Uncategorized

Thank you everyone. Thank you very much. Thank you. Once again, ladies and gentlemen, a very good evening and welcome to what promises to be a truly memorable evening. On behalf of Tanla Platforms and our group companies, Carex and Value First, I extend a warm and a hearty welcome to our enterprise and telco customers, our global strategic partners, board members, and our incredible team. At the core of Tanla’s DNA are three enduring principles, innovation, collaboration, and impact. For three decades, this DNA has driven us to build innovation at scale, touching billions of users, and what excites us the most is the greenfield landscape that we have always explored. Along with it, it has helped us work in close partnership with our esteemed customers, our regulatory ecosystem, our telco partners, and the broader ecosystem to ensure that every step has been collaborative and always ahead of the curve.

And lastly, it has helped us ensure that every innovation we pioneer creates a tangible and a measurable impact in the world. And it’s these principles that shaped Wisely .ai, our agentic AI platform built to identify, prevent, eliminate, and bring to the books the growing menace of spam and scam, not just in India, but world over. Today, Wisely .ai is live and delivering real impact at Indosat in Indonesia, at BSNL in India, and with our leading banks in India, safeguarding millions of users in real time every single day. Tonight, we don’t just want to talk about it, we want to bring it to life. So without further ado, let’s bring the story to life. Please welcome our guest for the fireside chat.

A fireside chat from the theme Vision to Impact, driving customer engagement with AI -driven trust. It gives me immense pleasure to invite our host for the fireside chat, who has spent nearly four decades as a distinguished global telecom leader, leading one of India’s most iconic companies as CEO of Bharti Airtel, shaping global mobile policy as a key voice on the board and executive committee of GSMA, and building a legacy that stretches from telecom to digital services and beyond. We are honored to have him as our board member at Tanla Platforms, where his global perspective and vision, continue to shape our journey towards becoming a world -class AI -driven communications enterprise. Ladies and gentlemen, please put your hands together for Mr.

Sanjay Kapoor. A guest on the fireside chat is a seasoned global telecom leader who has not only defined the arc of the industry, but also built it. A career spanning across some of the most dynamic markets across Asia and Africa, he has held senior leadership roles from being CEO of Bharti Airtel Africa and Managing Director of Bharti Airtel Seychelles to serving as a CEO of Orido Group in Maldives and Director -CEO of Indosat Orido before taking on his current role. Today, he leads one of Indonesia’s most transformative telcos, Indosat Orido Hutchison, driving its evolution from a telco into an AI tech co, anchored by a bold vision. of AI for all and a deep commitment to digital inclusion and security for every Indonesian.

Please join me in welcoming the President, Director and CEO of Indosat Oridu Hutchison, Mr. Vikram Sinha. I hand over the baton to our esteemed host, Sanjay, to take it forward and we all look forward to it. Thank you.

Thank you. Thank you for your kind words and welcome Vikram. Before we really get down to asking a few questions from the person who is going to be on the firing range for today’s chat, let me set up a pollute for what we are going to be discussing. We all know that the global economy is rapidly digitizing, but the trust has become its most crucial foundation. Digital payments are expected to surpass $14 trillion annually by 2027, with more than 5 billion people online. In South and Southeast Asia, nearly 2 billion people are coming online at a record speed, driven by affordable smartphones, low -cost data, and national digital infrastructure initiatives. India’s digital economy is projected to cross $1 trillion by 2030, while Indonesia’s has already exceeded $100 billion in GMV.

Yet, this scale brings vulnerabilities. Both markets are facing rising cybercrime, digital fraud, and organized scam operations, causing billions of dollars worth of losses each year. Globally consumers have lost over a trillion US dollars in scams Today’s fraud is no longer isolated phishing It is AI powered, it is cross border, it is automated and industrial in scale This is not just a consumer experience issue anymore It is an economic issue, it is a systemic risk issue, a trust issue and it demands great leadership to combat it It’s my privilege to welcome Vikram who I have known for years and years We worked together at ATL2 He is the President, Director, CEO of Indosat, Orido, Hutchison and is serving over 100 million customers in that country Under his leadership, Indosat has accelerated its transformation into a digital first AI technology access across both urban and rural communities.

Indosat has evolved as an AI tech company and partnered with Tanla, guided by a powerful vision, AI for all. And that’s a very powerful statement that they make. So Vikram, welcome here. And we’ll get down to some sharing of insights and questions to you. We’ve all known about digital fraud becoming more intense. We all know it’s eroding trust. And you as a CEO and with your lens, when did you really move it from being a customer complaint issue to a board level issue? Because there must

First of all, again, it’s an absolute honor and privilege, especially having it with Sanjay. You know, I have a long learning history. So thank you. Thank you, Sanjay. And it’s an absolute honor. I think coming back to your question, let me share with all of you a true story. I’m also on the advisory board of MasterCard. I still remember early 2024, one of the board meeting in London, advisory board meeting, the Asia SCAM and the GASA, Global Anti -Scam Association, presented the data and I was blown off. That report shows that in 2024 itself, 5 billion US dollar Indonesians have lost. What touched me is these are all middle income, lower income women, elderly women. This was an eye -opening data for me, number one.

Number two, the next key highlight, Sanjay, it was every Indonesian, 65 % of the Indonesians are facing spam or scam on a weekly basis. So that itself was a trigger for me that Indosat being such an iconic brand. Let me tell you, Indosat is like BSNL of Indonesia. 58 year old company, first company to connect Indonesia to the world. It became Indosat Oridu Hachisan. But people have lot of expectation. So as a CEO, that was the trigger Sanjay that our role is not only to connect. Our role is to also protect my 100 million customer. And that is where I got very serious that we need to solve this problem for our 100 million customer.

Yeah, I mean, I think every board worth its while today. gets intimidated by this problem that is hitting them. And I’m so glad to hear from you that your board is fully aligned with you on this cause and you’ve been able to convince them to say, I really want to go ahead making some serious investments and changes because of this. And my leading question from here is that I just said that scammers are using AI, voice cloning, automated phishing campaigns, synthetic identities. How did you think of AI in the middle of all this? You know, because it is a new technology. People are still surfacing where they’re headed. But you’ve picked it up as the foundational infrastructure for protecting, you know, this at a national scale.

So tell us about that.

So let me put it this way. I’m a very strong believer of fake it till you make it. So I started talking about AI two years back and I had very little understanding. Of what AI will do. I’m telling you a true story. I was in. many people still struggle with that today. But let me tell you fast forward. I was invited by, I think, Sundar and Google Circle. There were 15 CEOs. This was around a year back. I was on a breakfast table. The joke started by saying that AI is everywhere other than P &L. This is how the breakfast started. But within an hour, I understood companies, countries who have been all in and ahead of the curve and who are solving real problem, they have started seeing value.

So for me, if I have to solve a real problem at scale, and that is where we said that if these scammers are using AI, we have heard many stories and the way they clone their voice, you know, you will be so scared what all are happening. Then we were very clear, we want a partner, we don’t need a vendor We want a partner who can work with us and use AI to solve this real problem And I have to say Sanjay, I think you are on their board, Uday is here We work with 96 vendors, we categorize among them 20 strategic partners But there are 4 or 5 where I invest time, which becomes very strategic for us Because I was trying to solve a real problem, I met Uday and then our commitment was aligned And that is how we wanted to make sure that not only we solve it, we do it in a way which should become a global case study

And with an AI -led model that you have put in place What are the benefits that are accruing to you at a customer level to begin with?

Yes, because as I said, you know, there’s a lot of AI as a toy. We are very clear what we don’t want to do. So now that this is my first showcase, I put it on my quarterly result. And then you have been a CEO, you have reported quarterly after quarterly, until unless you have a substance, you don’t put any example on your investor deck. So if you look at my last quarterly investor deck, we have put it there that with Tanla platform, three things I’ll highlight. You know, quarter four, ARPU grew for the industry 3%, we grew 9%. Number one. Number two, our churn. Our churn, because markets are mature. You know, you don’t have to be over -obsessed with gross assets.

And you know, you have to deliver experience. our churn for serious base greater than 90 days from a level of 3 .6, 3 .7 have come down to 1 .6. And this is just a beginning, Sanjay, you know, because the model is getting trained. And I’m very confident that we will see much more value going forward.

And, you know, from here, being an ex -CEO and being a board member for very many years, now, this question of ROI always haunts every board to say, you’re making an investment in this, it seems to be doing good for your customers. What about the ROI on what you’ve done?

You know, this is, again, and it’s a fair question, you know, investment on AI is not small. So until and unless you see the impact of AI on your P &L, it will not be scalable. So. So very clearly, within six, eight months, we have seen whether it is ARPU, whether it is churn. And the most important thing is, Sanjay, where we lost, if I go back to my last two decades of experience, we as Telco, we were very inward looking. The biggest thing which we missed was focusing on customer love. I think this is very fundamental. This problem which I am solving is so, so fundamental that the role of Telco is not only to connect, it is also to give peace of mind.

Protection is a big statement. And the channel which is getting used is voice, WhatsApp. So you need to solve it for your customer. Otherwise, you have no business.

I mean, I hear you and you are a passionate CEO who believes in keeping his ears in and the ears out. And I see you

I had no idea about Tanla or anything. In fact, first time when Uday came to meet me, I thought it was a startup. I’m again being very honest. But then somebody told me they are solving for banks in India. I think we have to understand if somebody is solving this problem for banks. Because, you know, spam is one thing. But the bigger issue is scam. And these scams which happen, these are small tickets. These are like 50 rupees, 100 rupees, 500 rupees. These are like 50 rupees. and it goes up to as high as $10 ,000. You know, I’m just giving you example. But then I realized that they have done some good work in India. But I have to say, Sanjay, you know, where it moved from vendor to strategic partnership, we were also very keen that my team want to do a bit of an engineering with them.

So we have a full stack AI factory. We have our own GPU cluster. I think there are a few things where we have done before India. Our cluster of GB200, H100 was live. So I told Uday, let’s train the data because see the power of compute and GPU. You know, we all talk about TikTok. TikTok was all designed on GPU. They don’t even use CPU. So if you have to be ahead of the curve, today on that platform, let me give you two data points. Close to 2 billion spam instances. Scam. Clearly threat intelligence protected. 2 .3 million scammers flagged and customers are getting real time as you know we have grown up on the Airtel values I spent 5 days every month in the market I was in a village I was going to the new capital of Indonesia which is Far Flung on my way I stopped my car I saw an outlet I asked him my language still is not good in their language I asked him what do you like about Indosat IM3 he said this Sat Spam Spam Scam it’s solving real problem so again we just launched it on Whatsapp channel also I think Whatsapp is one of the challenge which is maximum getting misused so we have to continuously evolve and this is where Tanla has committed that we will make sure we do it together and we do it properly

excellent you know these fireside chats have a time limitation so we have to keep it up and my stopwatch is telling me we’ve exceeded time already. So let me wind it off. Vikram, first and foremost, thank you for these insights. What stands out from our conversation today is how digital trust moves from concept to reality, which is what you’ve just described. When over 100 million subscribers are protected by AI, when billions of communications are analyzed in real time, and when millions of malicious actors are stopped within the ecosystem, trust is no longer a promise. It becomes an infrastructure. And what Indosat has shown through AI for All is that inclusion and protection are not trade -offs.

They must advance together. So thank you for your insights, Vikram, and it is a pleasure having you today.

Thank you, Sanjay. Thank you.

Can I request both of you just pose for a picture, please? Thank you very much, Sanjay. Thank you very much, Vikram. Wow. Two global leaders, one who defined the yesteryears of telcos across the world, and the other who’s redefining and bending the arc to set the future of telecom leveraging AI. Thank you so much, Vikram and Sanjay, for this scintillating talk. Thank you very much. Our next session, ladies and gentlemen, is… going to be a panel discussion wherein we have Anshuman Kaur Chief Growth my apologies Chief Customer Success Officer of Tanla Platforms who is going to moderate the panel AI for Citizen Protection and Securing the Digital Economy May I request Anshuman to kindly come on to the podium please First of all panelists Mr.

Ratankesh Executive Director and Chief Operating Officer Bandhan Bank Bandhan Bank is one of the largest and the fastest growing banks in India with over 32 million customers served by the across 35 states He is leading multiple functions including technology, operations, customer experience and transformation functions second of our panelists Mr. Bipinpreet Singh founder and CEO MobiQuik a leading fintech entrepreneur at the forefront of India’s digital payments evolution Bipinpreet Singh has built MobiQuik into India’s largest digital wallet with over 180 million users please welcome Mr. Bipinpreet okay we’ll go ahead with the third panelist while we wait for Bipin our third panelist Ms. Nehaji Mahatme director Amazon Pay India a payments and fintech leader driving customer centric digital financial experiences at scale shaping how millions transact seamlessly and securely through the Amazon Pay India app please welcome Ms.

Nehaji Mahatme Ms. Neha Kavya can I request you to just check with Bipin please Maybe Anshuman. Yeah. Oh, Bipin is on his way. To Bipin, ladies and gentlemen, founder and CEO Moby Quick, leading fintech entrepreneur at the forefront of India’s digital payments evolution. Thank you.

Good evening, everyone. As we just heard in this fireside chat, the problem is big. We just heard numbers of over… over $1 trillion being lost in the global economy because of scams and frauds. If you think about India in particular, SMS as a channel, almost 70 % of the scams originate from that channel. And as messaging itself has expanded into other OTT channels as well, and I’ll share some numbers, now 65 billion SMSes are spent monthly in India. Another 15 billion are sent monthly over OTT channels. So when you look at these numbers, it is clear that it is a channel, while it is important and critical, it’s only proliferating even further. So in that context, when you think about, and I joined Tanla relatively recently, compared to the three decades of history as a chief customer officer, and it has been a privilege to see the build and the deployment of WISD AI platform.

And it is an honor to have Vikram here. As a CEO, and there is nothing better to hear that validation directly from the customer. And you heard the impact that it’s having. on the end users in terms of protecting them from scams and spams. In fact, the estimations are within six months of launch, we have protected almost $500 million in estimated losses. And then as you think about where this takes us in the future, scams and scamsters are continuing to evolve. They’re not sitting idle. And so we have to stay a few steps ahead in the innovation curve. That becomes critical. And when they’re becoming more sophisticated, they’re becoming more personalized, and they’re actually probably also becoming more successful at times.

