services) and the need for governance frameworks.
Speaker: Rishad Premji
India’s experience with UPI – processing over 20 billion transactions a month – shows that technology can scale rapidly when it is accessible, reliable, and inclusive.
Uses a concrete, home‑grown success story to illustrate how inclusive digital infrastructure can be a springboard for AI adoption.
Provides a tangible benchmark for the audience, reinforcing the argument that India has the foundational ecosystem needed for AI at scale and encouraging confidence in replicating similar models.
Speaker: Rishad Premji
Approximately 650,000 professionals in India work in AI‑related roles today, and this number will double by 2027. This talent brings not only technical capability but also practical experience in applying technology in complex real‑world situations.
Quantifies the talent pool and links it to real‑world problem‑solving, moving the discussion from abstract potential to measurable human capital.
Strengthens the narrative that India is not just a consumer of AI but a creator and implementer, prompting listeners to consider investments in education and apprenticeship programs.
Speaker: Rishad Premji
Models designed for specific processes or decisions tend to deliver the most reliable results. When AI is closely aligned to a defined workflow, it becomes more predictable, easier to govern, and more effective over time.
Challenges the hype around “general‑purpose” AI by advocating for narrow, workflow‑centric solutions that are easier to manage and scale.
Shifts the conversation toward pragmatic implementation strategies, influencing subsequent emphasis on governance, risk management, and change management within enterprises.
Speaker: Rishad Premji
For AI to deliver value, organizations must invest in change – reskilling people, redesigning roles, and building confidence in how AI is used. AI fluency must extend beyond engineering.
Highlights the human and cultural dimension of AI adoption, emphasizing that technology alone is insufficient without workforce transformation.
Introduces a new layer of discussion about education, continuous learning, and the societal shift required, prompting the audience to think about policy and corporate responsibility.
Speaker: Rishad Premji
Our foundation’s pilot in rural Tamil Nadu uses portable X‑ray devices and AI to screen for tuberculosis at home, eliminating the need for patients to travel to distant hospitals.
Provides a vivid, personal example of AI delivering tangible health outcomes in a low‑resource setting, illustrating the concept of AI as a multiplier of scarce expertise.
Humanizes the earlier abstract arguments, reinforcing the message that AI can solve pressing public‑health challenges and can be exported to other low‑resource regions globally.
Speaker: Rishad Premji
India’s advantage in the AI era will be defined by the choices we make – where we apply AI, how responsibly we deploy it, and whether we can translate capability into real impact for governments, citizens, and enterprises.
Summarizes the strategic thesis that success depends less on size of models or compute and more on ethical, purposeful deployment decisions.
Serves as a concluding turning point, urging policymakers, industry leaders, and academia to align on responsible, impact‑driven AI strategies, thereby framing the call to action for the rest of the audience.
Speaker: Rishad Premji
Overall Assessment

Rishad Premji’s remarks steered the discussion from a generic celebration of AI’s potential to a nuanced roadmap for India’s AI future. By first redefining AI as a technology that reshapes societal duties, he set a reflective tone. Subsequent pivot points—highlighting the shift from possibility to practical adoption, showcasing India’s UPI success, emphasizing narrow workflow‑aligned models, and underscoring the necessity of reskilling—each introduced fresh dimensions that deepened the conversation. The personal TB‑screening story anchored abstract concepts in real‑world impact, while the final call to make responsible choices framed a strategic agenda. Collectively, these key comments redirected the audience’s focus toward implementation, governance, talent development, and ethical stewardship, shaping the dialogue into a forward‑looking, action‑oriented discourse.

Follow-up Questions
How can AI models be designed to align with specific enterprise workflows to ensure reliability and ease of governance?
Understanding workflow-aligned model design is crucial for predictable performance, regulatory compliance, and scalable adoption in complex organizations.
Speaker: Rishad Premji
What effective strategies and programs are needed to reskill employees and teams to work confidently with AI tools?
Reskilling is essential to enable staff to interpret AI outputs, exercise judgment, and ensure sustainable AI integration across the workforce.
Speaker: Rishad Premji
What measurable impact does AI have on improving learning outcomes in local languages within the education sector?
Quantifying educational benefits validates AI investments and guides policy for multilingual, inclusive learning solutions.
Speaker: Rishad Premji
How can AI be leveraged to reduce leakages and improve efficiency in welfare delivery and public services?
Assessing AI’s role in welfare systems can enhance transparency, reduce corruption, and ensure resources reach intended beneficiaries.
Speaker: Rishad Premji
What guardrails and governance frameworks are needed to balance accountability with innovation in AI deployment across India?
Establishing appropriate safeguards protects against misuse while fostering an environment conducive to rapid AI advancement.
Speaker: Rishad Premji
What are the optimal approaches to expand and democratize compute infrastructure to support AI development at national scale?
Adequate compute resources are foundational for training large models and enabling widespread AI adoption across sectors.
Speaker: Rishad Premji
Will the AI‑powered portable X‑ray screening pilot for tuberculosis in rural Tamil Nadu be effective and scalable nationwide?
Evaluating clinical outcomes and scalability determines if this model can address critical health gaps across India and similar contexts globally.
Speaker: Rishad Premji
Can AI solutions developed in India be adapted and deployed effectively in other low‑resource regions of Asia, Africa, and Latin America?
Testing transferability ensures that Indian innovations have broader global impact and inform cross‑regional collaboration.
Speaker: Rishad Premji
How can fragmented and siloed enterprise data be curated and labeled to create high‑quality, context‑aware AI models?
Effective data preparation is a prerequisite for building accurate, reliable AI systems that operate within complex organizational landscapes.
Speaker: Rishad Premji
What specific AI guardrails are required to earn the confidence of security teams, risk leaders, and regulators in large enterprises?
Gaining stakeholder trust is vital for widespread AI adoption and for mitigating security and compliance risks.
Speaker: Rishad Premji
How can AI fluency be extended beyond engineers to teachers, nurses, administrators, and small business owners?
Broadening AI literacy ensures that diverse user groups can effectively interact with AI tools, maximizing societal benefits.
Speaker: Rishad Premji
What are the long‑term effects of AI‑driven early pest alerts on crop yields and farmer incomes across different Indian states?
Understanding agricultural outcomes informs policy support for AI in farming and measures its contribution to food security.
Speaker: Rishad Premji
How does AI‑enabled product cataloguing and price optimisation impact market access and profitability for small artisans and sellers?
Assessing economic benefits for micro‑enterprises validates AI’s role in inclusive growth and rural commerce.
Speaker: Rishad Premji
How effective are government initiatives aimed at training 10 million youth in AI for building a skilled, job‑ready workforce?
Evaluating these programs helps refine education policy and ensures alignment with industry needs.
Speaker: Rishad Premji
What metrics and methodologies should be used to track the transition of AI projects from pilots to scaled, sustainable impact?
Clear measurement frameworks are needed to assess progress, justify investments, and guide strategic scaling decisions.
Speaker: Rishad Premji
What are the best practices for modernizing legacy architectures to support AI integration in large, complex enterprises?
Legacy system modernization is a prerequisite for deploying AI at scale without disrupting existing operations.
Speaker: Rishad Premji
How can AI be responsibly deployed to ensure inclusivity across India’s diverse linguistic, geographic, and socioeconomic contexts?
Inclusivity safeguards that AI benefits all population segments, preventing digital divides and bias.
Speaker: Rishad Premji
What role should Indian enterprises play in setting global standards for responsible AI development and deployment?
Leadership in standards can shape international norms, enhance competitiveness, and promote ethical AI worldwide.
Speaker: Rishad Premji
How can AI be used to strengthen rural healthcare delivery beyond tuberculosis detection, addressing broader health challenges?
Exploring wider health applications can amplify AI’s impact on public health outcomes in underserved areas.
Speaker: Rishad Premji
What mechanisms can ensure that AI‑driven decision‑making retains human judgment where it matters most?
Balancing automation with human oversight is critical to maintain accountability, trust, and ethical outcomes.
Speaker: Rishad Premji

Disclaimer: This is not an official session record. DiploAI generates these resources from audiovisual recordings, and they are presented as-is, including potential errors. Due to logistical challenges, such as discrepancies in audio/video or transcripts, names may be misspelled. We strive for accuracy to the best of our ability.

Keynote-Vishal Sikka

Session at a glanceSummary, keypoints, and speakers overview

Summary

The session began with Speaker 1 thanking Sir Hasabis and introducing Vishal Sikka, founder and CEO of VNI and former Infosys CEO, as a leading thinker at the intersection of AI and enterprise [1-10]. Sikka opened by noting that users who truly understand how to employ generative AI can achieve dramatic productivity gains, citing a former classmate who rebuilt a nine-month, 15-person service in just 14 days using an AI coding tool-a more than 250-fold improvement [15-20]. He further illustrated the impact with a customer in the home-goods distribution sector who, using his company’s AI product, reduced a decision that would have taken a year to a few days, highlighting AI’s power to provide instant, multilingual knowledge synthesis [20-22]. Sikka emphasized that such effectiveness is uneven, describing it as a “jagged frontier” where only those who grasp AI’s capabilities can reap its benefits [24-26].


His second point stressed that mastering AI requires not only technical knowledge but also awareness of its limits and the ability to bridge the gap between large language models and enterprise users through reliable, verifiable systems [27-31]. He argued that closing this gap can transform legacy systems, simplify complex enterprise processes, and empower end users, a vision supported by remarks from Mukesh Bhai earlier in the summit [33-38]. Sikka highlighted India’s abundant talent and the Prime Minister’s call for a billion entrepreneurs who can create value with AI, positioning the country as poised to lead this transformation [40-43].


The third point warned that today’s AI still suffers from serious shortcomings such as hallucinations, limited physical understanding, and safety risks, which must be addressed before broader enterprise adoption [44-56]. He compared AI’s energy demands to a 100-watt light bulb, noting that current models consume far more power than the human brain and that substantial efficiency gains are still required [62-66]. Sikka invoked historical Indian innovations-the Green Revolution and mass connectivity through Jio and Airtel-to illustrate how the nation has repeatedly leapfrogged challenges and can do so again with AI [70-73].


He concluded that the summit demonstrates a path toward a “human revolution” powered by purposeful, safe AI, where billions of entrepreneurs not only earn livelihoods but also enrich lives [74-75]. The discussion underscored both the extraordinary opportunities AI offers for productivity and societal impact, and the urgent need for responsible development, safety, and energy efficiency to realize those benefits [22-56]. Ultimately, Sikka’s message positioned AI as a catalyst for India’s next wave of innovation, calling for coordinated effort to master, improve, and responsibly deploy the technology for broader human benefit [68-75].


Keypoints

Major discussion points


AI can deliver dramatic productivity gains for skilled users.


Sikka cites a Stanford-classmate who rebuilt a nine-month, 15-person service in just 14 days with a generative-coding tool – a “more than 250 times improvement in productivity” – and a home-goods distributor who reduced a year-long country-exit analysis to a few days using AI-driven simulations. [15-20]


Bridging the gap between large language models and enterprise users is essential.


He stresses that effectiveness with AI requires not only technical know-how but also awareness of limitations; enterprises need “trusted, verifiable, reliable systems” that sit above raw LLMs to deliver correct business value. [27-32]


Current AI systems have critical limitations that must be solved before wide-scale adoption.


Sikka highlights hallucinations, safety risks (e.g., reckless autonomous agents), and massive energy consumption of large models, arguing that AI must become as safe and regulated as nuclear power and far more energy-efficient. [44-56][66-67]


India has a unique opportunity to lead a “human-centered” AI revolution.


He references the Prime Minister’s call for a “billion entrepreneurs,” India’s past successes (Green Revolution, telecom expansion), and the nation’s abundant talent as the foundation for building the next generation of AI that empowers people rather than merely automates tasks. [40-42][68-73]


Overall purpose / goal


The speaker’s aim is to motivate the audience-particularly Indian technologists, entrepreneurs, and policymakers-to harness AI’s unprecedented productivity, to responsibly close the gap between AI capabilities and business needs, to confront the technology’s safety and sustainability challenges, and ultimately to position India as a global leader in a purposeful, human-centric AI transformation.


Overall tone


Opening (0:00-2:00): Warm, appreciative, and celebratory, thanking the previous speaker and praising Vishal Sikka’s credentials.


Middle (2:00-7:00): Energetic and optimistic when describing productivity breakthroughs, then becomes more analytical and cautionary as he outlines the need for trustworthy systems and the “jagged frontier” of AI effectiveness.


Later (7:00-10:42): Shifts to a serious, almost urgent tone when discussing AI’s limitations-hallucinations, safety, and energy concerns-while still maintaining a visionary optimism about overcoming these hurdles.


Closing: Returns to an inspiring, hopeful tone, envisioning a “human revolution” powered by good AI and a billion Indian entrepreneurs, ending on a note of excitement and fun.


The tone evolves from gratitude to excitement, through sober analysis of risks, and culminates in a hopeful call to action.


Speakers

Vishal Sikka – Founder and CEO of VNI; former CEO of Infosys; computer scientist and AI thought leader. [S2]


Speaker 1 – Event moderator/host (role not specified). [S3]


Additional speakers:


– None identified.


Full session reportComprehensive analysis and detailed insights

The session opened with a formal note of appreciation. Speaker 1 thanked Sir Hasabis for his “profound and illuminating address” and expressed sincere gratitude before introducing the next speaker, Vishal Sikka, founder and CEO of VNI, a former Infosys chief who led “one of the most ambitious transformations in Indian IT history” and is described as “a computer scientist by training, a philosopher by temperament” and a leading thinker at the AI-enterprise intersection [1-10].


Sikka began by echoing the courteous tone, thanking the audience twice and calling the event “wonderful” [11-13]. He then outlined that his remarks would be organised around three main observations drawn from his long experience in artificial intelligence [14].


First observation – productivity gains


He recounted a Stanford classmate who rebuilt a nine-month, 15-engineer service in 14 days using a generative-coding tool, achieving >250× productivity [15-20]. A second illustration involved a home-goods distributor who, with VNI’s product, reduced a decision that would normally require a year of analysis to a few days, providing instant, multilingual knowledge synthesis [20-22]. These anecdotes underscore the tangible, measurable impact of AI when wielded by knowledgeable practitioners [15-22].


Second observation – bridging the LLM-enterprise gap


Sikka argued that effective AI use requires not only technical know-how but also awareness of large-language models (LLMs) limitations, and that enterprises need “trusted, verifiable, reliable systems” that sit above the LLMs to deliver correct business value [27-32]. VNI is developing such a layer that abstracts the underlying models while ensuring correctness and reliability [31-32]. He linked this bridging effort to broader transformation potential: by closing the gap, “legacy systems” and “enormous complexities inside enterprises” can be removed, allowing industries to be reshaped and giving end users wings and amplifying them [33-38].


Third observation – mastering AI’s limitations


Sikka identified “hallucinations” as a primary obstacle, noted the lack of physical-world understanding, and warned of safety risks such as reckless autonomous swarms [44-56]. He likened the need for AI safety to the decades-long regulatory regime governing nuclear power, arguing that comparable rigor is essential for responsible AI deployment [55-56].


Energy consumption was presented as another critical constraint. Describing a 720 MW data centre on California’s Highway 101, he compared the power draw of generative-AI inference to a 100-watt light bulb and highlighted that the human brain operates on merely 15-20 W, implying that “many zeros still need to be removed” from today’s models [62-66]. To illustrate the inefficiency, he told the “32 000-step / two-burger” story, showing how far AI’s energy use is from human efficiency [69].


Sikka then invoked a quotation from the Sanskrit scripture Bhaj Govindam – “Samprapte sanihite kale…” – to stress that “knowledge without wisdom does not save us” [57-58]. He framed this cultural insight as a reminder that practical wisdom must accompany technical mastery.