So tonight, I will not, before I get into solutions, I want to focus on the problem. Is the problem really getting better? Or is it getting worse? And why? So we have a distinguished panel here, and they all provide very different vantage points in the industry. We have banking, who sees the transaction risk and, frankly, a lot of regulatory accountability as well. You have fintech, who sees a lot of velocity and scale, and they also are obsessed with customer experience. And then you have platforms like Amazon Pay that have the commerce side, they have the payment side. So they see a lot of behavior signals across multiple parts of the platform. But from a citizen perspective, an average user, it’s a seamless journey.

They don’t operate in this individually. So that’s something that we will delve into. And as part of that, we would love to deep dive in terms of, how the key stakeholders in this ecosystem need to work to thwart this menace that is in front of us. So with that, I want to welcome our distinguished panelists. Thank you for joining us for this discussion. Thank you. So let me start off with you, Pratham. Recent Supreme Court judgment, a couple of weeks back, talked about, I think, 54 or 56 ,000 crores being lost to scams. In fact, they called it dacoity. I don’t think I’ve heard that term lately. I think it was around Chambal and all. I used to watch movies when I had heard that term.

But it is of that magnitude and scale. So the question, Pratham, is why is this still a problem? And what is really not working?

mostly senior citizens and at times even IIT Bombay professors so that’s like the spectrum you can look at it they are being defrauded the second is a lot of customers are now able to open accounts in most of the banking companies they open banking accounts and those accounts are being utilized to siphon off the funds stolen from somewhere and getting routed so there are two parts of the problem in different sense the bigger trouble is the second one the second one is being done willingly with a country having 1 .5 billion population there are a lot of people who would be willing to open accounts and then across multiple banks the India stack makes it pretty simple to have an account onboarded very quickly in just about few minutes and then they go and rent that account and get a fee per month and that number and the lure of making easy money is so high and that’s why it’s so difficult and to me that’s not working So at one end, we celebrate India AI Summit, all the global leaders, heads of states, the big AI celebrities are coming all over here.

And we are talking about our countrymen who are actually defrauding poor senior citizens, and they have got hard -earned money of their entire life, and those are being siphoned off. So that’s very, very sad to see. So I think what is not working is the mindset. It’s not going to stop so soon. So it’s a bit of a philosophical response, but I’ll come to the bit more technical response a little later. The second is our customers getting defrauded using multiple things. That part I think it has got improved significantly because most of the banks now have very sophisticated rule engines. So now as banks like ours, we process. millions of transactions. Like I was just saying that my UPI volume, we are just an 11 -year -old bank.

My daily UPI volume is something like 60 lakhs per day. Now, the volume and velocity of transactions are very, very high. But the good part is that depending on the customer’s profile, the customer routine transaction pattern, we can identify an out -of -routine transaction. If a customer happens to withdraw 10 ,000 rupees from a particular ATM, generally, he or she withdraws from somewhere else, we can say it’s a non -routine transaction. Someone withdraws 10 ,000, suddenly withdraws 25 ,000, we know it’s a non -routine. Somebody never makes a rent payment, suddenly starts making a rent payment using one of the payment channels, we say it’s an out -of -routine. So once you have out -of -routine transactions, we are able to identify those.

Sometimes we prevent those transactions. Sometimes we go back, do enhanced due diligence, and then allow. So that part is working fine, and the tools are getting more and more mature. AI is helping us to build the algo a lot better. So that part is clearly working. So if you are seeing the numbers, the fact that the velocity and the volume has gone up in terms of percentage is coming down. But the mule part of it is very, very scary, which is customer account being utilized as a rent. You know, the other day, one of my employees from the fraud prevention team called up one of the fraudsters, saying that, you know, this is a senior citizen, you called so many times, you tried to defraud, why are you doing this?

So his response was that, can you tell me how much money you make a month? It must be 50 ,000, 70 ,000? I’ll give you double, you start giving me data. The fraudster is telling my employees, saying that, can you share with me more data? Don’t worry, I’ll not tell anybody, just give me data, I’ll give you 70 ,000, I can even pay you 1 ,50 ,000 per account. Now, there again, we are using a whole bunch of technology, including the algorithm in terms of the transaction monitoring algorithm to really prevent that. And I think Carix has developed one which we are implementing now, which is this anti -phishing tool. which has got some very interesting capabilities. If some of you are interested, you could talk to the Karik spokes.

I think that is quite interesting. They sit on the DLT platform. They scan that particular SMS, where is it originated from, some of the links that they provide in terms of collecting data, whether it is genuine or fake. They look at the keywords using some of the AI algo and then come back and use some of the techniques to stop preventing and sending the SMS. To the potential customer who could then get defrauded. So I think these are some of the things which are working. So largely that is what is the spectrum I would see, Ansuman. It is a long answer, but that is what it is.

No, this is fantastic insight and it is obviously great. In fact, my parents are so scared they do not even use ATM cards because of the risk of being defrauded. In fact, they wait for me to come and they will do recharge on the phone, otherwise they are going to the stores to do it. Because this is becoming really scary and especially vulnerable populations, like senior citizens and all, are particularly exposed. let me ask let me turn it over to you Neha right you sit at the intersection you see AI patterns I’m sure Amazon uses AI all over you analyze behavior patterns and all across commerce across payments right but why are we not able to stop scams across the whole journey

so I want to kind of talk about three four aspects that we’ve learned on why it is difficult so first of all I think GAMP is not at a transaction or a payment transaction level I think it is a behavioral journey it starts much before the payment really happens and that’s really where the fundamental issue is and that’s what I think as an industry we need to kind of solve for the second and if I kind of belabor on that point I think the social engineering happens much ahead it is not really when the transaction is happening or the payment is happening the deepfakes, the voiceovers the fake identities the layering of transactions, I think it’s all making it very difficult to stop scams at the point of the transaction.

The second point is the data side was limits visibility. So while as Amazon we have really good data internally on the platform but I think we miss the data of how these social engineering patterns are getting created outside of Amazon and that really prevents us. The third is the human psychology evolves faster than models. So while you can really build models, refine algorithms but can’t beat the offensive AI because the AI is being used on both sides. It’s not that the preventive AI is working, there is also offensive AI. And the offensive AI works unconstrained while the defensive AI has constraints of privacy, constraints of regulations, constraints of customer experience. And so I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s the reason why I think that’s benefits that you need to provide.

So I think that’s the third part and therefore the last part which is really the crux of the point is that AI is helping detect anomalies. It is not helping detect the malintent or the behavior and unless and until I think we solve the malintent or the behavior I think the scans, the

It’s a fantastic response. And basically what you’re saying is no one institution is seeing the whole journey. You’re seeing pieces of it. But there’s a lot of parts that are interconnected. Let me bring Vipin into the conversation. Vipin, you see a lot of transaction velocity and scale. But you’re also focused on customer experience because that can become a friction point if you do go overboard in times to protect. How does AI come in? How do you calibrate your model and AI to balance that potential friction that potentially can impact your growth and legitimate customers versus protecting from fraud?

Thank you. Thank you, first of all, Ashutosh and Tanla Solutions for inviting me here. It’s a privilege. We are also customers of Tanla, so happy customers. Thank you. I think when it comes to AI and the usage of AI for fraud, I want to give some perspective with respect to first the kind of fraud. So we operate in payments and financial systems. And I think what’s happened in the last 10 years is the financialization and digitalization of financials has happened at an exponential scale. And so many different entities have gotten interconnected, just like what you are saying, that a loophole in just one place is sufficient to create fraud and scam throughout the ecosystem. So one thing we have to be very clear that one entity cannot control scams.

It’s very, very difficult. Because like the experience that we have at MobiQuake is 99 % of whatever scams that our customers complain of are not the money. It’s not the money that they get stolen out of MobiQuake, but it is actually stolen out of some other bank. and come into MobiQuick. So we are the recipients and, you know, and we get the complaints that the money has come here and we need to take action, whether it is coming through UPI, coming through credit card fraud and all those things. Now, therefore, you know, the standards of education and the standards of 2FA, second factor authentication, they need to be there. Perhaps are not equally enforced. The awareness and the education is not equally enforced.

And that brings me to the second point, which is that, you know, the AI, the scams have also become very, very sophisticated. Right? In our company, and we are a fintech company, we employ so many smart people. There are people who have fallen fraud to scams where they got a WhatsApp from me asking to buy gift cards with my photo and people have bought it, you know, without trying to verify or seeing what the number is. On the WhatsApp thing, because, you know, this is… It’s becoming, with AI, it’s becoming harder and more sophisticated. You know, the modus operandi of the scamsters is becoming extremely smart. And they are getting very, very smart at understanding the profile of the customer.

It’s not, they don’t target everyone. They have a very clear idea on who is likely to fall for a scam. So, there is need, for example, and there is a, I think there is an effort going on at the RBI’s end. And the government said to create a intelligence body which will share data across the entire payments ecosystem. I think it’s called Digital Payments Intelligence Authority or something like that. And that is something that’s extremely important because until data sharing starts to happen at an India level scale, you cannot identify. I can identify patterns. I can identify a pattern which works for me. But, you know, the scamsters will get smarter because they will keep changing their MO outside of Mobiquit.

So it’s very very difficult to keep adapting to that. So there is a national level initiative that is required. Third thing which I want to say is on the LEA front. I think the almost the entire country all the police everyone knows where the scams are. And this I am not able to understand why no action gets taken. It is the same places. It is the same origins. Right. But somehow no action gets taken. And I think until there is fear of law until enforcement has happened that that which will fraud you know payments fraud will come. At our end you know what we are doing is we are creating our own in -house models.

So as far as technology is gone in terms of machine learning and now with AI. We have created our own models. We have created our own models trained on our own data sets because they work best. for our kind of transaction pattern. But they may not work best for the kind of solutions that other companies may need. In fact, we have explored solutions of fraud and machine learning from other companies, but they have a very poor performance because they are trained on some industry -level data, which is not the same patterns that we get. So, at our end, we are a tech company, we can adapt. But I cannot say the same for all the entities, at least on the financial domain.

And that’s a big problem because until there is a national… And I think the regulator, especially RBI, is very, very concerned about this. I think we have heard now, and in my recent conclave where I went to, that they are saying that enough of making transactions easy. Now we need to go in the reverse direction. Now we need to make transactions a little difficult so that there is some friction because otherwise, you know, people are losing money.

It’s great points. And as you said, I think the silos of data that we have, and in fact, you talked about training the models. In fact, you know, again, when we went to Indosat as well, you had to go and train them, even including language nuances as well. So these things become critical in terms of adapting. You mentioned about like RBI initiative and all, and I direct this to you, Ruthen. Who ultimately owns the responsibility of protecting citizens at national scale? Because can banks do it alone? Can RBI do it alone? Or are you dependent on upstream and downstream signals, upstream signals like that you get from telcos because a lot of these things originate from the channels?

How should responsibility be structured? to protect people and to help.

I think Vipin spoke about that point that look, for a fraud to happen, we know that somebody would have gone to an e -commerce platform to make a payment and that payment is coming through a payment channel and the account is held in Bandhan Bank and the payment is made to let’s say for a product or through some platform the payment is made to an access bank credit card. Now and then the fraudster is actually sitting somewhere out there who is pretty much somebody amongst us. Now I will give you a very funny story of I lived in Mumbai for 20 plus years the local trains are full of used to be those days pickpockets. I came from I was living out of India and came back and I was going to meet a friend of mine that was my first trip in the local train and my purse got stolen and he said don’t worry how much money I said that’s okay but I had credit cards and all of that so somehow managed to block those cards and he says let’s go to the police station I said but I had my identity cards there he says don’t worry it’s fairly ethical pickpockets in Mumbai you go and tell the police you have to tell which train which local train from what to what and what exactly it must have opened probably so I said from Borivali to Andhri it must have happened in between it was a whatever 950 local train so after two days the police called me and handed over me the identity cards so so there was nothing else there but then I got back the identity card so I think that’s like the police has an ability to really find out who the people are and I find it quite baffling to figure out that this fraudster is somewhere there the telephone numbers are issued multiple of them by a telecom authority the other pan cards were issued which are used to get this sort of account opening and video QIC and the telephone numbers yet we are not able to find out who these people are and technology has to protect all of that which as all of us are saying that we are trying our best to protect and the bad part and the sad part is that whenever there is a fraud happens and a customer goes to the regulator like the ombudsman and it says 20 lakh fraud it’s okay which bank is went from it says it went from HDFC bank where did it fall first it fell in Bandhan bank okay two of you together 10 10 lakhs pay it and be done with this that’s easy isn’t it I mean we of course are regulated entity regulator has no choice but to sort of do whatever best they can do and which we do and that’s absolutely fine we must have had some lacunae in our process but how do the whole chain has to work together I mean including the citizens instead of becoming gullible in terms of having the awareness about banking product the banks and the payment companies and the country has to create more awareness the police, cyber police and the local police they have to work together minister of home affairs is working very hard in terms of really making it happen so it’s an ecosystem problem and I think if all of us come together, all of us create more awareness and make it really ruthless probably that’s the only way to happen otherwise it’s no easy

Thank you, thanks a lot I am told time is up I wanted to solicit two questions as well from the audience but in the interest of time I will just summarize this session in fact we have all talked about going kind of end to end it’s not just identification using the AI models but the prevention, the elimination and ultimately holding the scampers accountable with law enforcement and that comes a big part there is law and the enforcement of the law they can sometimes mean two different things so ultimately one of the things I hope just one more point in the name of hyper personalization the amount of data that you collect and the amount of data that you collect and the amount of data that you collect and the amount of data that you collect and we have an ability to go back to Neha and say, you know what, you’ve been searching for a home.

I can tell you which is the right home for you. That’s great. Neha feels delighted because she can actually choose the right house. But the same data is getting misused to do other things. And as much as I can collect data as a bank or a real estate company, Proster can also collect the same data. So as you said, the AI is on both sides and they are like more offensive. And so it’s a question of who stays a few steps ahead of the other, right? What is striking from this discussion, hopefully, and I’m summarizing is like, as you can hear, everyone is doing something, right? Oh, please. Please go ahead. We can take one or two questions quickly.

Sure, please. Can you please help the gentleman with the mic? This is not the best way.

So there should be some integrated approach. So I think the government of India is already working on that and they have a digital payment intelligence platform. So it’s not for that purpose or you are referring to some other issue.

Sorry, is that a question? I think the question is there are some government initiatives.

Initiative is already there. Is it not enough? To have an integrated model for this fraud protection and other thing. Because you said. financial institutions are having their own trend model and accordingly they are protecting their customer this thing but if we have that integrated model and government of India through their one company there they have initiated with the collaboration because that RBI is working on mule hunter through RBI mule hunter is the initiative yeah from RBI in his hub so that is already all the banks are doing and they are doing with their own three month five month data individually not a complete umbrella like that so for that everything is coming in the this digital payment intelligence platform so with this initiative Anna that it is so which you have referred will not be addressed to

yes yes absolutely I think you know at least I feel that there is going to be there’s a strong potential because for the first time data across the financial economy is being used for the first time and it is being used for the first time ecosystem will come together in one place and I think that is a big deal and then once that data comes together then hopefully the best people will work on it and understand patterns at a national scale because the problem in digital is everything is connected so you have to study it in an integrated manner and I am very

Thank you for that question and we are obviously hoping that all of these show results but at the same time we are talking about national but scammers are not limited to even national geographies they are international as well so the scope and the breadth the surface area of the threats are only expanding so we have to really look beyond and if I may from my personal experience one of the things is in the world of AI data is actually the differentiator not so much the models because all public and the willingness to share the data itself is a potential barrier real time data and this is where it is not about just banks and financial institutions it’s also potentially telecom, right, because they see a lot of initial signals in terms of messages being sent and communication and so forth, as Vikram just talked about as well.

So let me summarize again this session.

Can I ask one question?

May I request you to take it offline, please, because of the time constraint, if you don’t mind. Thank you. Thank you for cooperating, sir. It’s great to hear. I’m sure there’s a lot of interest in this topic and it shows the resonance of what we are discussing. So, again, I think to summarize, as you said, the attack surface is all interconnected, but our defense is right now fragmented. And therein lies the opportunity. And while the next frontier cannot be just more smarter individual AI model, it has to be really coordinated intelligence. And that obviously has to happen real time. It has to happen across the ecosystem. And it has to happen within the guidelines of a national level trust architecture.

So with that I will want to thank you to all the panelists and also for all of you to participate in this discussion and really contributing to this to shaping what the future looks like because this is really not just trust, this is the foundation of creating the digital economy and the growth that underpins it. So thank you so much. Thank you very much all the panelists and Anshuman may I all request you to stay on the stage for a quick photograph. Wow. From using technology for transaction monitoring and layering it with AI to solving for behavioral intent and offensive AI along with solving for customer friction on customer experience by layering it on technology and captive models along with regulatory tenets.

It was a very insightful and a very meaningful panel. Thank you very much each one of you. Thank you. Thank you. Thank you. Thank you. Thank you. Thank you. Our last session for the evening is one of the most interesting ones. It’s what we do on our third element or the third pillar of DNA, that is impact. It’s the impact spotlight, wisely .ai, our client’s perspective. Very few leaders in India telecom landscape carry the depth of experience and the institutional weight that our next speaker brings to the stage. As chairman and managing director of BSNL, he has orchestrated one of the sector’s most compelling turnarounds, driving the rollout of India’s first indigenous 4G network and restoring the organization to a clear path of growth, profitability and purpose.

With over three decades of service spanning tri the government of Tamil Nadu and an advisory role to the government of Uganda, recognized with the Visist Sanchar. For distinguished service and a gold medalist in electronics and communication engineering, he remains one of the most. consequential voices in India’s telecom and digital governance story. Ladies and gentlemen, please put your hands together for A. Robert J. Ravi, Chairman and Managing Director, BSNL.

important step we are thinking about, that’s what I was talking about. On the network side, can I bring AI? In bringing AI in the network side, I can even get patents, customer patents, calling patents, network initiatives. We were able to even see exactly how and where most of the complaints were happening in the network. And this also helped me for tweaking my entire setup today, where I’m very sure at the end of probably the study and whatever research we are right now going on the AI, as a user, if you are a BSL customer, you can intelligently speak to your RAN. When I speak, when I tell intelligently speaking, it could be various things. So you can have…

you can request a specific dedicated data specific dedicated voice traffic that means today I am in this place I need to video stream I need a 1G so not a 1G at least under 10 throughput available all the time it will be made possible so that type of a user enabled platform which we are building that will control not only from the customer angle from the network angle if this becomes successful today no customer in future when this reality actually comes in no customer could be so easily fished or smacked that’s the way which we are trying to go ahead of course going into what exactly happened was the last one we can see that the user impact we were able to authoritatively say that that so many number of connections to close to 280 million spans we have installed today.

This is on one side. Now we are integrating this particular aspect on a customer experience platform also. How do we benefit out of it? In my customer experience today also, we have brought in we have something called the AI Vani system. It’s just a Vani which comes and says, whatever you want, you can speak to the particular agent. And then we brought in something called a BSNR recharge expert system. It’s a complete AI driven. So when user suppose even as a user because spam suppose when he stopped the spam for the SMS side, next thing we have to concentrate is on the data side, which again I’m talking of course we are speaking to you all how do we do for the data side.

Data is not only on whatsapps or social media how can we expand the horizon of this particular area that’s where what we thought can I build in intelligence into the system itself so when you wanted to do probably recharge or even it works as a worm in the network to easily identify the sites which needs to be blocked which should not be available for my customers such a sort of independent intelligent network needs to be built in which we are targeting up that’s the second pillar the last pillar before I wind up is going to be the rural thing the rural side with Bharatanat coming in at a very big space rolling out of Bharatanat’s network we are trying to see when you’re closing closely going close to the customer at the end we are seeing lot of fractured coming in can I put edge data centers using this edge data centers can I really run what we call even SLMs.

Today we talk about different LLM models. These LLM models require a lot of information. The data is a key engine for it. And we are all travesty to see why should I share my data. So this is the next concept of what we bring in what we call a federated learning. So your data resides with you. I just learn your data and I federate over it. So all this is possible when I go to the rural end of the edges. So there I will be able to protect the customer to a next level. I am very sure I think we can keep talking on this very interesting topic but since the time is short I thank the organizers for giving me opportunity.

But my request to you all still there is lot of work to be done. Unless we have built a system where we confidently say our citizens that you are 100 % safe in my network. our job is not done. And this is possible only when we bring in technology and play it across in a platform which really intelligently builds this network. Thank you.

Thank you, it’s been a wonderful evening absolutely thrilling to have two CEOs exchange and share with the audience the real life problems and how they converted into an opportunity that is going to shape the future of telecom in one part of the world followed by a panel thank you Anshuman and thank you once again to all the panelists who made the effort to come in and share their own perspectives on what could be changed structurally from a regulatory perspective, from an ecosystem collaboration perspective to customer experience without friction and lastly dear CMD Mr. Robert Ravi for sharing the deep collaboration that BSNL and Tandla platforms have come into and are trying to set a lighthouse to what could really be a customer experience driving safe and secure customer transaction thank you very much it’s been a true honor and a privilege to host everyone here thank you I am very thankful on behalf of Tanla platforms, our group companies Carex and Value First for hosting you all here Thank you very much Thank you Thank you.

I’d like to welcome you all to this session titled Responsible AI from Principles to Practice in Corporate India presented by Adobe in association with FICCI. India stands at this defining moment in its digital journey as AI becomes a powerful engine to innovation and productivity. But the real differentiator, is it about how quickly do we adopt AI? No. It’s about how responsibly we deploy it. Trust, transparency and accountability are no longer optional. They are foundational. And that’s exactly what we are going to be talking here today. The conversation will center on advancing safe and trusted AI in the corporate landscape. To set the context and to get us started, it’s my privilege to invite Andy Parsons, our Global Head for Content Authenticity at Adobe.

Andy, over to you.

Thank you so much. Whenever I speak publicly, they have to lower the microphone. They never raise it. I don’t know. Thank you. Thank you so much for having me. I know we’re against a tight time frame here, so I’m going to speak for just about five or six minutes and then turn it over to our wonderful panel where all the action will be. I’m Andy Parsons. I run the Content Authenticity Initiative at Adobe, and I’ll tell you a little bit about what we do. I think it’s a good example of how responsible AI can be adopted, promoted, and effective in enterprise. I want to start with a simple observation, and I promise not to talk too much about policy because I’m unqualified to do that.

I’m a mere engineer at Adobe. But 2026 is certainly going to be the year that responsible AI becomes a responsibility and an opportunity. I’m going to talk about that in a minute. I’m going to talk about that in a minute. I’m going to talk about that in a minute. of you in this room. This means it stops being a slide in a deck, and it will now be sort of a piece of our compliance strategy, but also, as I said, an important opportunity. The EU AI Act’s enforcement provisions take effect in August, as does the first law in the United States in California. And of course, we have the new IT rules here in India on SGI, and India is actively shaping its own path.

And it’s good to be here while that’s happening and talking to many of you about how AI and transparency can be effective in the very short term. So the question for everyone in this room has changed, I would say, this week and certainly will continue to change this year, from should we be responsible with AI? I think that debate is well settled at this point. But can your systems actually prove that you have been responsible with AI, and how do you go about doing that? And for all of us, what does it cost in terms of implementation and day -to -day usage? So we want to position responsibility, responsible AI today as a leadership and operating must and discipline rather than a regulatory obligation, although in 2026, I think it will be both.

So this shift from principles to provable practice is the theme of our panel today. It is what I want to spend just a couple of minutes on. I won’t speak about this from a theoretical perspective. I’ll tell you a little bit about what my team at Adobe does and why I think it matters and perhaps sets an example for others. The trust crisis with AI is real and it’s concrete and it’s happening every day to our children and our businesses and ourselves. Anyone who consumes media, especially across a cultural vastness and disparate languages that we have here in India, is certainly well aware of this. Generative AI has done remarkable things for creativity. Of course, we live and breathe that at Adobe every day and you’ll hear more about that from my colleague Pratibha in a moment.

But it’s made the trust problem absolutely impossible to address. So I think every enterprise in this room now produces or consumes AI. generated content at scale, whether it’s marketing assets or news or customer communications, product imagery, et cetera. The volume is extraordinary, and it’s absolutely accelerating every day. The potential is huge if the risks are managed, which is what we’re here to talk about today. In fact, I’d argue that we can all go faster with our adoption of generative AI if we do it responsibly and critically put in place a foundation for that responsibility now rather than wait and be reactive. India has the world’s largest digital population. Of course, you all know that. Hundreds of millions of people consuming digital content every day.

In this environment, synthetic content and AI -generated misinformation are not abstract risks. They’re real ones and operational ones for all of our businesses. So the corporate responsibility question is if you deploy AI that generates or modifies content, as almost all of you, I would imagine, do now or will shortly, can you demonstrate what was made, how it was made, and by which models and products? And that’s what my job at Adobe is. I’m very privileged to provide some leadership. I’m very privileged to have leadership in the C2PA, which hopefully many of you have. have heard about this week, which is the Coalition for Content Provenance and Authenticity. And our sort of piece of the responsible AI landscape is around transparency for content that’s generated by our tools, but also providing a global standard so anyone can do this without licensing totally free.

So perhaps this is a case study, and I’ll just spend my last couple of minutes on this. At Adobe, we decided five years ago now, around the time that I joined the team, that responsible AI via content transparency wasn’t a feature that could be grafted onto our products like Photoshop and Premiere, digital experience products, but had to be baked into the tools kind of at their very core. And we went about doing that. While we considered how to do it, we contemplated developing a global standard with partners like Microsoft, BBC, OpenAI, Sony, and many others. And now we’ve done that. Five years later, there is an open standard called the C2PA content credentials. If you browse LinkedIn and see this symbol, you have the C2PA content credentials.

I’ve encountered a content credential, and it will provide transparent context about a piece of media, whether it’s a video, audio, or video. or image. And as we see adoption increase, we realize this is perhaps a model for AI adoption in a responsible manner based on open standards and interoperability. So we are built on an open standard. It’s truly a cross -industry coalition that includes all the companies I mentioned, meta camera manufacturers like Sony, Nikon, a growing number of media organizations, perhaps some in this room, and also silicon manufacturers like Qualcomm and others. And the goal is for an infrastructure layer for content trust that anyone can adopt. It shouldn’t be owned by any one company.

It should be standards -based. It should not be proprietary, but available to everyone. And this shared philosophy I think is especially important here in India. And it should be conveyed by working code and products that leverage that working code, not theory and slides in a slide deck and statements on a website. So what are the principles that the Content Authenticity Initiative and the CTPA reflect? Transparency. Provenance information travels with assets when you make them. You can build an entire genealogy tree to understand where your corporate or daily content comes from, what models made that content, what products were used, what cameras were used. Simple ideas like knowing that a photograph is actually a photograph and not generated.

These are simple ideas. I’d say we’re well overdue having this ability, but we need it now more than ever. Accountability. You can trace those AI models, understand what was used, how it was used, and thereby understand the prominence if you wish to access it. We sometimes think of this as nutrition labels, which you heard PM Modi mention yesterday in his remarks. If you walk into a store in most democratic nations in the world, you can pick up a piece of food. You can decide if it is healthy or not healthy for your children. No one’s going to stop you from buying it or consuming it, but you have a right to know what’s in it, and we think that digital content has to have that same foundation of transparency.

Last, inclusivity. Perhaps this matters enormously for this room. Our standard is open and free. An independent creator here in India. We can apply the same kind of provenance at the same zero cost. as a Fortune 500 enterprise. I won’t say that everything is roses and happiness. There are challenges, and I’d be remiss if I didn’t spend 30 seconds on the challenges that standards adoption and AI transparency and responsible AI will present you, and I’m sure you’ll hear more about that with our esteemed panel. But adoption is uneven. Many social media platforms strip metadata and remove that transparency when content is uploaded. Legislation may help here. We’ll see. Consumer awareness is still very early. I saw many of you squint and look at this pin because you probably haven’t seen it.

Our hope is that you will see it ubiquitously across the world starting in 2026. And because consumer awareness is early, user interfaces are also quite early. But there are telephones, there are phones, cameras that are now beginning to show this symbol as a symbol of transparency. And the business case for provenance has been challenging. I have often said that doing something that helps perhaps preserve democracy and democratic discourse is maybe not a good way to make money. I’m not sure if that’s true. I’m not sure if that’s true. I’m not sure if that’s true. But it is critically important. And now we’re seeing that change as enterprises have to be compliant and have to be leaders when it comes to AI transparency and responsibility.

What’s changing, of course, is regulation. I view regulation, like what’s happening here in India, as a catalyst for good practices. We don’t want to be reactive, but we do want to help catalyze the responsible AI ecosystem. You need standards, not just principles. Responsible AI commitment on a website is a starting point, but not a meaningful milestone. And that’s really the difference that we’re here to talk about today, the difference between principle and practice. I think you need cross -industry infrastructure. Techniques you use for responsible AI should be interoperable, open, and standardized. And I think you have a long track record here in India of mobilizing things like UPI payment infrastructure that not a single bank could do or a single government agency, but required massive scale cooperation for openness, standards, and most importantly, interoperability.