Turning to the Indian context, Sikka invoked the Prime Minister’s call for “a billion entrepreneurs” who can harness AI to create value, stressing that India possesses abundant talent and an entrepreneurial spirit suited to this challenge [40-43]. He recalled the Green Revolution, which turned India from a food-deficit nation into a major exporter within a generation, and cited the nation-wide connectivity achieved by Jio and Airtel, now providing “billion-plus Indians” with data and internet access [70-73]. These precedents reinforce his belief that AI can be the next leap-frog for the country.


In closing, Sikka painted a vision of a “human revolution powered by AI, by good AI, by purposeful AI,” where a billion entrepreneurs are not merely earning a livelihood but “making a life” for themselves and others [74-75]. He reaffirmed the excitement of this endeavour, thanked the audience, and ended on an upbeat note, inviting participants to join the journey toward a responsible, empowering AI future [76].


Session transcriptComplete transcript of the session
Speaker 1

Thank you so much, Sir Hasabis, for your very profound and illuminating address. We really thank you. Sincere gratitude to your address. Ladies and gentlemen, and now I would like to invite Mr. Vishal Sikka. He’s the founder and CEO of VNI. As CEO of Infosys, Vishal Sikka has led one of the most ambitious transformations in Indian IT history. Before leaving to build VNI, a company focused on human -centered artificial intelligence. He is a computer scientist by training, a philosopher by temperament. He is one of the most original thinkers of the intersection of AI and enterprise. Please welcome the founder and CEO of VNI, Mr. Vishal Sikka.

Vishal Sikka

Thank you so much. Thank you so much. Wow, wonderful introduction and what an amazing event. I want to share three points from a long time that I have spent in AI. My first point is that what we see today in the world of AI, people who know what they are doing with AI are astonishingly effective with AI. Recently, a friend of mine, he and I were students together at Stanford. He has a large service that he runs, open public service. That service was built by 15 people, very world -class engineers, over nine months. Recently, he rebuilt that service entirely by himself in 14 days using one of the generative AI coding tools. So if you are counting, that is about more than 250 times improvement in productivity.

now he’s a genius and not everyone gets a 250 times productivity gain but you will see that people who understand how to use ai are astonishingly effective with it and i had a similar experience recently with a customer of mine who is a distributor of home goods and one of their main suppliers shut down their factories in one of their countries and they did analysis using our product all kinds of simulation scenarios and over a few days they reached the decision that they need to exit that country entirely i asked him you know such a monumental decision to get out of an entire country how long would that have taken you before and he said easily it would take you a year to get out of that country and i said you know what i mean by that it would have taken a year and it would have involved heavy duty consultants and things like that So, we now have instant access to knowledge in any language, a condensation of things that we can present in any way.

It is an incredible power. And yes, it is deeply disruptive to the ways that we have worked before, the way that we have done things in the past. But at the same time, and even more importantly, we can do unprecedented things with this, things that we could never do before. So, this astonishing effectiveness, and Yoshua referred to this as a jagged frontier, it is not uniform. Not everyone sees this. So, that’s my second point. Being effective with AI requires not only a knowledge of AI itself, but understanding its limitations and how to overcome those limitations. There is a huge gap between LLMs. and the business users inside enterprises especially and how to bring value to those users.

And overcoming that gap is where a lot of value -creating opportunity is. Bridging that gap requires delivering correct systems, trusted, verifiable, reliable systems that deliver value to people. My own company works in this area, a layer that sits above the language models and delivers value to business users while ensuring correctness and things like that. When we overcome that gap, we can deliver massive value. Mukesh Bhai talked about it just now. We can transform every existing system, legacy systems, enormous complexities inside enterprises can be removed. Industries can be transformed. We can give end users wings. We can end users. We can amplify them. to deliver things that were not possible to do before, that or in the best case, it required professionals to do this.

Doing that also requires not just overcoming the limitations of AI, but also imagination to see what is not there, to see what is possible. India has all of this in great abundance. The Honorable Prime Minister has called for a billion entrepreneurs, people who can overcome these and deliver value using AI. And I think this is exactly what the world needs and what India has the potential for. My third point is that we not only have to master today’s AI, but we have to leapfrog it. AI today has enormous limitations. I have worked in AI for the last 38 years. You know… One of our scriptures is Bhaj Govindam. It was written by the Shankaracharya. And it has a beautiful line.

Samprapte sanihite kale, nahi nahi rakshati dukhrin karane. What it means is that when you are faced with a life or death situation, the knowledge of a book does not help you. Knowledge without wisdom does not save us. That wisdom comes from living, from doing, from being in the world. AI today has plenty of limitations. Yoshua mentioned hallucinations. That’s one of the main issues blocking the use of AI in enterprises. But beyond hallucinations, understanding the world, understanding physical activities, the physical movement, this is one of the next frontiers. Safety. safety of AI and the Honorable Prime Minister talked about this today is an existential issue swarms of agents can be made to do completely reckless things and we don’t yet have ways to understand or deal with this and Yoshua Demis also talked about this we have to solve this issue we have to deliver AI that is safe we have done this with nuclear power for the last 80 plus years we can and we must do this with AI energy is another one of these issues where I live in California on highway 101 just north of San Tomas I drive by there every time I go to see my dad there is a massive data center that is coming up it’s 720 megawatts and this idea that I write a prompt and these gazillion genes GPUs blast into existence to produce a response and then I make a tiny change to that prompt and then I do that again.

It just seems like a completely absurd idea, especially to someone who has been around AI for such a long time. I have, thanks to the minister, Ministry of IT, Honorable Ashwini Vasanthaji, I have a colleague who has been accompanying me throughout this conference and he told me that yesterday he walked 32 ,000 steps. And I asked him, what did you eat? And he said, I ate two burgers. Shubham, if you’re here. Two burgers. You know, we normally eat 2 ,000 calories in a day. That’s about a 100 -watt light bulb, like less than one of these light bulbs. And out of that, our brain, our nervous system is maybe 15, 20 watts. That’s like when your laptop is in sleep mode, it takes more power than that.

So, there are… many zeros still to be removed from these models, and the models themselves have to be removed. So, I think that there is a tremendous opportunity here. India is a country of the human potential. We have plenty of times before delivered the ability to, you know, billion plus Indians. Mukesh Bhai talked about Jio, and Sunil talked about Airtel, and how now we have billion plus Indians who have data and connectivity. When I was young, one of my earliest childhood memories is of worry in my parents’ faces around food. They used to tell stories of how there was a shortage of food, and then the green revolution happened, and India is now one of the largest exporters of food in one generation.

So, I think that when you look at the time of intelligence, it is not only an opportunity to learn about this technology, to learn to master this technology, to understand its limitations, but to leapfrog that and to build the next generation of it. And as this summit so vividly demonstrates, we can be on our way to a human revolution powered by AI, by good AI, by purposeful AI, where every one of us, a billion entrepreneurs, is not just making a living but is making a life, not some artificial life or some artificial general life, but our own life and the life of others. And that would be so much fun. Fun to do. Thank you so much.

Related ResourcesKnowledge base sources related to the discussion topics (30)
Factual NotesClaims verified against the Diplo knowledge base (3)
Confirmedhigh

“Vishal Sikka is the founder and CEO of VNI and a former Infosys chief who led one of the most ambitious transformations in Indian IT history”

The knowledge base records that Vishal Sikka is the founder and CEO of VNI, former CEO of Infosys, and that he led one of the most ambitious transformations in Indian IT history [S1] and [S2].

Confirmedmedium

“Sikka is described as “a computer scientist by training, a philosopher by temperament””

The moderator introduced Sikka with exactly those descriptors in the source material [S2].

Confirmedmedium

“Sikka delivered a keynote address on artificial intelligence’s transformative potential and challenges”

The knowledge base notes that Sikka gave a keynote on AI’s transformative potential and challenges, matching the report’s description of his talk [S2].