And last, enterprises. Like many of yours in this room, that are willing and excited to go first that really look at transparency and AI responsibility as an opportunity, not a requirement and a consequence of AI. So let me bring this all together before I introduce our panel. The responsible AI conversation has matured, and now we have to move it to pragmatic implementation. It should favor fairness, accountability, transparency, privacy, inclusivity. We’ve all heard these words over and over again. And in 2026, I think it’s absolutely critical to put them to work and demonstrate that all of our enterprises are doing those things. And that’s the hard part. That’s going to look different depending on whether you’re operating aviation systems where safety is critical, running payments infrastructure at massive scale, governing AI across a conglomerate, or building creative enterprise tools as we do at Adobe, which is exactly why I carefully selected those industries, because we have representatives from all of them on our excellent panel today.

So we’re fortunate to have leaders from Air India and PCI, RPCI, and the U .S. Department of Defense, and Adobe. each of whom is navigating and translating the sort of remarks that I’ve made in different ways for different outcomes and providing, I would say, exemplary ways to find their pathway through the challenges I mentioned. And we have a fantastic guide for this conversation, Shantari Malaya, editor at the Economic Times, who covers these infrastructure and societal sort of breaking points every day in her coverage of our various industries. So I’m going to stop there. Thank you so much for having me. And let’s get on to our panel. Shantari.

Thank you so much, Andy. That was fantastic. It set the context very rightly for the next discussion coming up. So a warm welcome once again from my end. My name is Shantari Malaya. I’m editor at Economic Times. Welcoming you all to the panel. Right at the heart of the AI Impact Summit. It’s been spectacular. And to take this discussion forward, we are looking at responsible AI. very momentous time in India. India is really charting the course for the world. And as we look at building trustworthy and inclusive AI, industry, infrastructure, policy perspectives, it really becomes important to know what some premium leaders in the country are thinking about this. So this dialogue will examine some parts of responsible AI and how it’s really going to shape, reshape enterprise strategy.

So to help me with the discussion, may I have the pleasure of inviting Dr. Satya Ramaswamy, Chief Digital and Technology Officer at Air India Limited. Warm welcome, Satya. We have Mr. Vishal Anand Kanwati, Chief Technology Officer, National Payments Corporation of India, NPCI. We also have Mr. Amol Deshpande, Group Chief Digital Officer and Head of Innovation at the RPG Group. And I have the pleasure of inviting Prativa Mohapatra, Vice President and Managing Director of Adobe at India. This is a fantastic lineup and we shall get some very sharp insights from our panelists over the next half hour or so. So at the very outset, if you really see building trustworthy and inclusive AI is all going to be about how responsible AI principles, whether it’s fairness, accountability, transparency, privacy and inclusivity, is really going to actually realistically be translated into enterprise strategy frameworks and how we are going to go about it.

Right. So, Amol, let me call you into the discussion. Warm welcome. I’m keeping a tight watch on the timer right behind us. As we know, this is a summit of scale and we really need to help the organizers to clock good time. So, Amol, very quickly, as I invite you into this discussion, you represent enterprises at, you know, while you are participating. As part of the RPG group, you also represent enterprises that are deploying AI at scale. and many more that are just finding their footing as well. So there’s a huge spectrum and diversity in the kind of organizations that you are representing. Two things for you very quickly. One is in large multiple, you know, businesses such as the RPG group, how are you really preventing responsible AI from really becoming something like a just mere centralized compliance exercise or something that’s on the other flip side becoming a fragmented business unit wise, you know, checklist.

So there are two risks that can happen, right? In a group, in a conglomerate, large scale or decentralized. So how are you really looking at the balance here? And how do you really see your role in an industry body as well? So all yours.

Thank you, Shantiri. I’m very happy to be here. Thank you for having me with the, you know, esteemed panelists here. You ask a very pertinent question. You know, it’s to be or not to be, kind of a scenario when it comes to AI, but that not to be is not really a choice. and it did mention about the responsible AI and I would take a little stab at peeling and looking at where the responsible AI comes from when it comes to industries. It comes across all those five layers of the AI when we are looking at it. When any enterprise is looking at deploying AI and being responsible for the usage of those elements, whatever you are using for, the responsibility needs to be there at every layer.

It’s not one or the other. It has to be an orchestration of all the things. So far, AI in its very nascent forms had been a thing of center of excellences, trying use cases and seeing what it is happening, but now it has come to a scale. And when it comes to enterprises and manufacturing enterprises, which we have significantly higher share in terms of as a consumer of AI technologies, there is a very clear -cut view on how it is to be done. So you need to provide the playground for the enterprise, to operate function with agility. The other part is about people. The people is a very, very important stakeholder in the whole thing.

We are moving from generative AI to AI ML, more complex scenarios and agentic AI. So people is a very, very important aspect of it. The choice still remains with human sense. The awareness is important and enterprises like us spend a significant amount of time and effort in building that skill sets amongst the value chain of all the people who are doing it. Last but not least is the process and governance part which comes with it. It’s more of a guiding principles which need to be given so that it gives an opportunity. It’s more about, you know, if one can say it will be more of a bring your own AI kind of a scenario in every function.

You cannot provide one solution. One size doesn’t fit all. So when you are dealing with such scenarios, a scalable, safe environment with protected with guardrails is a key thing for us. Orchestration and getting a scale. That’s something which is there. And that. But those templates are being exercised, practiced within the enterprises, practice it in a very diverse group like RPC at ourselves. And then that can be deployed at multiple

Absolutely. So as you said, one size doesn’t fit all. Right. And I liked your coinage of bring your own AI. So let me quickly bring in Prativa here. Welcome, Prativa. So Adobe is consistently really positioned responsible as a product philosophy and a trust commitment. You may just have to switch that on. So Adobe is constantly spoken about responsible AI as a very fairly large commitment. So how do you really look at all these principles manifesting or panning out in terms of operationalizing it among all your product teams? And sending it out. as a strong positioning internally? And at the same time, as someone who’s led a lot of industry conversations, where do you really see enterprises struggling to get these things right?

So all yours here.

Okay, thank you. So I think Andy said the context. And since we are here not to learn about the principles, but the practices, I think everybody should go back with certain practices. So the first practice of AI governance, which we practice, is art. Which is accountability, responsibility, and transparency. So if every person goes back to their organizations and talks about art, which is our philosophy, that’s practicing philosophy number one. And we actually have been doing this for our own products for a while now. And of course, we are in the business of content for a very, very long time. And now the same content is becoming the currency which everybody’s debating. So our principles have been there for a while.

But. How it is actualized? and let me say how it’s translated into our products. And by the way, it’s in our products, it’s in our methodologies. Every new product that we have goes through a very strong, secure methodology with hundreds of steps inside it. So there’s principles embedded into how we create stuff. But a couple of examples. Firefly, which is our Gen AI tool, actually embeds what Andy said, those content traditions. Nutritions. It is anything that you have being generated out of this product will have that nutrition level. So when an enterprise is using something in Firefly, you can be super confident that you will not be violating any law, you will not be getting into any liability issues.

Because how you do it is by feeding. Because AI is all about the input and output. So the input has to be something which will not land you in the trouble. You cannot take somebody else’s data. So here it is everything licensed. So it goes into the models, and what you create, the output which comes, then you have to test that output. With that output, will we be accountable? Will we be responsible in showing the transparency of how this was created? So I think that loop has to be created in using any AI. Firefly is an example. Let me talk about Acrobat, which everybody has. I’m sure 100 % of you have PDF files on your phones or on your machines.

So Acrobat has this new feature called Acrobat Assistant. It is agentic, but we have so many chatbots in the market. But when you come to an assistant like Acrobat Assistant, it is following the same principles that PDF was used to be created. So everybody is confident when you’re using PDF. So today, you would have read in the papers recently, the Supreme Court was very worried. that there were certain lawyers who had the petitions which had reference to cases which do not exist or it had certain laws stated which are fictitious. So imagine somebody’s created certain content using some sources which were not authentic. Now, if you use acrobat kind of products for that, you feed the data or you feed files from your own machine.

So you’re confident that what comes out of it, you can go back. So wherever there is this usage of high -stake output, enterprise -grade, you have to look at this input -output process and follow the philosophies within it. And I think for every enterprise who is doing that today, they really have to already, Amal talked about the people process technology. I’m sure every organization today has a legal team, has a compliance team. But these teams have to re -opt and re -design and re -design to talk about AI compliance. Enterprises… they do business strategy, they have ethical strategies and then they have regulatory compliance. All the three anything that you do in AI, ensure that you tick all the three.

If you miss any one, you might not be ready for the future. So that’s how I see it.

Absolutely. So I guess the thread of most of what AI now entails is about are we not moving fast enough or are we moving so fast that we’re not really able to own the operational consequences of what we’re trying to do as well, setting out to do as well. So great point on that Pratibha. I’ll circle back to you, time permitting. Let’s see how best we can get back. Dr. Satya, calling you in here. So aviation volumes, landscape, scale, I mean you name it, it’s all there. So how are we really looking at balancing AI driven innovation really? Where you’re looking at regulation, you’re looking at accountability, you’re looking at operational efficiency. At the same time, you cannot really compromise on user and customer experience.

How do these things really fall in place in terms of vision and metrics?

Thank you, Shantiri. Since the audience is international, a real quick introduction about Air India. Air India is India’s national flag carrier. We operate about 300 aircraft, and we carry about more than 100 ,000 customers a day. And we have a few hundred airplanes on order. So once they are delivered, we will be one of the biggest airlines in the world on the size of one of the large three American carriers. So we are building it up to an airline of scale, and it brings about very interesting challenges that we talked about. So let me illustrate the way we handle it with one of our own examples in generative AI. So in May of 2023, we launched the global airline industry’s very first generative AI virtual assistant out of India.

So it was a global first in the whole airline industry. Today, it has handled about 13 .5 million queries so far from customers. about 40 ,000 queries a day and it operates at a cost which is 100 per query of a contact center and if you look at the customer preferences over a period of last two and a half years we have been operating this facing all the challenges you mentioned so from a customer preference perspective 50 % of the contact volume goes to the contact center they want to talk to human agent the remaining 50 % of the contact volume comes to A .G out of which it handles 97 % of the queries autonomously only 3 % are escalated further to the agent so pretty high success rate and we faced this challenge from day one we started working on it in November 2022 when the whole Azure OpenA services are available we started working on it and the whole approach to responsibility safety has evolved over a period of time so if you dial the safety knob too much then it is an inconvenience to the customer we practically cannot answer any question because the customers are always changing the way they ask certain thing and we have to be very flexible and clearly Generative AI takes us a large step towards that.

At the same time, we don’t want any jailbreak to happen. We don’t want problem injection to happen. We don’t want any inappropriate thing to happen. So we are watching the whole performance of the Gen A virtual assistant, A .G as we call it, all the time. So we use, in fact, Generative AI to watch the performance of the Generative AI chatbot and also we have given the voice to the customer, right? So at the end of the day, when we send a response, we also ask the customer, you know, did it answer your question? And also allow them to give their reactions, right? Is it appropriate, inappropriate? And thankfully, over the last 20 years, it has not answered one single question in an inappropriate fashion because we have embedded all the safety procedures all deep into the way that we handle it.

But now we are, as the technologies are maturing, for example, we now have… interesting technologies in terms of the prompt firewalls where we can centralize all these controls and obviously work with great partners like Adobe who do their diligence and the way that they have deployed some of these technologies, giving full indemnity to us in the event of a problem. That gives a lot of confidence in the way that we manage the risk. So it’s about managing the risk of something that is not within the bounds happening versus the convenience of the customer and we handle it in a variety of ways like I talked about just now.

Excellent. And given the kind of scale you’re operating at, I think every day is a new day. Yes, it is. We face challenges. There is new, brand new every day. Absolutely well stated. Thank you, Dr. Satya. Vishal, may I kind of call you into the discussion? We’re waiting to hear about you. Again, what do I say? NPCI, you know, largest, you know, digital payments, infrastructure platforms. you kind of call the shots in terms of taking some of those calls, you know, for want of a better coinage, you know, in terms of how the payments systems in this country move. So two quick questions here, or rather I’ll sort of, you know, phrase it into one so that, you know, we can get a comprehensive view from you.

How are you really looking at AI in terms of really, you know, being inclusive, you know, in trying to ensure fairness when it comes to two parts? One is when you’re looking at how India can play an important part in creating a responsible AI by design for a digital, national digital infrastructure platform such as yours. And B, how are we really looking at, given the volume, scale, size, fraud also becomes an unfortunate part of the entire, you know, discussion. So how are we really looking at AI to be, fair at the same time proactive and detective when it comes to looking at fraud? So how, what are the aspects that you look at? keenly here

I think we have to start slowly, ensure the accuracy can be a little lower, but the false positive, which is a genuine transaction being tagged as a fraud, should not be very, very high. I think that was the first principles on which we started. But over a period of time, once we had more data, once we collaborated with the industry and ecosystem, we saw that we were able to achieve higher accuracy. So I think this was the fundamental principles when, you know, I mean, and then once we started with the success, we were able to understand the customers better, you know, their patterns better. And that gave us a lot of insights into, you know, fine tuning the models and taking it forward.

Absolutely. So coming to the first question, you know, that you asked, I think, you know, I think there are obviously the governance principles are core to it. But I think I would like to call out two of the things that is there. One is a transparency. If a customer has a transaction that is failed, I think you should know why it has been failed. and today we build a small language model where you can go and actually chat and say what happened to this transaction why is it declined and you know even if it is declined due to a fraudulent transactions uh you know due to a suspicious activity we can actually tell him today saying that you know this is where we feel you know you normally do you know don’t send this transaction or you don’t scan a qr ever this is the first time you’re doing so this is the reason why we have sort of declined so this level of transparency and ensuring there’s a person you know obviously we can’t you know have the army of people sitting and answering these questions but building systems and answering those questions is very very important and i think we have a beautiful framework rba has also given the framework from a responsible ai meaty document is fairly comprehensive so i think all the principles have to be adopted there is absolutely no choice for us um and and i think i i don’t see it as a challenge at all because in our experience it’s been very very helpful to obviously ensure the trust in the payment system is not compromised

absolutely and also the fact that you know as you said given the scale that you’re operating in I’m also itching to ask you some things but maybe I’ll pick your brains offline in terms of the human in the loop there but yeah that’s a discussion for another time so Pratibha curious to know here so responsible AI while you know in letter and spirit remains there do you think it kind of remains or rather getting relegated to the risk of becoming something of an enterprise large enterprise luxury is it able to cut across and come down the line to you know aspiring businesses growing businesses MSMEs is it able to cut through the noise what’s the responsibility of industry leaders and the larger enterprises in kind of harmonizing the framework that can define responsible AI How would you look at this?