External Sources (77)
S1
Keynote-Vishal Sikka — -Honorable Ashwini Vasanthaji: Role/Title: Minister, Ministry of IT; Area of expertise: Information Technology -Moderat…
S2
Keynote-Vishal Sikka — Vishal Sikka, founder and CEO of VNI and former CEO of Infosys, delivered a keynote address on artificial intelligence’s…
S3
Keynote-Martin Schroeter — -Speaker 1: Role/Title: Not specified, Area of expertise: Not specified (appears to be an event moderator or host introd…
S4
Responsible AI for Children Safe Playful and Empowering Learning — -Speaker 1: Role/title not specified – appears to be a student or child participant in educational videos/demonstrations…
S5
Building Trusted AI at Scale Cities Startups &amp; Digital Sovereignty – Keynote Vijay Shekar Sharma Paytm — -Speaker 1: Role/Title: Not mentioned, Area of expertise: Not mentioned (appears to be an event host or moderator introd…
S6
Masterclass#1 — Gratitude was expressed towards both presenters and participants for engaging in the dialogue. The speaker expressed gr…
S7
Any other business /Adoption of the report/ Closure of the session — Looking ahead, the speaker proposed two rounds of informal consultations before the next meeting in New York, showing a …
S8
9821st meeting — Republic of Korea:Thank you, Mr. President. I’d like to begin by expressing my gratitude to the United States for conven…
S9
Scaling AI for Billions_ Building Digital Public Infrastructure — Rather than focusing on individual LLMs, enterprises need a comprehensive AI operating system that combines knowledge fr…
S10
HETEROGENEOUS COMPUTE FOR DEMOCRATIZING ACCESS TO AI — Arun Shetty made a crucial distinction between safety and security concerns in AI systems. Safety issues involve models …
S11
Responsible AI in India Leadership Ethics &amp; Global Impact part1_2 — Excellent. And given the kind of scale you’re operating at, I think every day is a new day. Yes, it is. We face challeng…
S12
Comprehensive Report: Preventing Jobless Growth in the Age of AI — -Benefit distribution mechanisms: What specific policies should redistribute productivity gains from capital to workers …
S13
World in Numbers: Jobs and Tasks / DAVOS 2025 — Andrew Ng predicts that AI tools will lead to a substantial increase in employee productivity. He suggests that this inc…
S14
AI drives productivity surge in certain industries, report shows — A recent PwC (PricewaterhouseCoopers International Limited) reporthighlightsthat sectors of the global economy with high…
S15
Open Forum #3 Cyberdefense and AI in Developing Economies — Philipp Grabensee: you know, follows up on the session you had in Riyadh and I think we all agreed that the bottleneck i…
S16
What policy levers can bridge the AI divide? — Lacina Kone: Before talking about the bridging of AI, bridging the gap of the AI, there are gaps already, digital gap. Y…
S17
[Parliamentary Session 3] Researching at the frontier: Insights from the private sector in developing large-scale AI systems — Ammari highlighted META’s open-source approach to large language models, explaining, “META has adopted an open source me…
S18
Large Language Models on the Web: Anticipating the challenge | IGF 2023 WS #217 — Emily Bender:Yeah, so on a slightly different topic, I want to say that all of these discussions become clearer if we st…
S19
Global AI Policy Framework: International Cooperation and Historical Perspectives — Werner identifies three critical barriers that prevent AI for good use cases from scaling globally. He emphasizes that d…
S20
Inclusive AI For A Better World, Through Cross-Cultural And Multi-Generational Dialogue — Factors such as restricted access to computing resources and data further impede policy efficacy. Nevertheless, the cont…
S21
From principles to practice: Governing advanced AI in action — Several critical issues remain unresolved:
S22
How Trust and Safety Drive Innovation and Sustainable Growth — This insight identifies a critical gap in current regulatory approaches – that AI creates an ‘enforcement invisibility’ …
S23
Indias Roadmap to an AGI-Enabled Future — Chariot’s focus on voice-native AI models exemplifies this strategy. India is fundamentally a voice-first country, with …
S24
Keynote_ 2030 – The Rise of an AI Storytelling Civilization _ India AI Impact Summit — “The first is the fact that we have demographic energy.”[27]”This is certainly a category where India can lead and show …
S25
HETEROGENEOUS COMPUTE FOR DEMOCRATIZING ACCESS TO AI — India’s unique position—combining technical talent, diverse datasets, a vibrant startup ecosystem, and supportive policy…
S26
Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session — Shree, Prime Minister Modiji, congratulations on an amazing conference, and thank you for your leadership. I think focus…
S27
Opening Remarks (50th IFDT) — The overall tone was formal yet warm and celebratory. Speakers expressed pride in the IFDT’s accomplishments and gratitu…
S28
Open Mic &amp; Closing Ceremony — The overall tone was formal yet appreciative. There was a sense of accomplishment and gratitude expressed throughout, wi…
S29
First round of informal consultations with member states, observers and stakeholders (2024) — The blend of diplomacy, multilateral cooperation, and procedural lucidity indicates a positive direction for internation…
S30
Masterclass#1 — Gratitude was expressed towards both presenters and participants for engaging in the dialogue. The speaker expressed gr…
S31
OPEN MIC – Taking Stock | IGF 2023 — Furthermore, Mitchell expresses his deep appreciation for the active participation of all individuals in the IGF and ext…
S32
AI/Gen AI for the Global Goals — Shea Gopaul: So thank you, Sanda. And like Sandra, I’d like to thank the African Union, as well as Global Compact. i…
S33
World in Numbers: Jobs and Tasks / DAVOS 2025 — Andrew Ng predicts that AI tools will lead to a substantial increase in employee productivity. He suggests that this inc…
S34
AI drives productivity surge in certain industries, report shows — A recent PwC (PricewaterhouseCoopers International Limited) reporthighlightsthat sectors of the global economy with high…
S35
Shaping AI to ensure Respect for Human Rights and Democracy | IGF 2023 Day 0 Event #51 — Artificial Intelligence (AI) carries the potential to revolutionise various sectors worldwide, due to its capacities for…
S36
Comprehensive Report: Preventing Jobless Growth in the Age of AI — Economic | Future of work Study of LLMs in call centers showing 14% average increase in productivity, up to 35%. Studie…
S37
Keynote-Vishal Sikka — And overcoming that gap is where a lot of value -creating opportunity is. Bridging that gap requires delivering correct …
S38
Keynote-Vishal Sikka — Sikka’s second point addressed the substantial gap between the raw capabilities of large language models and the practic…
S39
WS #187 Bridging Internet AI Governance From Theory to Practice — Vint Cerf: Well, thank you so much for this opportunity. I want to remind everyone that I am not an expert on artificial…
S40
From principles to practice: Governing advanced AI in action — Several critical issues remain unresolved:
S41
Transforming Agriculture_ AI for Resilient and Inclusive Food Systems — Despite promising developments, several critical challenges remain. The lack of adequate data sharing mechanisms, partic…
S42
Global AI Policy Framework: International Cooperation and Historical Perspectives — Werner identifies three critical barriers that prevent AI for good use cases from scaling globally. He emphasizes that d…
S43
AI Development Beyond Scaling: Panel Discussion Report — – Yoshua Bengio- Yejin Choi Infrastructure | Cybersecurity Notes that removing one machine from a cluster can crash th…
S44
Importance of Professional standards for AI development and testing — Several critical issues remained unresolved:
S45
Leaders’ Plenary | Global Vision for AI Impact and Governance- Afternoon Session — Shree, Prime Minister Modiji, congratulations on an amazing conference, and thank you for your leadership. I think focus…
S46
Keynote_ 2030 – The Rise of an AI Storytelling Civilization _ India AI Impact Summit — “The first is the fact that we have demographic energy.”[27]”This is certainly a category where India can lead and show …
S47
Indias Roadmap to an AGI-Enabled Future — Chariot’s focus on voice-native AI models exemplifies this strategy. India is fundamentally a voice-first country, with …
S48
HETEROGENEOUS COMPUTE FOR DEMOCRATIZING ACCESS TO AI — India’s unique position—combining technical talent, diverse datasets, a vibrant startup ecosystem, and supportive policy…
S49
Opening and introduction — The AU’s commitment to working with Member States in adopting the meeting’s recommendations was reaffirmed, alongside th…
S50
Bridging the AI innovation gap — The tone is consistently inspirational and collaborative throughout. The speaker maintains an optimistic, forward-lookin…
S51
Leaders’ Plenary | Global Vision for AI Impact and Governance Morning Session Part 1 — A breakthrough with immense promise but serious risks in the wrong hands. Democracy rests not on the rule of the most le…
S52
From summer disillusionment to autumn clarity: Ten lessons for AI — As we refocus on existing risks, some accountability is due:how and why did respected voices get carried away with AGI p…
S53
Powering AI _ Global Leaders Session _ AI Impact Summit India Part 2 — The discussion maintained a predominantly optimistic and forward-looking tone throughout, despite acknowledging signific…
S54
Artificial General Intelligence and the Future of Responsible Governance — The discussion maintained a serious, analytical tone throughout, characterized by cautious optimism mixed with genuine c…
S55
Keynote-Lars Reger — The tone is enthusiastic and visionary throughout, with Reger maintaining an optimistic, forward-looking perspective. He…
S56
AI in Mobility_ Accelerating the Next Era of Intelligent Transport — The discussion maintained a serious, urgent tone throughout, driven by the gravity of India’s road safety crisis. While …
S57
AI Innovation in India — The tone was consistently celebratory, inspirational, and optimistic throughout the discussion. Speakers expressed pride…
S58
Keynote-Nikesh Arora — Overall Tone:The tone begins optimistically, celebrating AI’s rapid progress and potential, then shifts to a more cautio…
S59
Keynote_ 2030 – The Rise of an AI Storytelling Civilization _ India AI Impact Summit — Overall Tone:The tone is consistently optimistic, visionary, and inspirational throughout. The speaker maintains an enth…
S60
AI in education: Leveraging technology for human potential — The tone is consistently optimistic and inspirational throughout, with Mills maintaining an enthusiastic and visionary a…
S61
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Hemant Taneja General Catalyst — Overall Tone:The tone is consistently optimistic, inspirational, and forward-looking throughout the speech. The speaker …
S62
World Economic Forum Annual Meeting Closing Remarks: Summary — The tone is consistently positive, celebratory, and grateful throughout the discussion. It begins with formal appreciati…
S63
Ad Hoc Consultation: Thursday 1st February, Morning session — In a formal statement, the representative from the delegation took the opportunity to express their appreciation to the …
S64
Opening of the session — The delegation commenced by expressing their sincere regards and thanks to the Chairman and their team for their diligen…
S65
Enhancing the digital infrastructure for all | IGF 2023 Open Forum #135 — The individual’s gestures and words demonstrated a gracious and courteous exit, leaving a positive impression on the aud…
S66
Open Microphone Taking Stock — The tone was largely positive and appreciative, with many speakers thanking the hosts and expressing enthusiasm for the …
S67
Centering People and Planet in the WSIS+20 and beyond — Guilherme de Aguiar Patriota from Brazil articulated three key priorities reflecting the Global South perspective. First…
S68
Shaping the Future AI Strategies for Jobs and Economic Development — Brazil’s Tech Ambassador Eugenio Vargas Garcia emphasized the importance of “tech diplomacy” for Global South countries….
S69
Governments, Rewired / Davos 2025 — Cina Lawson: So when the pandemic started, we wanted to implement mobility restriction measures like everybody else. B…
S70
Policy Network on Artificial Intelligence | IGF 2023 — Moderator – Prateek:Thanks. Thanks, Shamira. So we have three questions. Two are quite similar. But I’d request you to h…
S71
https://app.faicon.ai/ai-impact-summit-2026/keynote-rishad-premji — We are already seeing what this looks like on the ground. farmers across Karnataka, Maharashtra, Telangana, AP and Punja…
S72
FOREWORDS — FIGURE 48: Product illustration and user journey
S73
The mismatch between public fear of AI and its measured impact — Inknowledge work, AI tools help draft emails, summarize reports, or generate first drafts of text. This can save time, b…
S74
ElevenLabs Voice AI Session & NCRB/NPMFireside Chat — Impact:This anecdote shifted the discussion from theoretical benefits to tangible human impact, making the technology’s …
S75
The impact of AI on jobs and workforce — The ILO’s webinar was triggered by the recent impact of ChatGPT on our society and jobs. OpenAI’s ChatGPT, in particular…
S76
AI for Good Technology That Empowers People — Impact:This comment provided a practical methodology that other panelists could reference and build upon. It shifted the…
S77
Is the AI bubble about to burst? Five causes and five scenarios — Behind the diminishing returns are conceptual and logical limitations of Large Language Models (LLMs), which cannot be r…
Speakers Analysis
Detailed breakdown of each speaker’s arguments and positions
S
Speaker 1
1 argument128 words per minute110 words51 seconds
Argument 1
Expressed sincere gratitude to the preceding address (Speaker 1)
EXPLANATION
Speaker 1 publicly thanked the previous presenter, acknowledging the value of the address and showing appreciation to the audience. The remarks serve to maintain a courteous and respectful tone for the event.
EVIDENCE
Speaker 1 opened by saying “Thank you so much, Sir Hasabis, for your very profound and illuminating address,” followed by additional expressions of thanks and gratitude before introducing the next speaker [1-4].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Expressions of gratitude toward previous presenters are recorded in multiple session transcripts, e.g., S6, S7 and S8 all note speakers thanking the prior address and participants.
MAJOR DISCUSSION POINT
Gratitude to previous speaker
AGREED WITH
Vishal Sikka
V
Vishal Sikka
3 arguments134 words per minute1329 words592 seconds
Argument 1
Knowledgeable users can achieve massive productivity improvements (e.g., 250× faster coding) (Vishal Sikka)
EXPLANATION
Sikka argues that individuals who understand how to leverage generative AI can dramatically increase their output, citing examples where AI‑assisted work outpaces traditional methods by orders of magnitude. This effectiveness is presented as a key advantage of AI adoption.
EVIDENCE
He recounts a former Stanford classmate who rebuilt a service that originally took nine months and a 15-person engineering team in just 14 days using a generative-AI coding tool, representing more than a 250-fold productivity boost. He also describes a home-goods distributor who, using their AI product, reduced a country-exit decision from a year to a few days [16-20].
MAJOR DISCUSSION POINT
AI productivity gains
Argument 2
Necessity of building trustworthy, verifiable layers atop LLMs to close the gap between AI models and business users (Vishal Sikka)
EXPLANATION
Sikka stresses that raw large language models are insufficient for enterprise needs; a reliable, verifiable middleware is required to translate model outputs into trustworthy business value. Closing this gap creates significant commercial opportunities.
EVIDENCE
He notes that effectiveness with AI demands both knowledge of AI and awareness of its limits, highlighting a large gap between LLM capabilities and enterprise users. He argues that delivering correct, trusted, and reliable systems is essential, and describes his own company’s work on a layer that sits above language models to ensure correctness for business users [27-33].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
The concept of a trusted, verifiable middleware layer above large language models is described in S1 and further elaborated as part of an AI operating system in S9.
MAJOR DISCUSSION POINT
Trusted AI layers for enterprise
Argument 3
Highlighted AI safety challenges (hallucinations, reckless agent behavior), excessive energy consumption, and the need to create next‑generation, responsible AI systems (Vishal Sikka)
EXPLANATION
Sikka points out critical shortcomings of current AI, such as hallucinations and unsafe autonomous behavior, and draws attention to the massive energy demands of large models. He calls for the development of safer, more efficient, and next‑generation AI architectures.
EVIDENCE
He references Yoshua Bengio’s concerns about hallucinations, warns about the risk of reckless AI agents, and describes the enormous power usage of data-center-scale AI models (e.g., a 720 MW facility powering GPUs for inference) as evidence of unsustainable energy consumption [54-57].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
Safety issues such as hallucinations and risky agent behavior are discussed in S2, while S10 distinguishes safety from security concerns, underscoring the need for responsible AI design.
MAJOR DISCUSSION POINT
AI safety and sustainability
Agreements
Agreement Points
Both speakers opened their contributions with expressions of gratitude and appreciation for previous participants
Speakers: Speaker 1, Vishal Sikka
Expressed sincere gratitude to the preceding address (Speaker 1)
Speaker 1 thanked Sir Hasabis and the audience before introducing the next speaker [1-4], and Vishal Sikka began his remarks by thanking the audience twice [11-13], indicating a shared courteous tone at the start of the session.
POLICY CONTEXT (KNOWLEDGE BASE)
Expressing gratitude at the start of contributions is a recognized diplomatic norm that underscores inclusive and partnership-oriented dialogue; recent multilateral consultations highlighted such acknowledgments as part of procedural lucidity and collaborative spirit [S29], a masterclass on capacity building noted similar gratitude to prior presenters and participants [S30], and the IGF 2023 opening remarks emphasized appreciation for all contributors [S31].
Similar Viewpoints
Unexpected Consensus
Overall Assessment

The only clear point of agreement between the two speakers is the mutual expression of thanks at the outset of the session; substantive AI‑related arguments are presented solely by Vishal Sikka with no corresponding statements from Speaker 1.

Low consensus on policy or technical issues; the limited agreement is limited to procedural courtesy, suggesting that the discussion did not produce substantive shared positions on AI, safety, productivity, or broader development themes.

Differences
Different Viewpoints
Unexpected Differences
Overall Assessment

The transcript shows no substantive disagreement between the speakers. Speaker 1’s remarks are limited to gratitude and the formal introduction of Vishal Sikka ([1-4]), while Sikka’s extensive remarks focus on AI productivity gains, the need for trustworthy layers, safety and sustainability concerns ([14-24][27-33][54-57]). Their statements are complementary rather than contradictory, indicating alignment on the relevance and potential of AI for India.

Minimal – the interaction is largely harmonious, implying consensus on the importance of AI and its potential impact, which suggests a unified stance for the topics under discussion.

Takeaways
Key takeaways
AI can deliver massive productivity gains for users who understand how to leverage it (e.g., 250× faster coding). Realizing enterprise value requires bridging the gap between raw LLMs and business users by adding trustworthy, verifiable layers that ensure correctness and reliability. Current AI systems have critical limitations—hallucinations, safety risks, and high energy consumption—that must be addressed to create next‑generation, responsible AI. India possesses a large pool of talent and entrepreneurial potential that can drive the development and deployment of safe, purposeful AI at scale. Wisdom and practical experience are essential; knowledge alone (e.g., from models) is insufficient for safe and effective AI use.
Resolutions and action items
None identified
Unresolved issues
How to effectively mitigate hallucinations and other safety risks in AI agents deployed in enterprises. Strategies for reducing the massive energy consumption of large language models and making AI more sustainable. Specific approaches for building and standardising the trustworthy, verifiable layers that sit above LLMs. Mechanisms to scale AI literacy and expertise across the broader workforce to achieve the cited productivity gains. Policy or regulatory frameworks needed to ensure AI safety and responsible deployment at national scale.
Suggested compromises
None identified
Thought Provoking Comments
A friend rebuilt a service that originally took 15 engineers nine months to build, in just 14 days using a generative AI coding tool – a more than 250‑times productivity gain.
It provides a concrete, dramatic illustration of how AI can amplify human effort, moving the conversation from abstract hype to measurable impact.
This anecdote set the stage for the rest of the talk, prompting the audience to consider real‑world efficiency gains and opening the floor to discussions about who can achieve such gains and why the benefit is ‘jagged’ across users.
Speaker: Vishal Sikka
“People who understand how to use AI are astonishingly effective with it… the effectiveness is a jagged frontier – not uniform.”
It challenges the assumption that AI benefits everyone equally and introduces the idea that skill and understanding create a divide.
This observation shifted the tone from celebration to a more nuanced view, leading to the next point about bridging the gap between LLMs and business users and prompting listeners to think about education and tooling.
Speaker: Vishal Sikka
“Bridging that gap requires delivering correct, trusted, verifiable, reliable systems… a layer that sits above the language models and delivers value to business users while ensuring correctness.”
It moves the discussion from productivity to governance, emphasizing the need for safety, trust, and verification in enterprise AI deployments.
This comment introduced a new topic—AI reliability and trustworthiness—causing the audience to consider regulatory, ethical, and technical safeguards, and it linked directly to earlier remarks about the jagged frontier.
Speaker: Vishal Sikka
Quote from Bhaj Govindam: “Knowledge without wisdom does not save us,” used to argue that AI’s current knowledge‑only approach is insufficient.
By invoking a centuries‑old philosophical text, Sikka reframes the AI debate as a moral and experiential challenge, not just a technical one.
This cultural reference deepened the conversation, prompting listeners to reflect on the limits of data‑driven AI and the necessity of human judgment, thereby broadening the scope beyond engineering concerns.
Speaker: Vishal Sikka
“AI safety is an existential issue… we have done this with nuclear power for 80 years; we must do it with AI.”
It draws a powerful parallel between AI and nuclear energy, highlighting the gravity of safety and control, and challenges the audience to treat AI with comparable rigor.
This statement pivoted the discussion toward risk management, encouraging a shift from optimism to caution and setting up later points about hallucinations, swarm agents, and regulatory responsibility.
Speaker: Vishal Sikka
Energy analogy: a 720 MW data center versus a human brain’s 15‑20 W, illustrating that AI models consume “many zeros” of power compared to biological intelligence.
It quantifies the environmental and scalability challenges of current AI models, introducing sustainability as a critical dimension of the AI conversation.
The comment opened a new line of thought about the ecological footprint of AI, prompting the audience to consider efficiency improvements as part of responsible AI development.
Speaker: Vishal Sikka
Historical parallel: “When I was young, my parents worried about food; the Green Revolution turned India into a major food exporter in one generation. AI can be the next leap‑frog for India.”
It connects past national transformation to present AI opportunities, framing AI as a catalyst for a new ‘human revolution’ and invoking national pride.
This analogy shifted the tone toward a visionary, patriotic call to action, aligning the technical discussion with broader socio‑economic goals and inspiring the audience to envision large‑scale impact.
Speaker: Vishal Sikka
Visionary closing: “A human revolution powered by good, purposeful AI where a billion entrepreneurs are not just making a living but making a life.”
It synthesizes earlier points—productivity, trust, safety, sustainability, and societal uplift—into an aspirational narrative that reframes AI as a tool for human flourishing.
This concluding remark reinforced the earlier themes, left the audience with a hopeful yet responsible outlook, and served as a rallying cry that could shape subsequent policy and entrepreneurial initiatives.
Speaker: Vishal Sikka
Overall Assessment

Vishal Sikka’s remarks steered the discussion from a surface‑level celebration of AI capabilities to a layered exploration of productivity, inequality, trust, safety, sustainability, and national ambition. Each pivotal comment introduced a fresh dimension—whether technical (250× productivity), philosophical (knowledge vs. wisdom), regulatory (AI safety akin to nuclear power), or societal (leap‑frogging like the Green Revolution). These insights acted as turning points that redirected attention, deepened analysis, and broadened the conversation’s scope, ultimately shaping the dialogue into a balanced narrative that combined optimism with caution and positioned AI as a catalyst for a large‑scale human transformation.

Follow-up Questions
How can enterprises bridge the gap between large language models and business users to deliver correct, trusted, verifiable, and reliable AI systems?
Sikka highlights a huge gap that limits value creation; closing it is essential for widespread, trustworthy AI adoption in enterprises.
Speaker: Vishal Sikka
What methods can be developed to mitigate hallucinations in large language models and ensure safe, reliable outputs for enterprise use?
He notes hallucinations as a major barrier to enterprise adoption and calls for research into techniques that reduce or eliminate them.
Speaker: Vishal Sikka
How can AI be advanced to understand and reason about the physical world and physical activities, moving beyond purely textual knowledge?
Sikka identifies understanding physical actions as a next frontier, crucial for safety and broader applicability of AI systems.
Speaker: Vishal Sikka
What safety frameworks, analogous to those used in nuclear power, can be established to manage existential risks posed by AI, especially with autonomous swarms of agents?
He draws a parallel to nuclear safety, emphasizing the need for robust safeguards to prevent reckless AI behavior.
Speaker: Vishal Sikka
How can the energy consumption of AI models be dramatically reduced (e.g., removing unnecessary parameters or ‘zeros’) to make AI more sustainable and cost‑effective?
Sikka points out the massive power draw of current AI models and calls for research into more efficient architectures.
Speaker: Vishal Sikka
What strategies can enable India to cultivate a billion AI‑empowered entrepreneurs who can create purposeful, human‑centric AI solutions?
He envisions leveraging India’s talent pool to drive a human revolution powered by AI, suggesting a need for ecosystem‑building research.
Speaker: Vishal Sikka

Disclaimer: This is not an official session record. DiploAI generates these resources from audiovisual recordings, and they are presented as-is, including potential errors. Due to logistical challenges, such as discrepancies in audio/video or transcripts, names may be misspelled. We strive for accuracy to the best of our ability.

Fireside Conversation: 02

Session at a glanceSummary, keypoints, and speakers overview

Summary

The panel, moderated by Maria Shakil, featured Yann LeCun, who described the next generation of AI as an “amplifier for human intelligence” rather than a fully autonomous super-mind that will dominate all domains [14-17]. He suggested that while a future entity might eventually surpass human capability, the more immediate goal is to create systems that extend human reasoning and accelerate progress [16-17].


LeCun argued that the historical notion of “genius” has shifted from practical inventions such as agriculture to today’s theoretical and abstract achievements, and that AI will further evolve this concept [20-24]. He warned that large language models are often mistaken for true intelligence because they mainly function as advanced information-retrieval tools, compressing existing knowledge without genuine reasoning [27-33][35-38]. According to LeCun, true intelligence requires “world models” that let agents predict and plan in continuous, noisy environments-a capability current AI, including LLMs, lacks [41-42].


Economists estimate that AI could raise productivity by about 0.6 % per year, a modest but significant boost that could accelerate scientific and medical advances, though the distribution of these gains remains a political question [45-52][55]. LeCun emphasized that the long-term source of AI innovation will be countries with favorable demographics such as India and Africa, and that this requires substantial investment in higher education and PhD-level training [90-98][100-105].


He noted that for AI to be truly accessible in a nation of 1.4 billion people, the cost of inference must fall dramatically, especially the energy expenses that currently limit widespread deployment [108-112]. Demonstrations such as smart-glass assistants for Indian farmers illustrate how AI can improve agriculture by diagnosing plant diseases and advising on harvest timing [118-120][122-124]. LeCun likened the societal impact of AI to the printing press and the Internet, arguing that it will broaden knowledge access and make people more informed if deployed responsibly [131-136].


He rejected the idea of a single breakthrough event, describing AI progress as a continuous, incremental process that cannot be measured by a single test because intelligence is a collection of rapidly learnable skills [58-66][70-71]. Reflecting on past hype cycles, LeCun said that predictions of human-level AI within a few years have repeatedly failed, and that the remaining gap-such as teaching a robot to drive after only 20 hours of practice-shows the challenge ahead [149-158][162-165]. Ultimately, he concluded that humans will continue to set the agenda for AI development, and that building systems capable of handling the messy, real world is the key challenge for the next decade [168-181].


Keypoints

Major discussion points


AI as an amplifier of human intelligence, not a replacement – LeCun stresses that the most valuable AI we are building is a tool that extends human capabilities rather than an autonomous super-intelligence. He describes large language models (LLMs) as powerful information-retrieval systems that lack true “world models” needed for reasoning and interaction with the physical world [16-17][27-34][41-42][58-66].


Evolving notions of genius and intelligence – The conversation explores how historic definitions of “genius” (e.g., agricultural innovators) differ from today’s emphasis on theoretical invention, and how AI will further reshape what we consider intelligent or brilliant [20-24][68-71].


Economic impact and the question of abundance – LeCun cites economists who estimate AI will raise productivity by roughly 0.6 % per year, a modest but significant boost. He warns that the distribution of any resulting wealth is a political issue, not a technical one [45-53][57-64].


Education, upskilling, and democratizing AI for the Global South – AI will become “staff” that managers (academics, politicians, business leaders) work with, demanding higher-level talent. LeCun highlights the need for massive investment in education, especially in countries like India and Africa, and notes that current inference costs and energy consumption must fall for AI to be broadly accessible [75-84][90-105][108-113][125-136].


Realistic timeline and technical hurdles – The panel agrees that AI progress will be incremental, not a single breakthrough event. Past hype cycles have repeatedly over-promised rapid arrival of human-level AI. Key challenges include building continuous-world models and overcoming the Moravec paradox (the gap between symbolic language tasks and real-world sensorimotor skills) [58-66][149-165][168-178].


Overall purpose / goal of the discussion


The dialogue aims to provide a balanced, forward-looking assessment of artificial intelligence: highlighting its role as a catalyst for human productivity and knowledge dissemination, examining how it reshapes concepts of intelligence and genius, evaluating economic and societal implications, stressing the urgency of education and equitable access-especially for emerging economies-and grounding expectations in the technical realities of current research.


Overall tone and its evolution


– The conversation opens with an enthusiastic and optimistic tone, celebrating LeCun’s contributions and the promise of AI as an “amplifier” [1-8][14-17].


– It then shifts to a nuanced, analytical tone, dissecting misconceptions about LLMs, the need for world models, and the modest economic gains [27-34][45-53].


– As the discussion moves toward policy, education, and global equity, the tone becomes pragmatic and advisory, emphasizing concrete challenges such as inference cost and the necessity of upskilling [90-105][108-113].


– Finally, the tone settles into cautious optimism, acknowledging past over-hype, outlining realistic timelines, and expressing confidence that societies will eventually harness AI responsibly [149-165][168-178].


Overall, the exchange balances optimism about AI’s transformative potential with a sober appraisal of the technical, economic, and societal hurdles that must be addressed.


Speakers

Yann LeCun


– Role/Title: Executive Chairman, Advanced Machine Intelligence Labs; former Chief AI Scientist at Meta


– Area of Expertise: Deep learning, artificial intelligence research, world-model AI


– Source: [S1]


Speaker 1


– Role/Title: Event host / opening presenter (no specific title given)


– Area of Expertise:


Maria Shakil


– Role/Title: Managing Editor, India Today


– Area of Expertise: Journalism, media, AI policy and industry coverage


– Source: [S7]


Additional speakers:


Brad Smith – mentioned (no speaking role); known as President and Vice Chair of Microsoft (role not cited).


Prime Minister Narendra Modi – mentioned (no speaking role); Prime Minister of India.


Philippe Ackermann – mentioned (no speaking role); economist.


Jung – mentioned (no speaking role); economist.


Eric Brynjolfsson – mentioned (no speaking role); economist.


Hans Moravec – mentioned (no speaking role); roboticist known for the Moravec paradox.


Full session reportComprehensive analysis and detailed insights

Speaker 1 opened the session by thanking Brad Smith for his “energising address” and then introducing the next guest as “the godfather of deep learning”, Yann LeCun, Executive Chairman of Advanced Machine Intelligence Labs [179][1-8]. The conversation was moderated by Maria Shakil, Managing Editor of India Today [9].


LeCun began by tempering expectations of a near-term “super-mind”. He said that while a truly smartest entity might appear within the lifetime of some audience members, it is unlikely to happen in his own lifetime and will take “a while” [14-15]. He framed the immediate goal of AI as building an “amplifier for human intelligence” that accelerates progress rather than an autonomous entity that surpasses humans across all domains [16-17].


When asked about the evolving definition of genius [180], LeCun traced the term back to ancient practical achievements such as crop cultivation and animal domestication, noting that historic genius was tied to tangible inventions [20-23]. He suggested that today’s more abstract notion of genius may continue to evolve as AI changes how creation and invention are understood [24].


The moderator highlighted a common view that AI is “powerful but not intelligent”. LeCun agreed, explaining that large language models (LLMs) are essentially sophisticated information-retrieval systems that compress and provide rapid access to human-produced knowledge, likening them to a modern evolution of the printing press, libraries, the Internet and search engines [181-182][27-34]. Although LLMs excel in certain domains such as code generation or limited mathematical reasoning, they remain largely symbolic manipulators and lack genuine reasoning [35-38].


A key limitation, LeCun argued, is the absence of “world models” [183-184][41-42]. He described how babies and animals learn by observing and interacting with the physical world, forming mental models that allow them to anticipate novel situations and plan actions. Current AI, including LLMs, does not build such models, which explains why systems can pass exams yet fail to master embodied tasks like self-driving cars after only a few hours of practice [45-48][164-165].


Turning to the macro-economic picture, LeCun cited economists such as Philippe Ackermann and Erik Brynjolfsson who estimate AI will raise productivity by roughly 0.6 % per year [45-46]. He stressed that this modest boost can nevertheless accelerate scientific and medical progress, but warned that the distribution of any resulting wealth is a political question, not a technical one [51-55][57-58].


LeCun emphasized that AI development will be continuous rather than a single breakthrough event and that the real issue is ensuring policies allow the benefits to be shared broadly [58-71]. He questioned the usefulness of the term “artificial general intelligence”, noting that human intelligence is highly specialised and that true intelligence should be measured by the ability to learn new tasks quickly and perform them without prior training [61-71].


He portrayed AI as “our staff”, with every professional becoming a manager of intelligent machines that may be smarter than their human supervisors [75-84]. This metaphor underscores the need for a highly skilled workforce. LeCun highlighted that future AI innovation will likely emerge from demographically favourable regions such as India and Africa, provided they invest heavily in youth education and PhD-level training [90-105][95-104]. He rejected the myth that AI will eliminate the need for study, insisting that the demand for advanced scientists is growing worldwide [97-104].


Affordability, however, remains a barrier. LeCun pointed out that the cost of inference-primarily energy consumption-must fall dramatically before AI can be deployed at scale in a country of 1.4 billion people [185-186][108-113]. Without such reductions, the technology will stay out of reach for the majority of the population.


Illustrating practical benefits, LeCun described a pilot where smart glasses equipped Indian farmers with an AI assistant that could diagnose plant diseases, advise on harvest timing and provide weather forecasts [187-188][118-120]. He argued that, once costs drop, similar tools could improve agriculture, healthcare and education, acting as a “printing press” that broadens knowledge access [131-136].


Addressing education, the moderator asked whether AI will make students “more literate or more AI-dependent”. LeCun replied that dependence on technology is normal and that AI will facilitate access to knowledge much like the printing press, augmenting learning rather than creating harmful reliance [131-136].


LeCun also critiqued the term “artificial general intelligence”, arguing that human intelligence is highly specialised and that true intelligence should be measured by rapid task learning without prior training [61-71][168-171]. He warned against equating language proficiency with intelligence, noting that language is a finite set of discrete symbols-relatively easy for machines-whereas the real world presents high-dimensional, continuous, noisy signals, a disparity known as the Moravec paradox [172-176].


His new research programme focuses on building “intelligence for the real world”, i.e., systems capable of constructing robust world models that can predict consequences and plan actions in messy environments [176-178]. He acknowledged that achieving such capabilities will be the central challenge for AI over the next decade.


In concluding remarks, LeCun likened AI’s societal impact to that of the printing press rather than to electricity, suggesting that AI will amplify human intelligence and democratise knowledge, though the exact ways societies will need to adapt remain uncertain [139-146]. He expressed cautious optimism that societies will eventually discover how best to harness the technology for the public good [145-146].


LeCun added that historically we tend to over-estimate short-term impact and under-estimate long-term impact, a pattern he expects to continue with AI [190].


The dialogue revealed strong consensus on three fronts: (1) current AI, especially LLMs, is powerful yet not truly intelligent; (2) AI should be viewed as an augmentative tool that demands extensive up-skilling and higher-level education; and (3) reducing inference cost is essential for mass adoption, particularly in the Global South. Points of disagreement centered on the immediacy of a super-intelligent breakthrough, the future of openness amid economic growth, and whether AI-driven education will foster dependence or genuine literacy [14-15][57-64][124-125].


Overall, the conversation balanced optimism about AI’s transformative potential with a sober appraisal of technical limits, modest economic gains, and the societal infrastructure required to turn AI into a true amplifier of human capability. [Mr. Yeltsin]

Session transcriptComplete transcript of the session
Speaker 1

Thank you so much, Mr. Brad Smith, for that very energizing address, ladies and gentlemen. I think he really deserves an energetic applause from you all. His address has actually given a very constructive direction to the discourse on artificial intelligence. And well, now we are moving to the next conversation for which our guest is the person who’s often called the godfather of deep learning. Our guest is Mr. Yann LeCun, Executive Chairman, Advanced Machine Intelligence Labs. And his foundational work on convolutional neural networks underpins virtually every image recognition system in use today. Now at the frontier of next generation AI architectures, he’s one of the field’s most provocative and independent voices. Please welcome our next speaker, Mr.

Yann LeCun, and this conversation will be moderated by Ms. Maria Shakil, Managing Editor, India Today. Please welcome our guest and our moderator.

Maria Shakil

Mr. Yann LeCun. Welcome. Good afternoon, everyone. So let’s begin with a big idea here, Professor LeCun. Are we on a path to creating the smartest mind that humanity has ever known? And will that happen in our lifetime?

Yann LeCun

Maybe in the lifetime of some people here, possibly not in mine. We’ll see. It will take a while. But I think the more interesting… thing that we’re going to build is an amplifier for human intelligence. So maybe not an entity that surpasses human intelligence in all domain, although that will happen at some point, but it is something that will amplify human intelligence in ways that will accelerate progress.

Maria Shakil

So then will we end up defining and redefining genius? What will a genius be?

Yann LeCun

Well, you know, I think several thousand years ago, or even a few centuries ago, what people identified as genius is very different from what we currently identify as genius. And I think there will be more evolution of that concept of genius. You know, in the past, perhaps, you know, genius was, you know, some act of creation or invention, but maybe not at theoretical level like we are. We tend to think of it today. It was, you know, more practical, certainly in the very ancient past, people who figured out how to cultivate crops or domesticate animals probably were seen as genius.

Maria Shakil

So, you know, we have often seen, and this is a thought that you have all, you know, pretty openly shared, that AI is powerful but not intelligent. When we make that distinction and there are conversations around LLM, where do you see intelligence and AI -driven power?

Yann LeCun

Yeah, I think there’s a lot of confusion, really, because we tend to anthropomorphize systems that can reproduce certain human functions. So what’s, I mean, LLMs are incredibly useful. There’s no question about that. And they do amplify human intelligence, like computer technology going back to the 1940s. But LLMs, to some extent, except for a few domains, are mostly information retrieval systems. They can compress a lot of factual knowledge that has been previously produced by humans and can give easy access to it. In a way, it’s kind of a natural evolution of the printing press, the libraries, the Internet, and search engines, right? It’s just a more efficient way to access information. And there are a few domains where the intelligent capabilities of those systems actually is more than that.