Is it getting risked?

Absolutely, yes. I think we stand in that time when on the creators of the AI technology, the big guys versus the small guys, that divide might just become very stark. Coming down to the users of the AI, which is the big enterprises and the MSMEs who are in a big rush to make profit, do something, so that divide can happen. Hence, it is responsibility of all enterprises to be responsible for creating those responsible AI frameworks. It’s a tongue twister, but I think the responsibility is very, very big right now. Again, that Adobe’s example, while the entire AI big bang started happening after November 2022, so early 2023 -2024, but I think our models were there, or our content authentication initiative was from 2019.

So, I think that’s a big thing. I think that large enterprises who create technologies absolutely are responsible. And those frameworks now being taken up by many more is again an absolutely act of responsibility to back to the business. And so the creators of these technologies have to come together and keep on creating this method and methodology for others to adopt. Now the users of these enterprise AI -grade technologies. It’s very hard. I mean, I think 10 years back we had digital transformation. Now we are having AI transformation. So the big companies have to quickly create a new org structure, have to create the legal teams, which, by the way, had to just mull over the digital guidelines of various continents, countries, now have to go through the AI guidelines of countries.

So you have to infuse more people into that. Legal teams. So small organizations cannot do that. So the people, process, technology changes that is required to adopt this, big guys can maneuver, shift people, oh, we will take out people from here, put there. The MSMEs don’t have that luxury. So I guess it is creators have to create frameworks so the right technology is created. The users, the big guys, have to quickly tell the methodology. And then the other stakeholders, like the service providers, also quickly have to go from, like I come from this industry where we had custom software to ERPs to digital transformation and now AI transformation. So how do you do this transformation?

Because a technology on its own has no meaning unless there is a context behind it. I think all of that has to come together. And over and above this, since AI, as we have been hearing words like it’s a civilization change, it is similar to electricity and it will change everything. So because of the impact… at society level to each one of us, the governments have a big role here as well. So all of it has to come together to ensure that enterprises and society move in tandem. Absolutely.

Very rightly stated about the fact that there is a larger collective responsibility from the bigger players in trying to define the standards. I think it’s very critical. Amol, if I may ask you, like Pratibha said that there is an accountability. MSMEs are growing and there is also policy that supports the growth at this point of time, right? So in their hurry to really scale and innovate, they are often forgetting what guardrails and consequences they have to face when it comes to their AI policies, strategies, implementations. So what’s the role of the ecosystem? The industry, industry bodies and the entire ecosystem at large in helping responsible AI moving in letter and spirit. So

Shantiri, I think the first step towards being responsible towards anything is awareness, right? So first, as any part of the ecosystem, we need to be aware that this is our responsibility and we are accountable for that. That’s the first thing. Second is comes the action part of it, awareness, action, and then you demonstrate it through your product services or whatever you are trying to create and generate that kind of impact. How does that allocate? I echo the sentiment which Pravita has mentioned here is in terms of big players have to come up with those frameworks. Those frameworks need to get translated and industry partnership is a very key thing here through the industry bodies, right?

Where the learnings have to be disseminated into this. Second is it’s more of a demand and supply kind of a thing. So if the supply is there with the kind of right guardrails and responsible aspects which are there as a part of framework, then naturally the supplier starts aligning to it. For a business like us where we deal from, infrastructure to healthcare. and IT to agriculture, tires. See, it is a very diverse element and there is a different kind of templates which we need to do so. Organizations like us have the responsibilities of creating a framework which will be fair to us as well as to our customers and partners and those learnings can be shared across the value chain where MSMEs may not have access to that kind of an information and that to domain specific.

Mind you, this would change. It’s not like this one guardrail construct will work for everybody, but it would vary from industry to industry, function to function, and that kind of a cascading through the industry bodies like FICCI and others is I think very, very critical. Absolutely. Thank

you for that Amol. Dr. Satya, as we know now that there are enough global regulations or rules or recommendations that have come in. You have the EU AI Act, you have UNESCO’s recommendations, you have the OECD rules, I mean the principles, so on and so forth, and India is also kind of inching towards developing its own strategies, policies, and approaches at this point of time. so the real leadership question at this point of time that remains is how are we looking at marrying global best practices with the diversity the scale, the fire in the belly that India has at this point of time, we are really gearing to go but how are we really looking at it and besides of course we have a lot of domestic regulations industry wise as well, we have regulators we have even the DPDP Act, we have so many things that have come in how are we going to marry all of this and create harmony absolutely, I

think taking Air India, we are an international airline so we operate in many countries, for example we go to North America, US, where the federal aviation regulation is the key regulator, then we go to Europe all places in Europe, then obviously we operate in India where the DGCA is the regulator and they are doing a great job overseeing this industry and likewise in other parts of the world, so by nature we are geared to looking at the regulation in all parts of the world and I think and being in compliance, and our customers are international, and when we operate in this international geographies, we have to comply with the appropriate regulations. And again, you know, aviation is a very safety -critical industry, right?

So what we do has a direct impact on the safety of the customers that we carry. And the notions are well embedded in the industry because of this, because it’s highly regulated. For example, even, you know, many of these planes can practically land themselves, right? I was in a simulator last week for an Airbus A320 and landing that plane in San Francisco Airport. And as we were coming in, the plane was set at seven miles from touchdown, and, you know, my trainer pilot gave me the control. And, you know, so I could run the plane practically on autopilot all the way. At the same time, you know, there is a red button in the joystick, so at any moment I feel that the airplane is not doing the right thing.

the autopilot control is not in the right thing and quickly cancel and take our control, right? So that is – this concept is well embedded in the airline industry. So we know that we need to obey the regulations and do the right thing that is safe for the customer, but we also help the human in the loop take control if at any moment we feel the safety is at risk. So bottom line, you know, we comply with all the regulations, and it doesn’t in any way constrain Indian innovation. For example, again, like I mentioned, we launched the global airline industry’s first -generation virtual agent out of India, and we have not had any challenge with any of the regulations around because we comply with all the regulations and we work with partners who approach it in the same spirit like Adobe.

Absolutely.

So, Vishal, taking a thread from what Dr. Satya said, I’ll close with you here. Is industry -led governance realistically possible, or is regulatory intervention an inevitability? I did speak to people from other forums in some other related industries where they said, you know, it’s very difficult to answer this. But, yeah, self -regulation may be a way forward given the scale that we are operating on. I’d like to know your thoughts

Yeah, I think definitely the regulations are required, especially because AI can go berserk. And, you know, there are, I think, like I gave you the example on the transactions. You know, today all the UPI transactions can get declined, you know. And that’s where we have a check where we say, you know, this is the only percentage that I can decline, even if I have to let go of the other transactions, right. So those safeguards are very much, very much required. And when this has to be across the ecosystem, I think, you know, the regulations are mandatory. And obviously it has to be consulted and we have to work with everyone. But it’s important. While all of us realize.

it’s a great opportunity the innovation can really scale up I think regulation is one thing that I think we have to really take it as part of the initiatives, embed into our systems and then take it forward otherwise the chances of this having a challenge to us is really high

Fair enough, I think in an economy that is maturing regulatory intervention is an inevitability that we must welcome at some level so great discussion, I think this was fantastic, itching to ask you more but I think we’ll have to call to a close this discussion, thank you so much let’s put our hands together to our esteemed panelists this was really nice may I now invite Ms. Sarika Guliani who is the Senior Director, Head AI Technology in Industry 4 .0, Communications, Mobile Manufacturing and Language Technologies at FICCI Thank you so much Sarika, please

first of all what an insightful session I would say that if I start capturing the thoughts I think 2 minutes and 36 seconds would not justify that one but overall if I talk about it starting with Andy you mentioning about the initiatives what Eduvi was able to take in through and how you are responsibly developing the content Pratibha talking about the art which is really interesting whether you talk about accountability responsibility and transparency and of course Amol mentioning about all the five layers and how the of course the responsible development of AI and the permutation needs to be done Dr. Satya no second thoughts to it and that again goes to the NPCI also the kind of work which the national carrier of India is doing it or NPCI is handling it it has to be a balance of the responsible AI and the efficiency and actually the act which can be taken and the agency itself and the so we left it on the note that what regulation is required while that’s a sentence would require another session on this because they would have a people who will talk about a light touch regulation versus a balanced regulation to something I’ll say that as part of FICCI and as part of the discussions what we were hearing it here we feel it that responsibility is not anymore a compliance check which is supposed to be there it’s a commitment of the technology that we should develop it which has a shared human values that is something what decisions we take now not in terms of the words what we are discussing here is not going to define our future it’s not something which is what we create we define what we choose to create is something what will get defined that is something is very important so the choice comes out from the whole thing you have heard the panelist talking about whether it was from the input side to the output side give a very good example by taking it through the whole process it needs to be taken so we simply feel it whether it’s any layer it has to be developed in a way keeping people in mind and the theme of the summit people planet and progress should be kept in mind while doing any technological innovation which is keeping the principles of responsible AI into the mind.

That is something which we highly feel it and support it and I think with that I like to thank our esteemed panelists. Thank you Andy of course for joining us today. Shantri for moderating it and well capturing in time. And of course our panelists Dr. Satya Ramaswamy, Chief Digital and Technology Officer Air India. Mr. Amol Deshpande, Chief Digital Officer Head of Innovation RPG Group. Mr. Vishal Kanwati, CTO, NPCI Ms. Pratibha Mohapatra, Managing Director Adobe and the of course my lovely audience through which we could sail through this and as part of FICCI we are thankful that we could have this joint session with Adobe, the team of Adobe and Nita and Nanya who had worked with my team and the people at the background who delivered it.

So thank you all for joining us. We don’t end the discussion here. We end the session here. FICCI is committed to get this dialogue further into action with the support of the players. And we look forward to your joining in that time. Thank you. Thank you. Thank you.

Welcome to this session titled, Responsible AI from Principles to Practice in Corporate India, presented by Adobe in association with FICCI. India stands at this defining moment in its digital journey as AI becomes a powerful engine to innovation and productivity. But the real differentiator, is it about how quickly do we adopt AI? No. It’s about how responsibly we deploy it. Trust, transparency and accountability are no longer optional. They are foundational. And that’s exactly what we are going to be talking here today. The conversation will center on advancing safe and trusted AI in the corporate landscape. To set the context and to get us started, it’s my privilege to invite a guest. Andy Parsons, our Global Head for Content Authenticity at Adobe.

Andy, over to you.

Thank you so much. Whenever I speak publicly, they have to lower the microphone. They never raise it. I don’t know. Thank you. Thank you so much for having me. I know we’re against a tight time frame here, so I’m going to speak for just about five or six minutes and then turn it over to our wonderful panel where all the action will be. I’m Andy Parsons. I run the Content Authenticity Initiative at Adobe, and I’ll tell you a little bit about what we do. I think it’s a good example of how responsible AI can be adopted, promoted, and effective in enterprise. I want to start with a simple observation, and I promise not to talk too much about policy because I’m unqualified to do that.

I’m a mere engineer at Adobe. But 2026 is certainly going to be the year that responsible AI becomes a responsibility and an opportunity for innovation. I’m going to start with a simple observation. I’m going to start with a simple observation. I’m going to start with a simple observation. I’m going to start with a simple observation. of you in this room. This means it stops being a slide in a deck, and it will now be sort of a piece of our compliance strategy, but also, as I said, an important opportunity. The EU AI Act’s enforcement provisions take effect in August, as does the first law in the United States in California. And of course, we have the new IT rules here in India on SGI, and India is actively shaping its own path.

And it’s good to be here while that’s happening and talking to many of you about how AI and transparency can be effective in the very short term. So the question for everyone in this room has changed, I would say, this week and certainly will continue to change this year, from should we be responsible with AI? I think that debate is well settled at this point, but can your systems actually prove that you have been responsible with AI, and how do you go about doing that? And for all of us, what does it cost in terms of implementation and day -to -day usage? So we want to position responsibility, AI today as a leadership and operating must and discipline rather than a regulatory obligation, although in 2026, I think it will be both.

So this shift from principles to provable practice is the theme of our panel today. It is what I want to spend just a couple of minutes on. I won’t speak about this from a theoretical perspective. I’ll tell you a little bit about what my team at Adobe does and why I think it matters and perhaps sets an example for others. The trust crisis with AI is real and it’s concrete and it’s happening every day to our children and our businesses and ourselves. Anyone who consumes media, especially across a cultural vastness and disparate languages that we have here in India, is certainly well aware of this. Generative AI has done remarkable things for creativity. Of course, we live and breathe that at Adobe every day and you’ll hear more about that from my colleague Pratibha in a moment.

But it’s made the trust problem absolutely impossible to address. Ignore. So I think every enterprise in this room now produces or consumes AI. generated content at scale, whether it’s marketing assets or news or customer communications, product imagery, et cetera. The volume is extraordinary, and it’s absolutely accelerating every day. The potential is huge if the risks are managed, which is what we’re here to talk about today. In fact, I’d argue that we can all go faster with our adoption of generative AI if we do it responsibly and critically put in place a foundation for that responsibility now rather than wait and be reactive. India has the world’s largest digital population. Of course, you all know that.

Hundreds of millions of people consuming digital content every day. In this environment, synthetic content and AI -generated misinformation are not abstract risks. They’re real ones and operational ones for all of our businesses. So the corporate responsibility question is if you deploy AI that generates or modifies content, as almost all of you, I would imagine, do now or will shortly, can you demonstrate what was made, how it was made, and by which models and products? And that’s what my job at Adobe is. I’m very privileged to provide some leadership. I’m very privileged to provide some leadership in the C2PA, which hopefully many of you have heard of. have heard about this week, which is the Coalition for Content Provenance and Authenticity.

And our sort of piece of the responsible AI landscape is around transparency for content that’s generated by our tools, but also providing a global standard so anyone can do this without licensing totally free. So perhaps this is a case study, and I’ll just spend my last couple of minutes on this. At Adobe, we decided five years ago now, around the time that I joined the team, that responsible AI via content transparency wasn’t a feature that could be grafted onto our products like Photoshop and Premiere, digital experience products, but had to be baked into the tools kind of at their very core. And we went about doing that. While we considered how to do it, we contemplated developing a global standard with partners like Microsoft, BBC, OpenAI, Sony, and many others.