It’s more than just retrieval. So for generating code, maybe doing some type of mathematics, we’re getting the impression that it’s beyond this. But it’s still, to a large extent, domains where reasoning has to do with manipulating symbols. The problem is that… you know why do we have systems that can pass the bar exam and win uh mathematics olympiads but we don’t have domestic robots we don’t even have self -driving cars and we certainly do not have self -driving cars that can teach themselves to drive in 20 hours of practice like any 17 year old so we’re missing something big still

Maria Shakil

so what are we teaching a 17 year old then

Yann LeCun

well so the you know the question is how does how does a baby learn uh or even an animal right animals have a much better understanding of the physical world than any ai systems that we have today which is why uh you know we don’t have smart robots um and and so you know we learn we learned about about the world how the world works mostly by observation when we are babies a few months old and then we learn by interaction and we learn mental models of the world that allows us to apprehend any new situation even if we haven’t been uh you know exposed to it beforehand we can still handle it so a big buzzword in ai today is world models and this is really this idea that we we We develop mental models of the world that allow us to think ahead, to apprehend new situations, plan sequence of actions, reason, and predict the consequences of our actions, which is absolutely critical.

And LLMs don’t do this, really.

Maria Shakil

There is the sense, Professor, that perhaps AI will unlock an era of radical abundance. Will this abundance benefit us?

Yann LeCun

Well, if you talk to economists, they tell you, if we can measure the improvement that AI will bring to productivity, which is the amount of wealth produced per hour worked, it’s going to add up to maybe 0 .6 % per year. This is from economists that actually have studied the effect of technological revolutions on the labor market and the economy. People like Philippe Ackermann. Like Jung or Eric Brynjolfsson. And so that seems small. It’s actually quite big. And, you know, it’s certainly going to accelerate scientific progress, progress in medicine. I do not believe there’s going to be a singular identifiable point where the economy is going to take off and there’s going to be abundance. And there’s also the question of the policies surrounding this.

Are those benefits going to be shared across humanity or different categories of people in various countries? That’s a political question. It has nothing to do with technology.

Maria Shakil

So if economists see this as boom, will then openness survive?

Yann LeCun

It’s not going to be an event. It’s going to be progressive. There is this false idea that somehow at some point we’re going to discover the secret of human level intelligence. I don’t like the phrase AGI because human intelligence is specialized. So I don’t like the artificial general intelligence phrase. But it’s not going to be an event. We’re not going to discover one secret. We’re going to make… continuous progress. And we’re not going to be able to measure that progress by just having a series of tests that are going to test, you know, whether a machine is more intelligent than humans, because machines are already more intelligent than humans on a large number, a growing number of narrow tasks.

And so, you know, it’s not like a uniform, you know, scalar measurement of quality. It’s a collection of quality. But what’s more important is that intelligence is not just a collection of skills. It’s an ability to learn new skills extremely quickly, and even to accomplish new tasks without being trained to do it the first time we apprehend it. That’s really what, you know, intelligence should be measured at. So we’re not going to be able to just design a test that is going to figure out, you know, are machines more intelligent than humans.

Maria Shakil

So if it’s about upskilling and ensuring that you’re relevant, then only perhaps you’re intelligent. Will that then mean that the countries that adopt AI and the pace at which India and the scale at which India has adopted AI, the challenge would be to create talent which is upskilled and reskilled and have the required skills for this?

Yann LeCun

Absolutely. So the relationship that we’re going to have with intelligent AI systems is going to be similar to the relationship that a leader in business politics or academia or some other domain has with their staff. AI is going to be our staff. Every one of us is going to be a manager of a staff of intelligent machines. They’ll do our bidding. They might be smarter than us. But certainly if you are an academic or a politician, you work with staff that are smarter than you. In fact, that’s the whole point. attract people who are smarter than you because that’s what makes you more productive. For an academic, it’s students who are smarter than their professor.

It’s not the professor that teaches graduate students. It’s the other way around, actually. And certainly, we have a lot of examples of politicians who are surrounded by people who are smarter than them.

Maria Shakil

Earlier today, when Prime Minister Modi addressed the gathering, he said that India doesn’t fear AI. We are seeing this as our destiny future, which is Bhagya. Do you see that with a summit of this nature being hosted in India, it’s a message to the global south? And that’s where perhaps the next big innovation in AI could be coming from?

Yann LeCun

Well, long term, it’s going to come from countries that have, for example, favorable demographics. And that means India, Africa. You know, the youth is the most creative part of humanity and there’s sort of a deficit of that in the North, largely. So, you know, the scientists, the top scientists of the future, in fact, many of the present are from India and in the future will be from mostly Africa. So what does that mean, though? Right. It means having incentives for young people to kind of study, first of all. So the idea that somehow we don’t need to study anymore because AI is going to do it for us. And, you know, that’s completely false, absolutely completely false.

And it’s not because I’m a professor, OK, that I’m saying this. On the contrary, we’re going to have to study more. We see, for example, a trend. Where in industry, in the past, in certain countries, it’s certainly true for India, but it’s also true in European countries. And certainly in the U .S., we see. more demand for people with more education at the PhD level, for example. The demand for PhD -level scientists in industry has grown in the last 15 years, in part because of AI, but because of everything, because technological progress hinges on scientific progress, and scientific progress is brought about by scientists, and scientists mostly have done PhDs. And so there is more demand for education, not less.

And so for countries in the Global South, that means investing in education and youth.

Maria Shakil

And making AI more accessible, something that India believes in, democratizing AI, AI for all, is the theme of this summit as well. Do you think AI can become that accessible, particularly for a… country as large as ours with 1 .4 billion people?

Yann LeCun

Yeah, in all kinds of ways. Unfortunately, the cost of inference for AI system has to come down to kind of become practical for the vast majority of population in a country like India. Right now, the inference is just too expensive. And, you know, energy costs and things like that. It’s mostly energy costs, actually. So this has to come down, but then it will play a role in education. AI will improve the quality of education, not degrade it. Once we figure out how to use it best, it will improve agriculture and everything else. And healthcare in particular. Healthcare, of course, right? So I don’t work at Meta anymore, but there was an experiment a couple of years ago or a year ago that was run by my former colleagues where they gave smart glasses to people.

Agriculture, you know, to farmers in India. And they could talk to the AI assistant to figure out, like, you know, what is this disease on my plant or should I harvest now or what’s going to be the weather?

Maria Shakil

Yes, it is being used a lot in agriculture as well. That’s right. It is assisting farmers to ensure that their produce gets better. They make right choices. But when you say about education, will AI assist education in terms of making students or youth of the country more literate or will they become more AI dependent?

Yann LeCun

Well, I mean, we’re dependent on technology, right? I’m dependent on this pair of glasses. Otherwise, I don’t see you. So that has been with us for centuries. Yeah, we’ll be dependent on AI, of course. But AI will facilitate. Access to knowledge and thereby going to be a tool for education. I think the effect on society. might be extrapolated from what was observed in the 15th century when the printing press started enabling the production of printed matters and the dissemination of knowledge. It had a huge effect on society worldwide, at least in countries that allowed it to flourish. And I think it’s going to be a similar transformation with AI, of course, in the modern world, just more access to knowledge.

The Internet played also a similar role. And I think this is just going to make people more informed, smarter, able to make more rational decisions if it’s deployed in the proper way.

Maria Shakil

So if you were to define this moment, which we are witnessing in history, we are living it, how will you say it? Is it like the advent of electricity?

Yann LeCun

Yeah, people have made that claim. They have made that claim, yes.

Maria Shakil

Including?

Yann LeCun

The printing economy is the new electricity. I think it’s more like the new printing press, really. Again, in this vision of more dissemination and sharing of knowledge and amplification of human intelligence. But the impact on society and the way countries need to be run is very difficult to predict at this point. I’m sort of an optimist in the sense that I think societies would figure out how best to use that technology for the benefits of their population.

Maria Shakil

While I am an optimist, nevertheless, I’m going to ask this question to you, Professor. Are we overestimating the change or underestimating what has struck us?

Yann LeCun

So, usually in technological shifts of this type, we are overestimating. And the changes in the short term and overestimating them in the long term. Now, I think for AI, it’s a little bit different because there’s been a huge amount of hype and expectations that, you know, the transition to human -level AI, superhuman -level AI is going to be an event and is going to happen within the next few years. And people have been making that claim for the last 15 years, and it’s been false. In fact, they’ve been making it for the last 60 years or 70 years, and it’s been false. Every time in the history of AI that scientists have discovered kind of a new paradigm of AI, how you build intelligent machines, people have claimed, you know, within 10 years, the smartest entity on the planet will be a computer.

And that just proved to be wrong, you know, four or five times in the last 70 years. It’s still wrong. We’re still very far from that. We’re not very far. We’re getting close, right? We’re seeing the end of the tunnel. But it’s not like, you know, we’re going to have. Super intelligent systems within two years. It’s just not happening because of this gap. You know, where is the robot that can learn to drive? 20 hours of practice like a 17 -year -old, even though we have millions of hours of training data of people driving cars around, we should be able to train an AI system to just imitate that. That doesn’t actually quite work. It’s not reliable enough.

Maria Shakil

Okay, so let’s try and wrap up this conversation with who gets to define intelligence now onwards. Will it be actually humans, machines, or both together?

Yann LeCun

Probably both together, but mostly humans. We set the agenda, and the biggest difficulty is not to get fooled into thinking that a computer system is intelligent simply because it can manipulate language. We tend to think of language as the epitome of human intelligence, right? But in fact, it turns out language is easy to deal with because language is really a sequence of discrete symbols of which there is only a finite number. And that turns out to make things easy when you train a system to predict what the next word is in a text, which is what LLMs are based on. It turns out the real world is much, much more complicated. And it’s been known in computer science for many years.

It’s called the Moravec paradox after roboticist Hans Moravec. And so the company I’m building and the research program I’ve been working on for the last 15 years or so is intelligence for the real world. You know, how to deal with high -dimensional, continuous, noisy signal that the real world is, which your house cat is perfectly able to deal with or a squirrel or whatever, but not computers yet. That’s the big challenge for the next few years in AI, dealing with the real world. And that’s the point of the company I’m building.

Maria Shakil

So AI has to deal with the real world or real world has to deal with AI.

Yann LeCun

AI has to deal with the real world, the messiness of the real world, the unpredictability of the real world.

Maria Shakil

All right. Thank you so much for this conversation, Professor. Thank you, Mr. Yeltsin.

Related ResourcesKnowledge base sources related to the discussion topics (35)
Factual NotesClaims verified against the Diplo knowledge base (4)
Confirmedhigh

“Speaker 1 thanked Brad Smith for his “energising address”.”

The transcript records the host saying “Thank you so much, Mr. Brad Smith, for that very energizing address” confirming the description of his address as energising. [S9]

Confirmedhigh

“LeCun said a truly smartest entity might appear within the lifetime of some audience members but is unlikely in his own lifetime and will take “a while”.”

LeCun is quoted as saying “Maybe in the lifetime of some people here, possibly not in mine… It will take a while.” which matches the report’s wording. [S89]

Confirmedhigh

“LeCun framed AI’s immediate goal as building an “amplifier for human intelligence”.”

LeCun explicitly states “the more interesting… thing that we’re going to build is an amplifier for human intelligence.” [S89]

Additional Contextmedium

“LeCun emphasized that AI development will be continuous rather than a single breakthrough event.”

In another segment LeCun remarks that AI progress “is not going to be an event. It’s going to be kind of progressive innovations,” providing additional context for the claim about continuous development. [S1]

External Sources (96)
S1
Steering the future of AI — # Discussion Report: Yann LeCun on the Future of Artificial Intelligence This discussion featured Yann LeCun, Meta’s Ch…
S2
[Parliamentary Session 3] Researching at the frontier: Insights from the private sector in developing large-scale AI systems — She mentions advice from Yann LeCun, a professor at NYU and advisor at Meta, who advocates for this approach.
S3
Meta’s chief AI scientist Yann LeCun departs to launch world-model AI startup — Yann LeCun, one of the pioneers of deep learning and Meta’s chief AI scientist, isleavingthe company to establish a new …
S4
Keynote-Martin Schroeter — -Speaker 1: Role/Title: Not specified, Area of expertise: Not specified (appears to be an event moderator or host introd…
S5
Responsible AI for Children Safe Playful and Empowering Learning — -Speaker 1: Role/title not specified – appears to be a student or child participant in educational videos/demonstrations…
S6
Building Trusted AI at Scale Cities Startups &amp; Digital Sovereignty – Keynote Vijay Shekar Sharma Paytm — -Speaker 1: Role/Title: Not mentioned, Area of expertise: Not mentioned (appears to be an event host or moderator introd…
S7
Fireside Conversation: 02 — -Maria Shakil: Managing Editor, India Today (serving as moderator for the conversation)
S8
Conversation: 01 — Artificial intelligence
S9
Fireside Conversation: 02 — The moderator establishes LeCun’s credibility by highlighting his role as the ‘godfather of deep learning’ and his found…
S10
Debating Technology / Davos 2025 — Yann LeCun: By all measures, actually, META is leading in terms of content moderation, absolutely. Yann LeCun: OK. I…
S11
How AI Drives Innovation and Economic Growth — “So, you know, for all countries, but especially for emerging markets and developing economies, AI can be a game changer…
S12
Keynote-Ankur Vora — “Technologists can choose whether we use AI to take on the world’s greatest challenges or just the most precious.”[1]. “…
S13
The Foundation of AI Democratizing Compute Data Infrastructure — LeCun noted that while industry incentives exist to reduce power consumption, progress isn’t fast enough, and real break…
S14
AI for Social Good Using Technology to Create Real-World Impact — I think broadly speaking, I think, especially in the global south… the cost of AI inference has to drop dramatically b…
S15
The Foundation of AI Democratizing Compute Data Infrastructure — He provided a compelling comparison: LLMs are trained on approximately 10^14 bytes of text data, representing roughly ha…
S16
Safe and Responsible AI at Scale Practical Pathways — Ramaswami emphasizes that AI should be viewed as a tool that enhances human capabilities rather than replacing human int…
S17
Launch / Award Event #52 Intelligent Society Development &amp; Governance Research — Xunhua Guo: It’s a great pleasure to gather with you today at the 20th United Nations Internet Governance Forum to explo…
S18
From Technical Safety to Societal Impact Rethinking AI Governanc — Both speakers support government involvement but disagree on scope – Ioannidis wants to keep core technology development…
S19
Comprehensive Report: Preventing Jobless Growth in the Age of AI — Economic | Future of work While AI demonstrates substantial productivity improvements in specific applications, these g…
S20
Impact &amp; the Role of AI How Artificial Intelligence Is Changing Everything — “Technology may disrupt and may replace, but it will also create new jobs and new opportunities.”[54]. “For everybody, I…
S21
Developing capacities for bottom-up AI in the Global South: What role for the international community? — Amandeep Singh Gill: Thank you so much, Jovan, and thank you to you, Diplo Foundation, and its partners for convening th…
S22
Democratizing AI Building Trustworthy Systems for Everyone — “So this is both… Investments in data centres to power AI applications, but it’s also investments in connectivity as w…
S23
What policy levers can bridge the AI divide? — However, significant challenges remain, including connecting billions of unconnected people, ensuring affordability, dev…
S24
The Innovation Beneath AI: The US-India Partnership powering the AI Era — Tobias Helbig acknowledges that AI follows typical hype cycles with periods of disillusionment, but emphasizes that the …
S25
Policymaker’s Guide to International AI Safety Coordination — This comment crystallizes the fundamental tension at the heart of AI governance – the misalignment between market incent…
S26
AI/Gen AI for the Global Goals — Shea Gopaul: So thank you, Sanda. And like Sandra, I’d like to thank the African Union, as well as Global Compact. i…
S27
Comprehensive Report: Preventing Jobless Growth in the Age of AI — Distribution of Productivity Benefits and Inequality Economic | Future of work | Human rights Need for policies on tax…
S28
AI drives productivity surge in certain industries, report shows — A recent PwC (PricewaterhouseCoopers International Limited) reporthighlightsthat sectors of the global economy with high…
S29
UNSC meeting: Artificial intelligence, peace and security — Current AIs are information processing tools without real understanding
S30
Fireside Conversation: 02 — This discussion features AI pioneer Yann LeCun, known as the “godfather of deep learning,” speaking with moderator Maria…
S31
Fireside Conversation: 02 — This discussion features AI pioneer Yann LeCun, known as the “godfather of deep learning,” speaking with moderator Maria…
S32
Building Trusted AI at Scale – Keynote Anne Bouverot — Bouverot argues that the location of the summit in India, representing the global south, has both symbolic and strategic…
S33
The Role of Government and Innovators in Citizen-Centric AI — Speaker 1 expresses a desire to see increased collaboration between India and the European Union to build capacity for b…
S34
WS #82 A Global South perspective on AI governance — AUDIENCE: Ends up. We cannot hear. Rely on ISO 31,000 is what they see as the kind of framework for risk assessments…
S35
AI Impact Summit 2026: Global Ministerial Discussions on Inclusive AI Development — In the global south. the timing and the location are equally important. As AI technology has continued to advance so has…
S36
Press Conference: Closing the AI Access Gap — Moreover, the speakers argue that AI can drive productivity, creativity, and overall economic growth. It has the capacit…
S37
Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Keynote Cristiano Amon — Evidence:There is a process of jumping into a large -scale industrialization. India is becoming a global manufacturing h…
S38
Powering AI Global Leaders Session AI Impact Summit India — But I was a history major in college, so I get to play amateur historian, emphasis on amateur. Everyone has their own fa…
S39
AI for Social Good Using Technology to Create Real-World Impact — This argument emphasizes that for AI to be viable in developing countries, the cost of running AI models (inference) mus…
S40
AI for Social Good Using Technology to Create Real-World Impact — And I think that’s what we’re doing. And to give you another example of how it reduces the complexity, there’s a very in…
S41
Is AI the key to nuclear renaissance? — The global acceptance and widespread use of artificial intelligence are greatly affecting worldwide energy demands and t…
S42
Education, Inclusion, Literacy: Musts for Positive AI Future | IGF 2023 Launch / Award Event #27 — Connie Book:Thank you. We will now hear from law researcher of the University of Montreal. Her research focuses on the i…
S43
Empowering India & the Global South Through AI Literacy — Explanation:The unexpected consensus emerges around the government’s commitment to introduce AI education from class thr…
S44
Responsible AI for Children Safe Playful and Empowering Learning — Absolutely. We need to generate a fair amount of evidence before we rush to scale with something like this. Although we …
S45
Building the AI-Ready Future From Infrastructure to Skills — “And things are moving in a way that we cannot predict that the only way that anybody is going to be successful is an op…
S46
GOVERNING AI FOR HUMANITY — Open-source AI systems encourage innovation and are often a requirement for public funding. On the open extreme of the s…
S47
The strategic shift toward open-source AI — The release of DeepSeek’s open-source reasoning model in January 2025, followed by the Trump administration’s July endor…
S48
Generative AI: Steam Engine of the Fourth Industrial Revolution? — An open ecosystem that allows for the free use and accessibility of digital technologies and cloud services is advantage…
S49
AI: The Great Equaliser? — South Korean companies still have the opportunity to operate fabrication facilities in China, and trade plays a crucial …
S50
Artificial General Intelligence and the Future of Responsible Governance — This panel discussion focused on Artificial General Intelligence (AGI) and its implications for security, privacy, and e…
S51
The Gig Economy: Positioning Higher Education at the Center of the Future of Work (USAID Higher Education Learning Network) — In order to adapt to the changing landscape of upskilling and reskilling, higher education should take the lead in drivi…
S52
Turbocharging Digital Transformation in Emerging Markets: Unleashing the Power of AI in Agritech (ITC) — Moreover, while AI and new technologies have significant potential in agriculture, it is crucial to understand that they…
S53
Fireside Conversation: 02 — Despite their impressive capabilities, LeCun characterizes current Large Language Models as “mostly information retrieva…
S54
Safe and Responsible AI at Scale Practical Pathways — Ramaswami emphasizes that AI should be viewed as a tool that enhances human capabilities rather than replacing human int…
S55
Fireside Conversation: 02 — This discussion features AI pioneer Yann LeCun, known as the “godfather of deep learning,” speaking with moderator Maria…
S56
Steering the future of AI — ## LeCun’s Position on Large Language Models 3. **Reasoning capabilities**: While LLMs can simulate reasoning, they lac…
S57
MISUNDERSTOOD: THE IT MANAGER’S LAMENT – A CASE STUDY IN INTER-PROFESSIONAL MISCOMMUNICATION — Mental skills are far more varied in scope and nature than the now somewhat discredited IQ tests indicate, based as they…
S58
Empowering Workers in the Age of AI — Tom Wambeke: Good afternoon. This is the last input before we can go a little bit more interactive. As you see from the …
S59
The mismatch between public fear of AI and its measured impact — HAI is careful to distinguish between job exposure and job loss. Many occupations are exposed to AI tools, but exposure …
S60
Who Benefits from Augmentation? / DAVOS 2025 — Kumar argues that AI can lead to increased productivity and the creation of new job opportunities. He suggests that this…
S61
Comprehensive Report: Preventing Jobless Growth in the Age of AI — Economic | Future of work While AI demonstrates substantial productivity improvements in specific applications, these g…
S62
Leaders’ Plenary | Global Vision for AI Impact and Governance Morning Session Part 1 — The IMF calculated that AI has potential to provide up to 0.8% boost to global growth over the coming years, which would…
S63
Developing capacities for bottom-up AI in the Global South: What role for the international community? — Amandeep Singh Gill: Thank you so much, Jovan, and thank you to you, Diplo Foundation, and its partners for convening th…
S64
Upskilling for the AI era: Education’s next revolution — Doreen Bogdan Martin: Good afternoon, ladies and gentlemen. Yesterday morning on this very stage I spoke about skills. I…
S65
WS #100 Integrating the Global South in Global AI Governance — AUDIENCE: I think beyond skills programs and helping developers and people working in those industries in the click co…
S66
The Innovation Beneath AI: The US-India Partnership powering the AI Era — Tobias Helbig acknowledges that AI follows typical hype cycles with periods of disillusionment, but emphasizes that the …
S67
AI Infrastructure and Future Development: A Panel Discussion — Lessin acknowledges the very positive outlook presented by all panelists but probes for potential obstacles or risks tha…
S68
The AI Pareto Paradox: More computing power – diminishing AI impact?  — This is the gist of AI transformation. It isn’t a technical hurdle; it’s a human one. It requires workers who don’t just…
S69
‘The elephant in the AI room’: Does more computing power really bring more useful AI? — A less inflated AI debate won’t slow progress. But it may be the only way to ensure progress remains sustainable—and gen…
S70
AI Development Beyond Scaling: Panel Discussion Report — The tone began as optimistic and technically focused, with researchers enthusiastically presenting their innovative appr…
S71
Comprehensive Discussion Report: AI’s Transformative Potential for Global Economic Growth — The conversation maintains a consistently optimistic and enthusiastic tone throughout. Both speakers demonstrate genuine…
S72
Comprehensive Report: China’s AI Plus Economy Initiative – A Strategic Discussion on Artificial Intelligence Development and Implementation — The tone was consistently optimistic and collaborative throughout the conversation. Participants demonstrated mutual res…
S73
Partnering on American AI Exports Powering the Future India AI Impact Summit 2026 — The tone is consistently optimistic, collaborative, and forward-looking throughout the discussion. Speakers emphasize “l…
S74
AI Innovation in India — The tone was consistently celebratory, inspirational, and optimistic throughout the discussion. Speakers expressed pride…
S75
Bridging the Digital Skills Gap: Strategies for Reskilling and Upskilling in a Changing World — High level of consensus with complementary rather than conflicting viewpoints. The agreement suggests a mature understan…
S76
Session — The tone of the discussion was largely analytical and academic, with participants offering nuanced views on complex issu…
S77
Large Language Models on the Web: Anticipating the challenge | IGF 2023 WS #217 — One of the leading generative AI approaches is the so-called Large Language Models (LLMs), complex models capable of und…
S78
The Expanding Universe of Generative Models — He mentions that they provide a pretense of logical reasoning, can generate content, improve productivity and are being …
S79
Transforming Agriculture_ AI for Resilient and Inclusive Food Systems — The tone was consistently optimistic yet pragmatic throughout the conversation. Speakers maintained an encouraging outlo…
S80
Panel 4 – Resilient Subsea Infrastructure for Underserved Regions  — The discussion maintained a professional, collaborative tone throughout, with panelists building on each other’s insight…
S81
Advancing Scientific AI with Safety Ethics and Responsibility — The discussion maintained a collaborative and constructive tone throughout, characterized by technical expertise and pol…
S82
What policy levers can bridge the AI divide? — The discussion maintained a collaborative and optimistic tone throughout, with participants sharing experiences construc…
S83
AI for Safer Workplaces &amp; Smarter Industries Transforming Risk into Real-Time Intelligence — There was unexpected consensus that fear about AI is widespread across different age groups and demographics, but this f…
S84
AI for Safer Workplaces & Smarter Industries_ Transforming Risk into Real-Time Intelligence — Explanation:There was unexpected consensus that fear about AI is widespread across different age groups and demographics…
S85
How AI Drives Innovation and Economic Growth — The discussion maintained a balanced, pragmatic tone throughout, characterized by cautious optimism. While panelists ack…
S86
Cybersecurity in the Age of Artificial Intelligence: A World Economic Forum Panel Discussion — The discussion maintained a serious but measured tone throughout, with the moderator explicitly stating his hope for an …
S87
Skilling and Education in AI — The tone was cautiously optimistic throughout. Speakers acknowledged both the tremendous opportunities AI presents for I…
S88
Are we creating alien beings? — This comment is philosophically profound because it acknowledges the unprecedented nature of our current moment in histo…
S89
https://dig.watch/event/india-ai-impact-summit-2026/fireside-conversation-02 — Maybe in the lifetime of some people here, possibly not in mine. We’ll see. It will take a while. But I think the more i…
S90
AI Innovation in India — Deepak Bagla argues that India stands to benefit the most from AI as a transformative force due to its massive and growi…
S91
AI for Democracy_ Reimagining Governance in the Age of Intelligence — Chunggong acknowledges the significant positive potential of AI for social good, including improvements in healthcare de…
S92
The myth of the lone genius: How scientific revolutions really happen — And I’m not talking of the wholesalestealing that learned minds did from ‘artisans and low mechanicks’ (seeA People’s Hi…
S93
BILATERAL AAV — covers the performing and the plastic arts, and involves exchanges of delegations of artistes and exhibitions, visits by…
S94
Engineering Accountable AI Agents in a Global Arms Race: A Panel Discussion Report — The discussion maintained a thoughtful but somewhat cautious tone throughout, with speakers acknowledging both opportuni…
S95
WSIS Action Line C2 Information and communication infrastructure — This clarification was so impactful that the moderator specifically highlighted it in his closing remarks. It provided a…
S96
Hard power of AI — He argues that his models are objectively better, emphasizing innovation and development. There is an ongoing debate sur…
Speakers Analysis
Detailed breakdown of each speaker’s arguments and positions
S
Speaker 1
1 argument129 words per minute143 words66 seconds
Argument 1
The introductory framing emphasizes LeCun’s role in shaping AI discourse and his pioneering work on deep learning (Speaker 1)
EXPLANATION
Speaker 1 introduces Yann LeCun as a leading figure in AI, highlighting his foundational contributions to convolutional neural networks and his influence on current AI debates. The introduction sets the tone for the discussion by positioning LeCun as a key authority.
EVIDENCE
Speaker 1 thanks the previous speaker, applauds his address, and then describes LeCun as “the godfather of deep learning,” noting that his work on convolutional neural networks underpins virtually every image-recognition system in use today, and calls him a provocative and independent voice at the frontier of next-generation AI architectures [1-8].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
The moderator’s introduction of LeCun as the “godfather of deep learning” and his foundational work on convolutional neural networks is documented in [S9].
MAJOR DISCUSSION POINT
LeCun’s stature and influence in AI
Y
Yann LeCun
14 arguments153 words per minute2329 words908 seconds
Argument 1
AI will serve as an amplifier for human intelligence rather than a fully autonomous supermind (Yann LeCun)
EXPLANATION
LeCun argues that the most valuable AI we will build is a tool that augments human intellect, not a separate entity that eclipses human cognition in every domain. This amplification will accelerate scientific and societal progress.
EVIDENCE
He says, “the more interesting… thing that we’re going to build is an amplifier for human intelligence… maybe not an entity that surpasses human intelligence in all domain… it is something that will amplify human intelligence in ways that will accelerate progress” [14-17].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
LeCun explicitly states that the most interesting development will be an amplifier for human intelligence, not a supermind that surpasses humans in all domains [S9].
MAJOR DISCUSSION POINT
AI as an intelligence amplifier
AGREED WITH
Maria Shakil
DISAGREED WITH
Maria Shakil
Argument 2
The emergence of a truly smartest mind may occur within some participants’ lifetimes, but not necessarily within the speaker’s own (Yann LeCun)
EXPLANATION
LeCun acknowledges that a super‑intelligent system could appear within the lifetime of some audience members, but he is uncertain it will happen within his own lifespan, emphasizing the long‑term nature of the challenge.
EVIDENCE
He replies, “Maybe in the lifetime of some people here, possibly not in mine… We’ll see. It will take a while.” [14-15].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
LeCun’s measured response that a super-intelligent system might appear “maybe in the lifetime of some people here, possibly not in mine” is recorded in [S7]; a more concrete timeline (“no less than five years… could be 10, it could be more”) appears in [S1].
MAJOR DISCUSSION POINT
Timeline for super‑intelligence
Argument 3
Current large language models are chiefly advanced information‑retrieval tools, comparable to a modern printing press, not true reasoners (Yann LeCun)
EXPLANATION
LeCun characterises LLMs as sophisticated systems for storing and retrieving factual knowledge, likening them to the evolution of the printing press, libraries, and search engines, rather than genuine reasoning engines.
EVIDENCE
He notes that LLMs “are mostly information retrieval systems… they can compress a lot of factual knowledge… a natural evolution of the printing press, the libraries, the Internet, and search engines… just a more efficient way to access information” [30-34].
MAJOR DISCUSSION POINT
Limits of LLMs
Argument 4
Economists estimate AI‑driven productivity growth at roughly 0.6 % per year, which is modest yet significant for scientific and medical progress (Yann LeCun)
EXPLANATION
LeCun cites economic research suggesting AI will raise productivity by about 0.6 % annually, a figure that, while seemingly small, can have substantial cumulative effects on scientific and medical advancements.
EVIDENCE
He references economists who say AI will add “maybe 0 .6 % per year” to productivity, naming researchers such as Philippe Ackermann, Jung, and Eric Brynjolfsson, and stresses its importance for scientific progress [45-52].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
LeCun cites economists who estimate AI will add about 0.6 % per year to productivity, a point highlighted in [S9].
MAJOR DISCUSSION POINT
Economic impact of AI
Argument 5
Whether AI‑generated wealth benefits all humanity is a political, not a technological, issue (Yann LeCun)
EXPLANATION
LeCun stresses that the distribution of AI‑driven gains depends on policy choices rather than technical capabilities, framing the question of shared abundance as a political challenge.
EVIDENCE
He asks, “Are those benefits going to be shared across humanity…? That’s a political question. It has nothing to do with technology.” [54-56].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
LeCun frames the distribution of AI benefits as a political question, a view echoed in [S9]; related governance perspectives are discussed in [S12].
MAJOR DISCUSSION POINT
Political nature of AI benefits
DISAGREED WITH
Maria Shakil
Argument 6
The AI era demands more, not less, advanced education; demand for PhD‑level scientists is rising worldwide (Yann LeCun)
EXPLANATION
LeCun argues that contrary to fears of AI replacing study, the field actually requires deeper and more extensive education, with industry increasingly seeking PhD‑qualified researchers.
EVIDENCE
He states, “we’re going to have to study more… demand for PhD-level scientists in industry has grown in the last 15 years… there is more demand for education, not less” [95-104].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
LeCun argues that “we’re going to have to study more” and points to growing industry demand for PhD-level researchers, documented in both [S7] and [S9].
MAJOR DISCUSSION POINT
Education demand in AI era
AGREED WITH
Maria Shakil
Argument 7
Countries with favorable demographics, like India and Africa, will become future hubs of AI innovation if they invest in youth education (Yann LeCun)
EXPLANATION
LeCun predicts that regions with young, creative populations—particularly India and Africa—will lead future AI breakthroughs, provided they create incentives for youth to pursue scientific study.
EVIDENCE
He notes, “long term, it’s going to come from countries that have… favorable demographics… India, Africa… the youth is the most creative part of humanity… top scientists of the future will be from India and Africa” [90-94].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
LeCun highlights India and Africa’s favorable demographics as future sources of AI talent in [S9]; the discussion also references India’s non-fearful stance toward AI in [S7].
MAJOR DISCUSSION POINT
Geographic shift of AI talent
Argument 8
Inference costs, driven mainly by energy consumption, must drop dramatically for AI to be practical for billions of users (Yann LeCun)
EXPLANATION
LeCun points out that the high energy and monetary cost of running AI inference is a barrier to widespread adoption, especially in populous countries, and that reducing these costs is essential for scalability.
EVIDENCE
He says, “the cost of inference for AI system has to come down… right now, the inference is just too expensive… it’s mostly energy costs” [109-113].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
LeCun identifies high inference energy costs as a barrier to widespread AI adoption [S9]; further commentary on the need for cost reductions appears in [S13] and [S14].
MAJOR DISCUSSION POINT
Cost barrier to AI accessibility
AGREED WITH
Maria Shakil
Argument 9
AI can improve agriculture, healthcare, and education once affordable, as illustrated by smart‑glass pilots for Indian farmers (Yann LeCun)
EXPLANATION
LeCun gives an example where AI‑enabled smart glasses helped Indian farmers diagnose plant diseases and make harvesting decisions, demonstrating AI’s potential to enhance agriculture, health, and education when deployment costs fall.
EVIDENCE
He recounts an experiment where “smart glasses… were given to farmers in India… they could talk to the AI assistant to figure out… what is this disease on my plant or should I harvest now or what’s the weather?” [118-120].
MAJOR DISCUSSION POINT
AI applications in essential sectors
AGREED WITH
Maria Shakil
DISAGREED WITH
Maria Shakil
Argument 10
The term “artificial general intelligence” is misleading; human intelligence is highly specialized, and intelligence should be judged by rapid learning of new tasks, not by a single test (Yann LeCun)
EXPLANATION
LeCun critiques the AGI label, arguing that human intelligence is domain‑specific and that true intelligence should be measured by the ability to quickly acquire new skills and solve novel problems, rather than by static benchmarks.
EVIDENCE
He says, “I don’t like the phrase AGI because human intelligence is specialized… intelligence should be measured at the ability to learn new skills extremely quickly… not by a single test” [61-71].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
LeCun rejects the AGI label, arguing that human intelligence is specialized and true intelligence is the ability to learn new skills quickly, as noted in [S7].
MAJOR DISCUSSION POINT
Redefining AGI
Argument 11
Intelligence will be co‑defined by humans and machines, but humans set the agenda and must avoid equating language proficiency with true understanding (Yann LeCun)
EXPLANATION
LeCun asserts that while machines will play a role in shaping intelligence, humans remain the primary definers, warning against conflating language mastery—an easy symbolic task—with genuine comprehension of the complex real world.
EVIDENCE
He states, “Probably both together, but mostly humans… we set the agenda… don’t get fooled into thinking a computer system is intelligent simply because it can manipulate language… language is easy… the real world is much more complicated… Moravec paradox” [168-176].
MAJOR DISCUSSION POINT
Human‑machine co‑definition of intelligence
Argument 12
Humans and animals learn physical world models through observation and interaction; current AI lacks such robust world models (Yann LeCun)
EXPLANATION
LeCun explains that babies and animals acquire mental models of the world by observing and interacting, enabling them to handle novel situations, whereas current AI, especially LLMs, does not build such world models.
EVIDENCE
He describes how “babies… learn by observation… then by interaction… develop mental models… LLMs don’t do this, really” [41-42].
MAJOR DISCUSSION POINT
Missing world models in AI
Argument 13
The biggest upcoming challenge is enabling AI to handle high‑dimensional, continuous, noisy real‑world signals—a problem highlighted by the Moravec paradox (Yann LeCun)
EXPLANATION
LeCun identifies the core research hurdle as teaching AI to process the messy, continuous sensory data of the real world, a difficulty encapsulated by the Moravec paradox, which notes that tasks easy for animals are hard for computers.
EVIDENCE
He references “the Moravec paradox… dealing with high-dimensional, continuous, noisy signal that the real world is… not computers yet” [175-179].
MAJOR DISCUSSION POINT
Real‑world perception challenge
Argument 14
The speaker’s new company focuses on building intelligence for the real world, moving beyond symbolic manipulation toward genuine perception and action (Yann LeCun)
EXPLANATION
LeCun mentions that his current venture is dedicated to creating AI systems capable of real‑world intelligence, emphasizing perception, planning, and interaction rather than mere symbol‑level processing.
EVIDENCE
He says, “the company I’m building… intelligence for the real world… dealing with high-dimensional… that’s the big challenge… and that’s the point of the company I’m building” [176-179].
MAJOR DISCUSSION POINT
Company’s research direction
M
Maria Shakil
2 arguments128 words per minute479 words223 seconds
Argument 1
AI is powerful but not intelligent; we risk anthropomorphising systems that merely mimic human functions (Maria Shakil)
EXPLANATION
Maria cautions that while AI can perform impressive tasks, it does not possess true intelligence, and people tend to attribute human-like qualities to systems that only replicate specific functions.
EVIDENCE
She remarks, “we have often seen… AI is powerful but not intelligent… When we make that distinction… where do you see intelligence and AI-driven power?” [25-26].
MAJOR DISCUSSION POINT
Distinguishing power from intelligence
AGREED WITH
Yann LeCun
Argument 2
The moderator highlights India’s stance that AI is not feared but embraced as destiny, questioning if this signals a global‑south opportunity (Maria Shakil)
EXPLANATION
Maria points out Prime Minister Modi’s statement that India does not fear AI and frames the summit as a signal to the Global South that AI could become a source of future innovation.
EVIDENCE
She notes, “Earlier today, when Prime Minister Modi addressed… India doesn’t fear AI… Do you see that with a summit of this nature being hosted in India, it’s a message to the global south?” [86-89].
EXTERNAL EVIDENCE (KNOWLEDGE BASE)
The discussion references Prime Minister Modi’s statement that “India doesn’t fear AI” and frames the summit as a signal to the Global South in [S7] and [S9].
MAJOR DISCUSSION POINT
India’s positioning on AI
DISAGREED WITH
Yann LeCun
Agreements
Agreement Points
Current AI systems, especially large language models, are powerful tools for information retrieval but do not possess genuine reasoning or intelligence, and should not be anthropomorphized.
Speakers: Yann LeCun, Maria Shakil
Current large language models are chiefly advanced information-retrieval tools, comparable to a modern printing press, not true reasoners (Yann LeCun) AI is powerful but not intelligent; we risk anthropomorphising systems that merely mimic human functions (Maria Shakil)
Both speakers agree that today’s AI, exemplified by LLMs, functions mainly as an efficient knowledge-access mechanism rather than a truly intelligent system, and that people tend to over-attribute human-like cognition to it [30-34][25-26].
POLICY CONTEXT (KNOWLEDGE BASE)
This view mirrors the United Nations Security Council’s characterization of current AIs as information-processing tools without real understanding, underscoring the need to avoid anthropomorphism [S29].
AI should be viewed as an amplifier of human intelligence and will require massive upskilling and higher‑level education to realise its benefits.
Speakers: Yann LeCun, Maria Shakil
AI will serve as an amplifier for human intelligence rather than a fully autonomous supermind (Yann LeCun) The AI era demands more, not less, advanced education; demand for PhD‑level scientists is rising worldwide (Yann LeCun) So if it’s about upskilling and ensuring that you’re relevant… the challenge would be to create talent which is upskilled and reskilled… (Maria Shakil)
LeCun stresses that AI’s main value is to augment human intellect and that this will drive a surge in demand for advanced education and continual reskilling; Maria echoes this by asking how India can create up-skilled talent to stay relevant [14-17][95-104][73-74].
POLICY CONTEXT (KNOWLEDGE BASE)
The emphasis on upskilling aligns with policy recommendations for higher-education institutions to lead reskilling efforts and with national AI-literacy programmes in India, reflecting broader workforce-development strategies [S51][S43].
For AI to be widely usable, especially in populous countries, the high cost of inference—largely driven by energy consumption—must be reduced.
Speakers: Yann LeCun, Maria Shakil
Inference costs, driven mainly by energy consumption, must drop dramatically for AI to be practical for billions of users (Yann LeCun) Do you think AI can become that accessible, particularly for a… country as large as ours with 1.4 billion people? (Maria Shakil)
LeCun points out that inference energy costs are a barrier to mass adoption, and Maria raises the question of making AI affordable for India’s 1.4 billion population, indicating shared concern over cost and accessibility [109-113][106-107].
POLICY CONTEXT (KNOWLEDGE BASE)
Calls to lower inference costs echo discussions at AI-for-Social-Good forums that stress affordability for mass adoption in the Global South and highlight the environmental impact of AI energy use [S39][S41].
When affordable, AI can transform key sectors such as agriculture, healthcare, and education, similar to how the printing press reshaped knowledge dissemination.
Speakers: Yann LeCun, Maria Shakil
AI can improve agriculture, healthcare, and education once affordable, as illustrated by smart‑glass pilots for Indian farmers (Yann LeCun) Yes, it is being used a lot in agriculture as well… But when you say about education, will AI assist education… or will they become more AI dependent? (Maria Shakil)
LeCun cites a pilot where smart glasses helped Indian farmers and predicts broader benefits for health and education; Maria acknowledges AI’s current use in agriculture and probes its future role in education, showing consensus on sectoral impact [118-120][121-124][125-136].
POLICY CONTEXT (KNOWLEDGE BASE)
Summit statements and sector reports describe AI’s potential to boost agriculture, health and education, drawing historical parallels to the printing press as a catalyst for knowledge diffusion [S36][S52][S38].
Similar Viewpoints
Both see today’s AI as powerful yet lacking true intelligence, cautioning against anthropomorphising it [30-34][25-26].
Speakers: Yann LeCun, Maria Shakil
Current large language models are chiefly advanced information-retrieval tools, comparable to a modern printing press, not true reasoners (Yann LeCun) AI is powerful but not intelligent; we risk anthropomorphising systems that merely mimic human functions (Maria Shakil)
Both agree that AI’s value lies in augmenting human capabilities and that societies must invest in up‑skilling and education to harness it [14-17][73-74].
Speakers: Yann LeCun, Maria Shakil
AI will serve as an amplifier for human intelligence rather than a fully autonomous supermind (Yann LeCun) So if it’s about upskilling and ensuring that you’re relevant… the challenge would be to create talent which is upskilled and reskilled… (Maria Shakil)
Both recognize cost and accessibility as critical hurdles for large‑scale AI adoption in developing contexts [109-113][106-107].
Speakers: Yann LeCun, Maria Shakil
Inference costs, driven mainly by energy consumption, must drop dramatically for AI to be practical for billions of users (Yann LeCun) Do you think AI can become that accessible, particularly for a… country as large as ours with 1.4 billion people? (Maria Shakil)
Unexpected Consensus
Both speakers see the Global South—especially India and Africa—as future hubs of AI innovation and as a strategic audience for AI advancement.
Speakers: Yann LeCun, Maria Shakil
Countries with favorable demographics, like India and Africa, will become future hubs of AI innovation if they invest in youth education (Yann LeCun) Do you see that with a summit of this nature being hosted in India, it’s a message to the global south? (Maria Shakil)
LeCun predicts that demographic advantages will shift AI leadership to India and Africa, while Maria frames the Indian-hosted summit as a signal to the Global South, revealing a shared belief in the emerging role of these regions [90-94][86-89].
POLICY CONTEXT (KNOWLEDGE BASE)
The strategic focus on India and Africa is reflected in the AI Impact Summit’s location choice, EU-India collaboration initiatives, and broader efforts to close the AI divide in the Global South [S32][S33][S35][S37].
Overall Assessment