And now we’ve done that. Five years later, there is an open standard called the C2PA content credentials. If you browse LinkedIn and see this symbol, you have the C2PA content credentials. I have encountered a content credential, and it will provide transparent context about a piece of media, whether it’s a video, audio. or image. And as we see adoption increase, we realize this is perhaps a model for AI adoption in a responsible manner based on open standards and interoperability. So we are built on an open standard. It’s truly a cross -industry coalition that includes all the companies I mentioned, meta camera manufacturers like Sony, Nikon, a growing number of media organizations, perhaps some in this room, and also silicon manufacturers like Qualcomm and others.

And the goal is for an infrastructure layer for content trust that anyone can adopt. It shouldn’t be owned by any one company. It should be standards -based. It should not be proprietary, but available to everyone. And this shared philosophy, I think, is especially important here in India. And it should be conveyed by working code and products that leverage that working code, not theory and slides in a slide deck and statements on a website. So what are the principles that the Content Authenticity Initiative and the CTPA reflect? Transparency, provenance information travels with assets when you make them. You can build an entire genealogy tree to understand where your corporate or daily content comes from, what models made that content, what products were used, what cameras were used.

Simple ideas like knowing that a photograph is actually a photograph and not generated. These are simple ideas. I’d say we’re well overdue having this ability, but we need it now more than ever. Accountability. You can trace those AI models, understand what was used, how it was used, and thereby understand the prominence if you wish to access it. We sometimes think of this as nutrition labels, which you heard PM Modi mention yesterday in his remarks. If you walk into a store in most democratic nations in the world, you can pick up a piece of food. You can decide if it is healthy or not healthy for your children. No one’s going to stop you from buying it or consuming it, but you have a right to know what’s in it, and we think that digital content has to have that same foundation of transparency.

Last, inclusivity. Perhaps this matters enormously for this room. Our standard is open and free. An independent creator here in India. We can apply the same kind of provenance at the same zero cost. as a Fortune 500 enterprise. I won’t say that everything is roses and happiness. There are challenges, and I’d be remiss if I didn’t spend 30 seconds on the challenges that standards adoption and AI transparency and responsible AI will present you, and I’m sure you’ll hear more about that with our esteemed panel. But adoption is uneven. Many social media platforms strip metadata and remove that transparency when content is uploaded. Legislation may help here. We’ll see. Consumer awareness is still very early. I saw many of you squint and look at this pin because you probably haven’t seen it.

Our hope is that you will see it ubiquitously across the world starting in 2026. And because consumer awareness is early, user interfaces are also quite early. But there are telephones, there are phones, cameras that are now beginning to show this symbol as a symbol of transparency. And the business case for provenance has been challenging. I’ve often said that doing something that helps perhaps preserve democracy and democratic discourse is maybe not a good way to make money. I’m sure that you will see it. I’m sure that you will see it. I’m sure that you will see it. But it is critically important. And now we’re seeing that change as enterprises have to be compliant and have to be leaders when it comes to AI transparency and responsibility.

What’s changing, of course, is regulation. I view regulation, like what’s happening here in India, as a catalyst for good practices. We don’t want to be reactive, but we do want to help catalyze the responsible AI ecosystem. You need standards, not just principles. Responsible AI commitment on a website is a starting point, but not a meaningful milestone. And that’s really the difference that we’re here to talk about today, the difference between principle and practice. I think you need cross -industry infrastructure. Techniques you use for responsible AI should be interoperable, open, and standardized. And I think you have a long track record here in India of mobilizing things like UPI payment infrastructure that not a single bank could do or a single government agency, but required massive scale cooperation for openness, standards, and most importantly, interoperability.

And last, enterprises. Like many of yours in this room, I’m sure you’ve all heard the phrase, that are willing and excited to go first, that really look at transparency and AI responsibility as an opportunity, not a requirement and a consequence of AI. So let me bring this all together before I introduce our panel. The responsible AI conversation has matured, and now we have to move it to pragmatic implementation. It should favor fairness, accountability, transparency, privacy, inclusivity. We’ve all heard these words over and over again, and in 2026, I think it’s absolutely critical to put them to work and demonstrate that all of our enterprises are doing those things, and that’s the hard part. That’s going to look different depending on whether you’re operating aviation systems where safety is critical, running payments infrastructure at massive scale, governing AI across a conglomerate, or building creative enterprise tools as we do at Adobe, which is exactly why I carefully selected those industries because we have representatives from all of them on our excellent panel today.

So we’re fortunate to have leaders from. Air India, PCI, RPG Group, and Adobe. each of whom is navigating and translating the sort of remarks that I’ve made in different ways for different outcomes and providing, I would say, exemplary ways to find their pathway through the challenges I mentioned. And we have a fantastic guide for this conversation, Shantari Malaya, editor at the Economic Times, who covers these infrastructure and societal sort of breaking points every day in her coverage of our various industries. So I’m going to stop there. Thank you so much for having me. And let’s get on to our panel. Shantari.

Thank you so much, Andy. That was fantastic. It set the context very rightly for the next discussion coming up. So a warm welcome once again from my end. My name is Shantari Malaya. I’m editor at Economic Times. Welcoming you all to the panel. Right at the heart of the AI Impact Summit. It’s been spectacular. And to take this discussion forward, we are looking at responsible AI. very momentous time in India. India is really charting the course for the world. And as we look at building trustworthy and inclusive AI, industry, infrastructure, policy perspectives, it really becomes important to know what some premium leaders in the country are thinking about this. So this dialogue will examine some parts of responsible AI and how it’s really going to shape, reshape enterprise strategy.

So to help me with the discussion, may I have the pleasure of inviting Dr. Satya Ramaswamy, Chief Digital and Technology Officer at Air India Limited. Warm welcome, Satya. We have Mr. Vishal Anand Kanwati, Chief Technology Officer, National Payments Corporation of India, NPCI. We also have Mr. Amol Deshpande, Group Chief Digital Officer and Head of Innovation at the RPG Group. And I have the pleasure of inviting Prativa Mohapatra, Vice President and Managing Director of Adobe at India. This is a fantastic lineup and we shall get some very sharp insights from our panelists over the next half hour or so. So at the very outset, if you really see building trustworthy and inclusive AI is all going to be about how responsible AI principles, whether it’s fairness, accountability, transparency, privacy and inclusivity, is really going to actually realistically be translated into enterprise strategy frameworks and how we are going to go about it.

Right. So Amol, let me call you into the discussion. Warm welcome. I’m keeping a tight watch on the timer right behind us. As we know, this is a summit of scale and we really need to help the organizers to clock good time. So Amol, very quickly, as I invite you into this discussion, you represent enterprises at, you know, while you are participating. As part of the RPG group, you also represent enterprises that are deploying AI at scale. and many more that are just finding their footing as well. So there’s a huge spectrum and diversity in the kind of organizations that you are representing. Two things for you very quickly. One is in large multiple, you know, businesses such as the RPG group, how are you really preventing responsible AI from really becoming something like a just mere centralized compliance exercise or something that’s on the other flip side becoming a fragmented business unit wise, you know, checklist.

So there are two risks that can happen, right? In a group, in a conglomerate, large scale or decentralized. So how are you really looking at the balance here? And how do you really see your role in an industry body as well? So all yours.

Thank you, Shantiri. I’m very happy to be here. Thank you for having me with the, you know, esteemed panelists here. You ask a very pertinent question. You know, it’s to be or not to be, kind of a scenario when it comes to AI, but that not to be is not really a choice. and it did mention about the responsible AI and I would take a little stab at peeling and looking at where the responsible AI comes from when it comes to industries. It comes across all those five layers of the AI when we are looking at it. When any enterprise is looking at deploying AI and being responsible for the usage of those elements, whatever you are using for, the responsibility needs to be there at every layer.

It’s not one or the other. It has to be an orchestration of all the things. So far, AI in its very nascent forms had been a thing of center of excellences, trying use cases and seeing what it is happening, but now it has come to a scale. And when it comes to enterprises and manufacturing enterprises, which we have significantly higher share in terms of as a consumer of AI technologies, there is a very clear -cut view on how it is to be done. So you need to provide the playground for the enterprise. It has to operate function with agility. The other part is about people. The people is a very, very important stakeholder in the whole thing.

We are moving from generative AI to AI ML, more complex scenarios and agentic AI. So people is a very, very important aspect of it. The choice still remains with human sense. The awareness is important and enterprises like us spend a significant amount of time and effort in building that skill sets amongst the value chain of all the people who are doing it. Last but not least is the process and governance part which comes with it. It’s more of a guiding principles which need to be given so that it gives an opportunity. It’s more about, you know, if one can say it will be more of a bring your own AI kind of a scenario in every function.

You cannot provide one solution. One size doesn’t fit all. So when you are dealing with such scenarios, a scalable, safe environment with protected with guardrails is a key thing for us. Orchestration and getting a scale. That’s something which is there. But those templates are being exercised, practiced within the enterprises, practice it in a very diverse group like RPC at ourselves. And then that can be deployed at multiple levels.

Absolutely. So as you said, one size doesn’t fit all. Right. And I liked your coinage of bring your own AI. So let me quickly bring in Pratibha here. Welcome, Pratibha. So Adobe is consistently really positioned responsible as a product philosophy and a trust commitment. You may just have to switch that on. So Adobe has constantly spoken about responsible AI as a very fairly large commitment. So how do you really look at all these principles manifesting or panning out in terms of operationalizing it among all your product teams? And sending it out. as a strong positioning internally and at the same time as someone who’s led a lot of industry conversations, where do you really see enterprises struggling to get these things right?

So all yours here.

Okay, thank you. So I think Andy set the context and since we are here not to learn about the principles but the practices, I think everybody should go back with certain practices. So the first practice of AI governance which we practice is art, which is accountability, responsibility and transparency. So if every person goes back to their organizations and talks about art, which is our philosophy, that’s practicing philosophy number one. And we actually have been doing this for our own products for a while now. And of course we are in the business of content for a very, very long time and now the same content is becoming available. It’s becoming the currency which everybody’s debating. So our principles have been there for a while, but how it is actualized.

And let me say how it’s translated into our products. And by the way, it’s in our products. It’s in our methodologies. Every new product that we have goes through a very strong, secure methodology with hundreds of steps inside it. So there’s principles embedded into how we create stuff. But a couple of examples. Firefly, which is our Gen AI tool, actually embeds what Andy said, those content traditions. Nutritions. It is anything that you have being generated out of this product will have that nutrition level. So when an enterprise is using something in Firefly, you can be super confident that you will not be violating any law. You will not be getting into any liability issues. Because how you do it is by feeding.

Because AI is all about the input and output. So the input has to be something which will not land you in the trouble. You cannot take somebody else’s data. So here it is everything licensed. So it goes into the models, and what you create, the output which comes, then you have to test that output. With that output, will we be accountable? Will we be responsible in showing the transparency of how this was created? So I think that loop has to be created in using any AI. Firefly is an example. Let me talk about Acrobat, which everybody has. I’m sure 100 % of you have PDF files on your phones or on your machines. So Acrobat has this new feature called Acrobat Assistant.

It is agentic, but we have so many chatbots in the market. But when you come to an assistant like Acrobat Assistant, it is following the same principles that PDF was used to be created. So everybody is confident when you’re using PDF. So today, you would have read in the papers recently, the Supreme Court was very worried. that there were certain lawyers who had the petitions which had reference to cases which do not exist or it had certain laws stated which are fictitious. So imagine somebody’s created certain content using some sources which were not authentic. Now, if you use acrobat kind of products for that, you feed the data or you feed files from your own machine.

So you’re confident that what comes out of it, you can go back. So wherever there is this usage of high -stake output, enterprise -grade, you have to look at this input -output process and follow the philosophies within it. And I think for every enterprise who is doing that today, they really have to… Already Amal talked about the people process technology. I’m sure every organization today has a legal team, has a compliance team. But these teams have to re -opt to talk about AI compliance. Enterprises… they do business strategy, they have ethical strategies and then they have regulatory compliance. All the three anything that you do in AI ensure that you tick all the three. If you miss any one you might not be ready for the future.

So that’s how I see it.

Absolutely. So I guess the thread of most of what AI now entails is about are we not moving fast enough or are we moving so fast that we’re not really able to own the operational consequences of what we’re trying to do as well, setting out to do as well. So great point on that Prithiva. I’ll circle back to you time permitting. Let’s see how best we can get back. Dr. Satya calling you in here. So aviation volumes, landscape, scale, I mean you name it, it’s all there. So how are we really looking at balancing AI driven innovation really? Where you’re looking at regulation, you’re looking at accountability, you’re looking at operational efficiency. At the same time, you cannot really compromise on user and customer experience.

How do these things really fall in place in terms of vision and metrics?

Thank you, Shantiri. Since the audience is international, real quick introduction about Air India. Air India is India’s national flag carrier. We operate about 300 aircraft, and we carry about more than 100 ,000 customers a day. And we have a few hundred airplanes on order. So once they are delivered, we will be one of the biggest airlines in the world on the size of one of the large three American carriers. So we are building it up to an airline of scale, and it brings about very interesting challenges that we talked about. So let me illustrate the way we handle it with one of our own examples in generative AI. So in May of 2023, we launched the global airline industry’s very first generative AI virtual assistant out of India.

So it was a global first in the whole airline industry. Today it handles. It has handled about 13 .5 million queries so far from customers. about 40 ,000 queries a day and it operates at a cost which is 100 per query of a contact center and if you look at the customer preferences over a period of last two and a half years we have been operating this facing all the challenges you mentioned so from a customer preference perspective 50 % of the contact volume goes to the contact center they want to talk to a human agent the remaining 50 % of the contact volume comes to A .G out of which it handles 97 % of the queries autonomously only 3 % are escalated further to the agent so pretty high success rate and we faced this challenge from day one we started working on it in November 2022 when we were the whole Azure OpenA services are available we started working on it and the whole approach to responsibility safety has evolved over a period of time so if you dial the safety knob too much then it is an inconvenience to the customer we practically cannot answer any question because the customers are always changing the way they ask certain thing and we have to be very flexible and clearly Generative AI takes us a large step towards that.