The discussion shows strong convergence on three fronts: (1) current AI systems are powerful but not truly intelligent; (2) AI should be treated as an intelligence‑amplifying tool that demands extensive up‑skilling and higher education; (3) cost and accessibility are pivotal for widespread adoption, especially in large developing economies. Additionally, both speakers unexpectedly align on the strategic importance of the Global South for future AI breakthroughs.

High consensus on the nature of present‑day AI, its role as a human‑augmenting technology, and the need for education and cost reductions. This consensus suggests policy focus should prioritize affordable infrastructure, education investment, and inclusive innovation pathways to maximise AI’s societal benefits.

Differences
Different Viewpoints
Timeline for achieving a super‑intelligent AI system
Speakers: Maria Shakil, Yann LeCun
AI will serve as an amplifier for human intelligence rather than a fully autonomous supermind (Yann LeCun) The moderator highlights India’s stance that AI is not feared but embraced as destiny, questioning if this signals a global‑south opportunity (Maria Shakil)
Maria asks whether we are on a path to creating the smartest mind ever and if it will happen in our lifetime [12-13]. Yann replies that a super-intelligent system might appear in the lifetime of some audience members but possibly not in his own, emphasizing uncertainty and a longer horizon [14-15]. The two positions differ on how imminent such a breakthrough is.
POLICY CONTEXT (KNOWLEDGE BASE)
Yann LeCun’s remarks acknowledge the possibility of future super-intelligent systems while emphasizing uncertainty about the timeline, providing an authoritative framing of the debate [S30][S31].
Future of openness and AI‑driven economic boom
Speakers: Maria Shakil, Yann LeCun
The moderator highlights India’s stance that AI is not feared but embraced as destiny, questioning if this signals a global‑south opportunity (Maria Shakil) Whether AI‑generated wealth benefits all humanity is a political, not a technological, issue (Yann LeCun)
Maria wonders whether openness will survive if economists predict a productivity boom from AI [57]. Yann answers that AI progress will be progressive, not a single event, and that the distribution of benefits is a political question, not a technical one [58-56]. The exchange shows differing views on whether openness is at risk and how it should be safeguarded.
POLICY CONTEXT (KNOWLEDGE BASE)
Policy discussions highlight open-source AI as a driver of innovation and economic growth, with calls for open ecosystems to lower total cost of ownership and stimulate competition [S46][S48][S45].
Impact of AI on education – literacy versus dependence
Speakers: Maria Shakil, Yann LeCun
The moderator highlights India’s stance that AI is not feared but embraced as destiny, questioning if this signals a global‑south opportunity (Maria Shakil) AI can improve agriculture, healthcare, and education once affordable, as illustrated by smart‑glass pilots for Indian farmers (Yann LeCun)
Maria asks whether AI will make students more literate or more dependent on AI [124-125]. Yann acknowledges dependence on technology but argues AI will facilitate access to knowledge and act as a tool for education, likening its impact to the printing press [126-135]. Their perspectives differ on whether dependence is a problem or an acceptable outcome.
POLICY CONTEXT (KNOWLEDGE BASE)
IGF 2023 and research on AI in higher education raise concerns about balancing AI-enhanced literacy with risks of over-dependence, especially for children, informing the policy debate [S42][S44][S43].
Unexpected Differences
Magnitude of AI‑driven productivity gains
Speakers: Maria Shakil, Yann LeCun
The moderator highlights India’s stance that AI is not feared but embraced as destiny, questioning if this signals a global‑south opportunity (Maria Shakil) Economists estimate AI‑driven productivity growth at roughly 0.6 % per year, which is modest yet significant for scientific and medical progress (Yann LeCun)
Maria’s line of questioning suggests an expectation of a large, transformative economic boom, whereas Yann cites a modest 0.6 % annual productivity increase, indicating a more restrained view of AI’s macro-economic impact [57][45-52]. This contrast was not anticipated given the generally optimistic framing of AI’s potential.
POLICY CONTEXT (KNOWLEDGE BASE)
PwC’s analysis shows AI-intensive industries achieving productivity growth nearly five times faster than others, while other reports stress the need for redistribution policies to ensure broader benefit sharing [S28][S27].
Overall Assessment

The conversation shows limited but notable disagreement. The main points of contention revolve around the expected timeline for super‑intelligent AI, the future of openness and equitable distribution of AI‑driven wealth, and the role of AI in education—whether dependence is acceptable or problematic. Most of the dialogue reflects consensus that AI will act as an amplifier of human intelligence, that education and capacity building are essential, and that AI can benefit agriculture and health if costs fall.