At the same time we don’t want any jailbreak to happen we don’t want problem injection to happen we don’t want any inappropriate thing to happen so we are watching the whole performance of the Virtual Assistant A .G as we call it all the time. So we use in fact Generative AI to watch the performance of the Generative AI chatbot and also we have given the voice to the customer so at the end of the day when we send a response we also ask the customer did it answer your question and also allow them to give their reactions is it appropriate, inappropriate and thankfully over the last 20 years it has not answered one single question in an inappropriate fashion because we have embedded all the safety procedures all deep into the way that we handle it but now we are as the technologies are maturing for example we now have interesting technologies in terms of the prompt firewalls where we can centralize all these controls and obviously work with great partners like Adobe who do their diligence and the way that they have deployed some of these technologies, giving full indemnity to us in the event of a problem.

That gives a lot of confidence in the way that we manage the risk. So it’s about managing the risk of something that is not within the bounds happening versus the convenience of the customer and we handle it in a variety of ways like I talked about just now.

Excellent. And given the kind of scale you’re operating at, I think every day is a new day.

Yes, it is. We face challenges. There is new, brand new every day.

Absolutely well stated. Thank you, Dr. Satya. Vishal, may I kind of call you into the discussion? We’re waiting to hear about you. Again, what do I say? NPCI, you know, largest, you know, digital payments, infrastructure platforms. you kind of call the shots in terms of taking some of those calls, you know, for want of a better coinage, you know, in terms of how the payments systems in this country move. So two quick questions here or rather sort of, you know, phrase it into one so that, you know, we can get a comprehensive view from you. How are you really looking at AI in terms of really, you know, being inclusive, you know, in trying to ensure fairness when it comes to two parts?

One is when you’re looking at how India can play an important part in creating a responsible AI by design for a digital, national digital infrastructure platform such as yours. And B, how are we really looking at, given the volume, scale, size, fraud also becomes an unfortunate part of the entire, you know, discussion. So how are we really looking at AI to be? Fair at the same time proactive and detective when it comes to looking at fraud. So how what are the aspects that you look at? keenly here

I think we have to start slowly, ensure the accuracy can be a little lower, but the false positive, which is a genuine transaction being tagged as a fraud, should not be very, very high. I think that was the first principles on which we started. But over a period of time, once we had more data, once we collaborated with the industry and ecosystem, we saw that we were able to achieve higher accuracy. So I think this was the fundamental principles when, you know, I mean, and then once we started with this success, we were able to understand the customers better, you know, their patterns better. And that gave us a lot of insights into, you know, fine tuning the models and taking it forward.

Absolutely. So coming to the first question, you know, that you asked, I think, you know, I think there are obviously the governance principles are core to it. But I think I would like to call out two of the things that is there. One is a transparency. If a customer has a transaction that is failed, I think you should know why it has been failed. and today we build a small language model where you can go and actually chat and say what happened to this transaction why is it declined and you know even if it is declined due to a fraudulent transactions uh you know due to a suspicious activity we can actually tell him today saying that you know this is where we feel you know you normally do you know don’t send this transaction or you don’t scan a qr ever this is the first time you’re doing so this is the reason why we have sort of declined so this level of transparency and ensuring there’s a person you know obviously we can’t you know have the army of people sitting and answering these questions but building systems and answering those questions is very very important and i think we have a beautiful framework rba has also given the framework from a responsible ai meaty document is fairly comprehensive so i think all the principles have to be adopted there is absolutely no choice for us um and and i think i i don’t see it as a challenge at all because in our experience it’s been very very helpful to obviously ensure the trust in the payment system is not compromised

absolutely and also the fact that you know as you said given the scale that you’re operating in I’m also itching to ask you something but maybe I’ll pick your brains offline in terms of the human in the loop there but yeah that’s a discussion for another time so Pratibha curious to know here so responsible AI while you know in letter and spirit remains there do you think it kind of remains or rather getting relegated to the risk of becoming something of an enterprise large enterprise luxury is it able to cut across and come down the line to you know aspiring businesses growing businesses MSMEs is it able to cut through the noise what’s the responsibility of industry leaders and the larger enterprises in kind of harmonizing the framework that can define responsible AI How would you look at this?

Is it getting risked?

Absolutely, yes. I think we stand in that time when on the creators of the AI technology, the big guys versus the small guys, that divide might just become very stark. Coming down to the users of the AI, which is the big enterprises and the MSMEs who are in a big rush to make profit, do something, so that divide can happen. Hence, it is responsibility of all enterprises to be responsible for creating those responsible AI frameworks. It’s a tongue twister, but I think the responsibility is very, very big right now. Again, that Adobe’s example, while the entire AI big bang started happening after November 2022, so early 2023 -2024, but I think our models were there, or our content authentication initiative was from 2019.

So… I think that large enterprises who create technologies absolutely are responsible. And those frameworks now being taken up by many more is again an absolutely act of responsibility to back to the business. And so the creators of these technologies have to come together and keep on creating this method and methodology for others to adopt. Now the users of these enterprise AI -grade technologies. It’s very hard. I mean, I think 10 years back we had digital transformation. Now we are having AI transformation. So the big companies have to quickly create a new org structure, have to create the legal teams, which, by the way, had to just mull over the digital guidelines of various continents, countries, now have to go through the AI guidelines of countries.

So you have to infuse more people into that. Legal teams. So small organizations cannot do that. So the people, process, technology changes that is required to adopt this, big guys can maneuver, shift people, oh, we will take out people from here, put there. The MSMEs don’t have that luxury. So I guess it is creators have to create frameworks so the right technology is created. The users, the big guys, have to quickly tell the methodology. And then the other stakeholders, like the service providers, also quickly have to go from, like I come from this industry where we had custom software to ERPs to digital transformation and now AI transformation. So how do you do this transformation?

Because a technology on its own has no meaning unless there is a context behind it. I think all of that has to come together. And over and above this, since AI, as we have been hearing words like it’s a civilization change, it is a… similar to electricity and steam, it will change everything. So because of the impact… at society level to each one of us, the governments have a big role here as well. So all of it has to come together to ensure that enterprises and society move in tandem.

Absolutely. Very rightly stated about the fact that there is a larger collective responsibility from the bigger players in trying to define the standards. I think it’s very critical. Amol, if I may ask you, like Pratibha said that there is an accountability. MSMEs are growing and there is also policy that supports the growth at this point of time, right? So in their hurry to really scale and innovate, they are often forgetting what guardrails and consequences they have to face, you know, when it comes to their AI policies, strategies, implementations. So what’s the role of the ecosystem, the industry, industry bodies and the entire ecosystem at large in helping responsible AI moving in letter and spirit?

Shantari, I think the first step towards being responsible towards anything is awareness, right? So first, as any part of the ecosystem, we need to be aware that this is our responsibility and we are accountable for that. That’s the first thing. Second is comes the action part of it, awareness, action, and then you demonstrate it through your product services or whatever you are trying to create and generate that kind of impact. How does that allocate? I echo the sentiment which Pravita has mentioned here is in terms of big players have to come up with those frameworks. Those frameworks need to get translated and industry partnership is a very key thing here through the industry bodies, right?

Where the learnings have to be disseminated into this. Second is it’s more of a demand and supply kind of a thing. So if the supply is there with the kind of right guardrails and responsible aspects which are there as a part of framework, then naturally the supplier starts aligning to it. For a business like us where we deal from, infrastructure to healthcare. and IT to agriculture, tires. See, it is a very diverse element and there is a different kind of templates which we need to do so. Organizations like us have the responsibilities of creating a framework which will be fair to us as well as to our customers and partners and those learnings can be shared across the value chain where MSMEs may not have access to that kind of an information and that to domain specific.

Mind you, this would change. It’s not like this one guardrails construct will work for everybody, but it would vary from industry to industry, function to function, and that kind of a cascading through the industry bodies like FICCI and others is, I think, very, very critical.

Than you for that, Amol. Dr. Satya, if, you know, as we know now that there are enough global regulations or rules or recommendations that have come in. You have the EU AI Act, you have UNESCO’s recommendations, you have the OECD rules, I mean, the principles, so on and so forth. And India is also kind of inching towards developing its own strategies, policies, and approaches at this point of time. so the real leadership question at this point of time that remains is how are we looking at marrying global best practices with the diversity the scale, the fire in the belly that India has at this point of time we are really gearing to go but how are we really looking at it and besides of course we have a lot of domestic regulations industry wise as well, we have regulators we have even the DPDP Act, we have so many things that have come in how are we going to marry all of this and create harmony

Absolutely, I think taking Air India, we are an international airline so we operate in many countries for example we go to North America US, where the federal aviation regulation is the key regulator, then we go to Europe all places in Europe, then obviously we operate in India where the DGCA is the regulator and they are doing a great job overseeing this industry and likewise in other parts of the world so by nature we are geared to looking at the regulation in all parts of the world in all parts of the world and we are looking at the regulation and we are looking at the regulation and we are looking at the regulation and being in compliance.

And our customers are international, and when we operate in this international geographies, we have to comply with the appropriate regulations. And again, you know, aviation is a very safety -critical industry, right? So what we do has a direct impact on the safety of the customers that we carry. And the notions are well embedded in the industry because of this, because it’s highly regulated. For example, even, you know, many of these planes can practically land themselves, right? I was in a simulator last week for an Airbus A320 and landing that plane in San Francisco airport. And as we were coming in, the plane was set at seven miles from touchdown, and, you know, my trainer pilot gave me the control.

And, you know, so I could run the plane practically on autopilot all the way. At the same time, you know, there is a red button in the joystick, so at any moment I feel that the airplane is not doing the right thing. the autopilot control is not in the right thing and quickly cancel and take our control, right? So that is, this concept is well embedded in the airline industry. So we know that we need to obey the regulations and do the right thing that is safe for the customer, but we also help the human in the loop take control if at any moment we feel the safety is at risk. So bottom line, you know, we comply with all the regulations, and it doesn’t in any way constrain Indian innovation.

For example, again, like I mentioned, we launched the global airline industry’s first -generation virtual agent out of India, and we have not had any challenge with any of the regulations around because we comply with all the regulations and we work with partners who approach it

So, Vishal, taking a thread from what Dr. Satya said, I’ll close with you here. Is industry -led governance realistically possible, or is regulatory intervention an inevitability? I did speak to people from other forums in some other related industries where they said, you know, it’s very difficult to answer this. But, yeah, self -regulation may be a way forward given the scale that we are operating on. I’d like to know your thoughts here.

Yeah, I think definitely the regulations are required, especially because AI can go berserk. And, you know, there are, I think, like I gave you the example on the transactions, you know. You know, today all the UPI transactions can get declined, you know. And that’s where we have a check where we say, you know, this is the only percentage that I can decline, even if I have to let go of the other transactions, right. So those safeguards are very much, very much required. And when this has to be across the ecosystem, I think, you know, the regulations are mandatory. And obviously it has to be consulted and we have to work with everyone. But it’s important.

While. All of us realize. it’s a great opportunity the innovation can really scale up I think regulation is one thing that I think we have to really take it as part of the initiatives, embed into our systems and then take it forward otherwise the chances of this having a challenge to us is really high

Fair enough, I think in an economy that is maturing regulatory intervention is an inevitability that we must welcome at some level so great discussion, I think this was fantastic, itching to ask you more but I think we’ll have to call to a close this discussion, thank you so much let’s put our hands together to our esteemed panelists this was really nice may I now invite Ms. Sarika Guliani who is the Senior Director, Head AI Technology in Industry 4 .0, Communications, Mobile Manufacturing and Language Technologies at FICCI Thank you so much Sarika, please

first of all what an insightful session I would say that if I start capturing the thoughts I think 2 minutes and 36 seconds would not justify that one but overall if I talk about it starting with Andy you mentioning about the initiatives what Eduvi was able to take in through and how you are responsibly developing the content. Pratibha talking about the art which is really interesting whether you talk about accountability responsibility and transparency and of course Amol mentioning about all the five layers and how the of course the responsible development of AI and the permutation needs to be done. Dr. Satya no second thoughts to it and that again goes to the NPCI also the kind of work which the national carrier of India is doing it or NPCI is handling it it has to be a balance of the responsible AI and the efficiency and actually the act which can be taken thank you so we left it on the note that what regulation is required while that’s a sentence would require another session on this because they would have a people who will talk about a light touch regulation versus a balanced regulation to something I’ll say that as part of Vicky and as part of the discussions what we were hearing it here we feel it that responsibility is not anymore a compliance check which is supposed to be there it’s a commitment of the technology that we should develop it which has a shared human values that is something what decisions we take now not in terms of the words what we are discussing here is not going to define our future it’s not something which is what we create we define what we choose to create is something what will get defined that is something is very important so the choice comes out from the whole thing you have heard the panelist talking about whether it was from the input side to the output side give a very good example by taking it through the whole process it needs to be taken so we simply feel it whether it’s any layer it has to be developed in a way keeping people in mind and the theme of the summit people planet and progress should be kept in mind while doing any technological innovation which is keeping the principles of responsible AI into the mind.

That is something which we highly feel it and support it and I think with that I like to thank our esteemed panelists. Thank you Andy of course for joining us today. Chantri for moderating it and well capturing in time. And of course our panelists Dr. Satya Ramaswamy, Chief Digital and Technology Officer Air India. Mr. Amol Deshpande, Chief Digital Officer Head of Innovation RPG Group. Mr. Vishal Kanwati, CTO, NPCI Ms. Pratibha Mohapatra, Managing Director Adobe and the of course my lovely audience through which we could sail through this and as part of FICCI we are thankful that we could have this joint session with Adobe, the team of Adobe and Nita and Nanya who had worked with my team and the people at the background who delivered it.

So thank you all for joining us. We don’t end the discussion here. We end the session here. FICCI is committed to get this dialogue further into action with the support of the players. And we look forward to your joining in that time. Thank you. Thank you. Thank you.

Again, to my immediate right, we have Stephen Clare, who wrote the International AI Safety Report as the co -lead author, if I’m not mistaken. And he earned that applause, because that report is a remarkable document that I do think is the foundation upon which all conversations about AI governance now must rest for the next year. It’s the sort of minimum amount of knowledge that you must have to participate in the conversation, which I think is really a tribute to him. Then we have Hiroki Hibuka, who is currently a research professor at the Kyoto University Graduate School of Law, and was also deeply involved in drafting Japan’s first set of soft law regulations, and is an expert on all things AI, but also especially astute at what’s going on in Japan.