Low to moderate. While the speakers largely agree on the broad direction (AI as an augmentative tool, need for education, and potential societal benefits), they diverge on expectations about speed, economic magnitude, and policy implications. These differences suggest that policy and research agendas should address timeline uncertainty, ensure openness, and manage educational dependence, but they do not undermine the overall shared vision of AI as a catalyst for human progress.

Partial Agreements
Both acknowledge that today’s AI systems, especially LLMs, are impressive tools for information access but do not possess genuine reasoning or intelligence, describing them as akin to a modern printing press or information‑retrieval system [25-26][30-34].
Speakers: Maria Shakil, Yann LeCun
AI is powerful but not intelligent; we risk anthropomorphising systems that merely mimic human functions (Maria Shakil) Current large language models are chiefly advanced information‑retrieval tools, comparable to a modern printing press, not true reasoners (Yann LeCun)
Both stress the importance of education and capacity building: Maria emphasizes democratizing AI for a large population, while Yann points to growing demand for highly educated scientists and the need for more study [106-107][95-104].
Speakers: Maria Shakil, Yann LeCun
The AI era demands more, not less, advanced education; demand for PhD‑level scientists is rising worldwide (Yann LeCun) And making AI more accessible, something that India believes in, democratizing AI, AI for all, is the theme of this summit as well (Maria Shakil)
Both see AI as a tool that can benefit large‑scale sectors such as agriculture and health in the Global South, provided it becomes affordable and accessible [118-120][86-89].
Speakers: Maria Shakil, Yann LeCun
AI can improve agriculture, healthcare, and education once affordable, as illustrated by smart‑glass pilots for Indian farmers (Yann LeCun) The moderator highlights India’s stance that AI is not feared but embraced as destiny, questioning if this signals a global‑south opportunity (Maria Shakil)
Takeaways
Key takeaways
AI is best viewed as an amplifier of human intelligence rather than an autonomous supermind; true human‑level AI may appear in some people’s lifetimes but not imminently. Current large language models function mainly as advanced information‑retrieval and compression tools, comparable to a modern printing press, and lack genuine reasoning or world‑model capabilities. Productivity gains from AI are estimated at about 0.6 % per year; the distribution of resulting wealth is a political issue, not a technical one. The AI era demands more advanced education and upskilling, with growing demand for PhD‑level scientists; demographics in the Global South (India, Africa) position them as future innovation hubs if they invest in youth education. For AI to be truly accessible to billions (e.g., in India), inference costs—primarily energy consumption—must drop dramatically; affordable AI can transform agriculture, healthcare, and education. The term “artificial general intelligence” is misleading; intelligence should be measured by rapid learning of new tasks, not by static benchmarks, and will be co‑defined by humans and machines, with humans setting the agenda. A major research challenge is building robust world models that handle high‑dimensional, continuous, noisy real‑world signals (the Moravec paradox); LeCun’s new company focuses on this problem.
Resolutions and action items
Invest in reducing AI inference costs, especially energy consumption, to enable large‑scale deployment in populous regions. Increase investment in education and talent development in the Global South, emphasizing advanced (PhD‑level) training to meet AI industry demand. Promote the development of AI systems with real‑world world‑model capabilities, moving beyond symbolic language prediction.
Unresolved issues
Exact timeline for achieving human‑level or superhuman AI capabilities remains uncertain. How the economic gains from AI will be equitably shared across societies and nations. Concrete pathways to create affordable, low‑energy inference hardware for billions of users. Effective metrics and tests for evaluating true intelligence and rapid learning in AI systems. Strategies for integrating AI into education without fostering over‑dependence or loss of critical thinking.
Suggested compromises
Adopt a realistic, progressive view of AI progress rather than expecting a single breakthrough event (balancing hype with measured expectations). Recognize that AI will both amplify human capabilities and require humans to manage and guide its development, sharing responsibility between humans and machines.
Thought Provoking Comments
Maybe in the lifetime of some people here, possibly not in mine… the more interesting thing that we’re going to build is an amplifier for human intelligence. So maybe not an entity that surpasses human intelligence in all domains, although that will happen at some point, but it is something that will amplify human intelligence in ways that will accelerate progress.
Frames AI not as a competitor to human minds but as a tool that extends and multiplies human capabilities, shifting the narrative from fear of replacement to partnership.
Set the tone for the rest of the interview, prompting Maria to ask about the definition of genius and leading LeCun to discuss how AI will reshape concepts of creativity and expertise rather than replace them.
Speaker: Yann LeCun
LLMs are incredibly useful… they are mostly information retrieval systems… a natural evolution of the printing press, libraries, the Internet… they compress factual knowledge produced by humans and give easy access to it. In a few domains they go beyond retrieval, but they still lack true world‑model reasoning.
Draws a clear distinction between raw computational power and genuine intelligence, demystifying the hype around large language models and grounding them in a historical continuum of knowledge‑access tools.
Prompted a deeper dive into the limitations of current AI, leading to the discussion of world models, the gap between symbolic manipulation and physical understanding, and the subsequent question about why robots still lag behind humans.
Speaker: Yann LeCun
Animals have a much better understanding of the physical world than any AI systems we have today… we learn mental models of the world that allow us to think ahead, plan, reason, and predict consequences. LLMs don’t do this, really.
Introduces the concept of ‘world models’ as a missing piece for embodied AI, highlighting the difference between language‑only systems and agents that interact with the physical environment.
Shifted the conversation from abstract capabilities to concrete embodiment challenges, leading Maria to explore how this gap affects practical applications like self‑driving cars and robotics.
Speaker: Yann LeCun
If you talk to economists, they tell you AI will add maybe 0.6 % productivity per year. It’s not a single boom point; the benefits depend on policies and who gets to share them. That’s a political question, not a technological one.
Counters the common narrative of an imminent AI‑driven economic explosion, emphasizing modest, measurable gains and the central role of policy in distributing those gains.
Redirected the dialogue from pure technology optimism to socioeconomic realities, prompting Maria to ask about openness, equity, and the role of governments, especially in the Global South.
Speaker: Yann LeCun
I don’t like the phrase AGI because human intelligence is specialized. Intelligence is not just a collection of skills; it’s the ability to learn new skills extremely quickly and to accomplish new tasks without prior training.
Challenges the prevailing definition of Artificial General Intelligence, reframing intelligence as rapid adaptability rather than a static benchmark, which deepens the philosophical underpinnings of the debate.
Led to a discussion about how to measure progress, why traditional tests are insufficient, and set up later remarks about over‑estimating AI timelines.
Speaker: Yann LeCun
AI will be our staff. Every one of us will be a manager of a staff of intelligent machines. They’ll do our bidding, they might be smarter than us, but we still set the agenda.
Provides a vivid organisational metaphor that clarifies the future human‑AI relationship, moving from abstract speculation to a concrete workplace model.
Encouraged Maria to connect the metaphor to national strategies, leading to questions about talent development in India and the need for up‑skilling and reskilling.
Speaker: Yann LeCun
Long term, innovation will come from countries with favorable demographics—India, Africa—because youth is the most creative part of humanity. But that means we must invest in education; the idea that we won’t need to study because AI will do it for us is completely false.
Links demographic trends to future AI leadership while debunking the myth of AI replacing human learning, emphasizing education as the critical bottleneck.
Shifted the conversation toward global equity, prompting Maria to ask about democratizing AI, cost of inference, and the role of AI in agriculture and education in large‑population countries.
Speaker: Yann LeCun
The cost of inference for AI systems has to come down to become practical for a country like India. Right now it’s too expensive, mainly because of energy costs.
Highlights a concrete technical‑economic barrier to widespread AI adoption, moving the discussion from lofty visions to actionable engineering and policy challenges.
Prompted follow‑up questions about accessibility, the potential of AI in education and agriculture, and reinforced the earlier point about the need for infrastructure investment.
Speaker: Yann LeCun
Usually in technological shifts we overestimate short‑term impact and underestimate long‑term impact. For AI the hype about reaching human‑level or super‑human AI within a few years has been false for the last 60‑70 years.
Provides a historical perspective that tempers current hype, reminding listeners of repeated cycles of over‑promising, which adds nuance to the optimism‑pessimism debate.
Served as a turning point that brought the conversation back to realistic timelines, influencing the final segment about who defines intelligence and the Moravec paradox.
Speaker: Yann LeCun
The real challenge is the Moravec paradox: language is easy for machines because it’s discrete symbols, but the real world is high‑dimensional, continuous, noisy. My current research is about intelligence for the real world.
Summarizes the core technical obstacle—bridging symbolic language models with embodied, sensory-rich understanding—offering a clear research agenda for the next decade.
Concluded the interview with a forward‑looking technical focus, reinforcing earlier points about world models and setting the stage for future breakthroughs beyond current LLM capabilities.
Speaker: Yann LeCun
Overall Assessment

The discussion was driven primarily by Yann LeCun’s nuanced framing of AI as an intelligence‑amplifying partner rather than a rival, his clear demarcation between current AI’s information‑retrieval strengths and its lack of world‑model reasoning, and his grounding of hype in historical and economic context. Each of these comments acted as a pivot, steering the conversation from abstract speculation to concrete challenges—such as the need for world models, the cost of inference, and the importance of education in the Global South. By repeatedly reframing expectations (e.g., rejecting the AGI label, emphasizing adaptability over task collections), LeCun deepened the dialogue, prompting the moderator to explore policy, equity, and practical deployment issues. Collectively, these insights shaped the interview into a balanced, forward‑looking analysis that blended technical realities with societal implications.

Follow-up Questions
How can AI systems develop world models that enable them to think ahead, plan actions, and predict consequences similarly to humans and animals?
World models are essential for moving AI beyond information retrieval toward genuine reasoning and embodied intelligence, addressing current gaps in robotics and autonomous learning.
Speaker: Maria Shakil, Yann LeCun
What research is needed to dramatically reduce the cost and energy consumption of AI inference to make AI services affordable for billions of users in countries like India?
High inference costs limit AI accessibility; lowering them is crucial for democratizing AI benefits across large, low‑resource populations.
Speaker: Yann LeCun
How will the economic productivity gains from AI be distributed across societies, and what policy frameworks are required to ensure equitable sharing of these benefits?
Understanding distribution mechanisms is vital to prevent widening inequality and to guide political decisions on AI deployment.
Speaker: Yann LeCun
What new metrics or evaluation frameworks can capture AI’s ability to learn new skills rapidly and perform zero‑shot tasks, beyond performance on narrow, well‑defined benchmarks?
Current task‑specific tests miss the core of intelligence; better metrics are needed to track progress toward truly general AI capabilities.
Speaker: Yann LeCun
What strategies should India and other Global South nations adopt to upskill and reskill their workforce to meet the growing demand for AI talent?
A skilled talent pool is essential for leveraging AI for economic growth and maintaining competitiveness in the global AI landscape.
Speaker: Maria Shakil
How can AI be integrated into education to improve literacy and learning outcomes without creating over‑dependence on AI tools?
Balancing AI assistance with independent learning is critical to ensure that AI enhances education rather than undermining critical thinking skills.
Speaker: Maria Shakil, Yann LeCun
Will the openness of AI research and open‑source models survive as AI drives significant economic growth, or will commercialization restrict access?
Open AI fosters innovation and equitable access; understanding future openness informs decisions on research funding and regulation.
Speaker: Maria Shakil
What approaches are needed for AI to effectively handle high‑dimensional, continuous, and noisy real‑world signals, addressing the Moravec paradox?
Advancing real‑world AI is necessary for practical robotics, autonomous systems, and bridging the gap between language models and physical interaction.
Speaker: Yann LeCun
What are the long‑term societal transformations that AI will trigger, comparable to the impacts of the printing press or electricity?
Anticipating AI’s broad societal effects helps policymakers, educators, and the public prepare for changes in work, communication, and culture.
Speaker: Maria Shakil
How can the Global South, particularly India and Africa, become leading sources of AI innovation, and what investments are required to nurture this potential?
Leveraging youthful demographics can diversify AI research and ensure that breakthroughs reflect a wider range of perspectives and needs.
Speaker: Maria Shakil, Yann LeCun
Who should define intelligence in the era of advanced AI—humans, machines, or a collaborative partnership—and what criteria should be used?
A clear, shared definition influences research directions, ethical guidelines, and public perception of AI capabilities.
Speaker: Maria Shakil
What further research is needed to validate and scale AI‑driven tools (e.g., smart glasses) for agriculture and healthcare in developing regions?
Demonstrating real‑world impact of AI assistants can improve productivity and health outcomes, but requires rigorous field studies and adaptation to local contexts.
Speaker: Yann LeCun
Why do current AI systems fail to learn tasks like autonomous driving in a few hours despite massive datasets, and what learning paradigms could close this gap?
Understanding this failure points to fundamental limitations in current learning algorithms and is key to achieving efficient, adaptable AI.
Speaker: Yann LeCun
How can we develop more realistic forecasts of AI’s short‑term and long‑term impacts to avoid overestimation and better guide investment and regulation?
Accurate forecasting prevents hype‑driven misallocation of resources and helps set appropriate expectations for stakeholders.
Speaker: Yann LeCun

Disclaimer: This is not an official session record. DiploAI generates these resources from audiovisual recordings, and they are presented as-is, including potential errors. Due to logistical challenges, such as discrepancies in audio/video or transcripts, names may be misspelled. We strive for accuracy to the best of our ability.

Leaders’ Plenary | Global Vision for AI Impact and Governance Morning Session Part 2

Leaders’ Plenary | Global Vision for AI Impact and Governance Morning Session Part 2

Speakers Analysis
Detailed breakdown of each speaker’s arguments and positions

Disclaimer: This is not an official session record. DiploAI generates these resources from audiovisual recordings, and they are presented as-is, including potential errors. Due to logistical challenges, such as discrepancies in audio/video or transcripts, names may be misspelled. We strive for accuracy to the best of our ability.