We also have a privilege of collaborating with him at CSIS, where he’s a non -resident senior associate. And I must say, he is probably the best person writing about Japanese AI policy in Japanese, but he is definitely the best person writing about it in English. And so I often tell Hiroki that, like, if he doesn’t write about it, nobody in Washington, D .C. knows about it. So it’s important, his work. And then finally, we have Shana Mansbach, who’s the vice president of strategy and communications at Fathom, which is a young think tank, started only two years ago, but has already succeeded as one of the best conveners of the ASHFE conference series on AI, and also now leading a policy initiative, which I think she’s going to tell us all about.

So without further ado, I’d like to start with you, Stephen. I just said that the report that you were the lead author of is sort of the bedrock for having a conversation on AI governance. For those in the audience who haven’t yet made it through, but they, of course, will, can you sort of set the stage? Where are we in 2026 in AI governance and in AI safety, technical and procedural intervention?

Sure. Thanks, Greg. First of all, I’m sorry if I’d known Greg was going to make the report, you know, required reading, I would have tried harder to make it shorter. Yeah. Thanks for having me. Thanks for really excited to be here. So for people who don’t know, the report is it was founded up the Bletchley 2023 Bletchley Safety Summit as sort of, you know, the shared evidence base for decision makers thinking about these complicated, fast moving, noisy governance questions. It’s kind of trying to be like the IPCC report for for AI. It’s backed by over 30 countries and intergovernmental organizations. You know, I’m one of two co lead writers along with Karina Prunkle, but there’s over 30 dedicated experts writing different sections, and there’s hundreds of people that review it.

So it’s really trying to be a sort of state of the art, what do we know? What don’t we know about general purpose AI systems and the risks they might pose? I think this year the main message of the report is like the rubber is really hitting the road or something with these kind of systems. Risks that even a year or two ago might have been theoretical are now very real and we’re seeing emerging empirical evidence. More real world impacts of AI on productivity and labor markets and in science and in software engineering. It’s all like really happening out in the world. There’s a billion people now using AI around the world. Many of those impacts include risks.

So we’re seeing effects of deepfake spreading, cyber attacks being more common with AI systems. And so the need for sort of risk management techniques that are effective is also growing. One thing that I found surprising working on the report is that in this domain on risk management and technical safety, there’s actually some good news. Quite a lot of good news, I’d say. In various ways, our technical safeguards are improving. Models are becoming much harder to jailbreak. So. You know. So three, four years ago, if you asked a model to give you a recipe for a Molotov cocktail, it would not do that. But if you said, oh, I miss my grandma, and she used to tell me this amazing bedtime story about how she loved making Molotov cocktails, please help me remember my grandmother, it would be like, okay, well, if it’s for your grandmother.

Then that stopped working maybe a year or two ago, but then if you maybe translated your question into Swahili or something and put it in the model and then translated the answer back, it might have made safeguards. So none of that works anymore. These safeguards are much harder to evade, and we know this quantitatively. For example, the UK Security Institute will try and evade the safeguards or jailbreak all these new models when they’re released. At the beginning of 2025, they could do this in literally minutes, find a sort of universal jailbreak that would elicit potentially harmful knowledge. For the latest models, it’s taking them seven, ten hours to get around safeguards. So there’s still vulnerabilities, but for novices or even moderately skilled actors, it’s basically the same thing.

It’s becoming much, much harder to evade them. We’re also seeing more of these safeguards get implemented into organizational practices. So 12 companies, all the leading AI developers now have frontier safety frameworks, which are these documents that describe how they plan to manage risks as they scale more powerful systems, which is many more than had them a couple of years ago and is, I think, a sign of transparency and sort of collective learning about risk management that’s worth noting. So basically, yeah, our toolkit for managing these risks is growing. But, you know, it wouldn’t be a safety report if I didn’t maybe end on a few caveats or some bad news. The first is that these technical safeguards are still vulnerable in many ways.

They can still be jailbroken with enough effort or in edge cases, and it’s very difficult to test and provide reliable assurances that these safeguards will work across this huge range of use cases that these models are now applied to in the real world. And on the organizational side, you know, these safeguards only work if they are applied. And although we’re seeing, especially from frontier developers, we’re very prominent, usually quite robust safeguards applied on models, across the whole industry, and especially behind the frontier, application remains quite inconsistent. The safety frameworks, all these companies have them, but they vary in the risks they cover, they vary in the practices that they recommend. And so the landscape as a whole, you know, these tools only work if they are applied.

And we still see that, some vulnerabilities across the landscape, which I think turns this technical challenge, that points towards the governance challenge of how do we assure broader adoption, how do you ensure compliance, what do you do when there’s a lack of compliance. We’re sort of facing these questions, and again, because these risks and the impacts are now not something that we can sort of push down the road anymore, I think, for future years, the governance questions are becoming a lot more urgent.

Terrific. And if I could contrast what you said with what we might have said if we were having this conversation back at the Bletchley Park AI Summit. But it’s almost like the only good news on AI safety, AI security, and AI governance at Bletchley was, well, at least we’re all here talking about it. And now, three years later, the good news is we’ve done a lot about it. We have techniques that can provide demonstrable increases in safety. We don’t know everything that we need to work, but we know a lot of stuff that does work. And really, a lot of the challenges, I think, as the report says, it’s now in the hands of policymakers to make sure that these safeguards get implemented robustly and diversely.

So with that, I now want to turn to Hiroki, who I hope can give us a state of where we are in the story of AI governance around the world. If the next steps are really in the hands of policymakers, where are we globally?

Thank you, Greg. And again, congratulations. Stephen was the publisher of the great report. And I think, first of all, I feel very glad that now the discussion on AI governance is such advanced compared to three years ago. I’m a lawyer and I’m a former policymaker. I worked for the Japanese government for four years, designing the Japanese AI policies, mainly in terms of regulation and governance. And as a lawyer and policymaker, the question after reading the report is, where is the end? And to what extent stakeholders have to manage the risks? Because in the end, you can’t remove all the risks. AI is black box and the technology advances so fast. And even though there is advance and progress of Godwins, the next day you may find another risk.

So there is no end to the story of how regulators should design the regulations. That is the main question. All countries. Countries are facing and different nations, regions take different approaches. Maybe the most famous regulation is the EU AI Act. And in that context, a lot of people say, hey, EU takes a hard law regulatory approach on AIs while Japan or UK or United States takes a software approach. But I think it’s a completely wrong understanding of the regulatory framework because, as you know, there are already lots of regulations that can be applied to AI systems. Privacy protection laws, copyright laws, or sector -specific laws such as finance, automotive or healthcare. We already have a lot of regulations out there.

So the real question is not whether or not to regulate AIs, but the real question is how to update our existing regulations and whether or not we need additional regulations targeting AI systems. In addition to the existing regulatory framework, so in that sense all countries take the hard law approach and also all countries have soft laws because European Union there are a lot of technical standards to implement the EU AI Act that are now under discussion but anyways all countries have both hard laws and soft laws that is the start of the discussion and then when we compare EU approach and Japan approach the clear difference is whether to regulate AI holistically or not sector -specific and when I compare the Japanese policy and the US policy we are on the same position as to taking a sector -specific regulation the main difference I understand is whether you prioritize the exempt approach or exposed approach the US takes more exposed approach you can do whatever you want to do and the regulation is usually very high level the principle is very high But once you have a problem, if you damage others’ properties or lives, then you go to the court and you fight in the court.

The Japanese society is not like that. In Japan, actually the number of losses is very low. People prefer to set the rules in advance. Japanese companies are very, very good at complying with the given rules. But they are not very good at creating their own governance mechanisms or explaining to stakeholders why you are doing that. And now Japanese stakeholders are starting to realize that it doesn’t work. So we need to have more agile and multi -stakeholder approach. So we are trying to leverage the power of soft laws, negotiating among different stakeholders, and give the standards, guidances. But in the end, again, if you violate the existing hard laws, of course you will be sanctioned. So that’s the main differences in American approach and Japan approaches.

And in the end, all countries are facing difficult questions of how to deal with this cutting -edge technologies that are black box and there are unlimited risk scenarios. And sometimes we don’t know how to evaluate the values such as privacy or transparency or fairness. There has been no clear benchmark standards so far in the society. So how to design those benchmarks and regulation methods are the challenges all countries are facing.

Terrific, Hiroki. And Shaina, I know you have a unique perspective on this because your organization is now proposing sort of additional models of AI governance that are not really reflected in existing law, whether in the United States or Europe or Japan or India. So walk us through what you see as the important work we’re doing now.

Sure. My panelists have set me up very well to say this. So I think as the International AI Safety Report shows, the capabilities around these models are surging. And as the capabilities surge, so too does the uncertainty around the risks, by which I mean, do these systems work safely, securely, and as advertised? That uncertainty creates a trust problem, a trust problem for the public, which doesn’t have a way of figuring out what is actually safe, a trust problem for deployers, by which I mean hospital systems, retail, banks, who want to and indeed need to use these systems, but have no idea what they can actually trust. So there’s a trust problem for the regulators, too.

They don’t know, how do you confer not just trust, but how do you confer earned trust? And I would say there’s a trust problem for the developers also, because if and as trust starts to grow, there’s a trust problem for So if the trust starts to decline, you’re going to see adoption decline as well, so this is something that developers should be focused on too. The current approach is just not the current approach to tech governance is not equipped to handle this trust problem very well. Traditional command and control governance says here are the rules, here are all the things you have to do, here are the procedures, here is what compliance actually here’s what compliance actually looks like.

There are a bunch of problems with this approach in the context of AI, but I’ll focus on two, which is the speed problem. AI moves really, really quickly, and even well -intentioned regulations are going to become outdated very, very quickly, and then there’s the technical capacity problem. Even with the rise of the AI safety institutes, which are doing amazing work, the talents, the expertise for understanding these systems and understanding their risks is largely concentrated in the frontier labs, which of course leads some people to say, well, let’s just go to the frontier labs. They can regulate themselves. I don’t think I have to spend too much time explaining why there are problems with that approach but it’s simple incentives I think all of us know people in the labs who are doing amazing, amazing work they are the people who make sure that I can because of them I sleep better at night but the incentives are just not there there are always going to be trade -offs between investing in safety testing and tooling and investing in development so we’re going to have problems with self -regulation in terms of addressing that trust gap so where does that lead us?

at Fathom, my organization, we’re very focused on coming up with new models that can solve this trust gap so we’re very focused on independent verification specifically the marketplace of independent verification organizations by which I mean a government -authorized and overseen marketplace of independent verifiers which are trying to be charged with creating testing and tooling to determine whether these AI systems are actually safe The difference here is that this is an outcomes -based approach. Instead of, as I said, having procedures, here are the rules, here are all the things you need to do, here are all the boxes you must check to be certified as being good, you have an outcomes -based approach where you have a government saying, here are the things that we care about.

We care about children’s safety. We care about data privacy and protection. We care about controllability and interpretability. And then you have independent verifiers that can actually go out, do the testing, have updated testing constantly to make sure that those outcomes are being met. We think that independent verification solves for a couple of these deficits in the trust context. First, they are independent. The labs are not grading their own homework. Second, democratic accountability. You have governments that are creating outcomes instead of the industry doing it itself. Third, flexibility. Under this system, the IVOs, independent verification organizations, are constantly updating their testing and criteria to make sure that they’re keeping up with the pace of technology and the pace of risks as well.

And I think the fourth thing, which is pretty interesting, is it creates a race to the top here. Right now, the only people working on safety testing and tooling are in the labs. What we’re envisioning is a marketplace that incentivizes ever better testing and tooling here. I could talk about IVOs for days and days, but let me just end on one point. I was talking to Greg about this earlier, and Greg asked, are there analogous systems or industries or sectors that we could talk about? And I said, yeah, sort of. I mean, in America, we have Underwriters Lab. There’s LEED certification. There are some analogies. But the honest answer is there’s not a perfect analogy.

We have had the same regulatory system for the last century. And I think that with the rise of AI, we’re seeing that system is no longer built for purpose. And when we try to use old systems, hard law, soft law, any of these things, we’re really struggling to make it work. So what I’m trying to do, what I’d encourage all of us to do is to say, you know, we do need to think a little bit differently. Because this is what this technology in this time calls for.

Well, that’s great. So there’s a few points I want to pull together there. The first is, you know, as Hiroki pointed out, in the U .S. system, liability law looms extremely large, right? The lawsuits at the end of this story when things go wrong. And when you have, as, for example, ChatGPT does, 800 million weekly average users, something’s going to go wrong every week, right? And the question is… How is that going to intersect with our existing body of regulation? How is that going to intersect with liability law? The second thing is this is going to, because we’re talking about these general purpose technologies, this is going to be adopted in so many different sectors of the economy.

And right now, as Shana pointed out, the number of people who have, you know, Steven’s expertise on what it takes to really make AI systems safe and well -governed and perform reliably as intended across the whole range of potential applications, that’s not a lot of humans on planet Earth who are good at that stuff. And because these AI models are going to be deployed in just about every sector of the economy, we need some level of those capabilities in every sector of the economy. And so the question is, you know, if I am a financier, if I am a finance company, if I am a health care company, you know, how am I going to know and how are my consumers going to know?

that when they use AI -related capabilities, it’s going to work reliably as intended over the full range of acceptable use cases. And so, Stephen, I want to come to you and ask, when it comes to governance, when it comes to oversight and verification, how do you see the balance of responsibilities in terms of what responsibilities need to fall upon the model developers, what responsibilities need to fall upon the users, what responsibilities need to fall on independent third parties, whether that’s the government, whether that’s auditors, whether that’s this marketplace of verification that Shana is talking about. So what do you see as the balance of responsibilities, and how might this go wrong, how might this go right?

In 30 seconds or less.

I mean, I’m sure it’s kind of the boring but true answer. It’s the boring part of it. depends and it’ll vary a lot across use cases and sectors. I think probably it’s not the case that it’s fair or helpful or true to allocate to one actor or another, but instead we need this layered approach of just many different policies and practices at different parts of the stack. Because none of our approaches are foolproof, they all have vulnerabilities, and so we have, instead of safety by design, we have this safety by degree situation where we want defense in depth. So for developers, there will be training techniques that they can implement to make models less likely to elicit dangerous knowledge in the first place.

If there are people building on top of those models and then deploying them, there will be monitoring systems they can put in place and classifiers that identify dangerous queries and stop models from answering them. and then probably for ecosystem monitoring bodies which could be deployers but could also be other institutions in the world there can be tracking how AI content is spreading across borders and around the world and then I think there’s this other aspect of we’re focusing a lot on sort of model or developer safety but as we are moving into this world where many people around the world are having access to powerful, helpful intelligent technologies and we also just need to adapt for that reality and think about resilience at the societal level too of how do we adapt to the beneficial use cases and the various use cases that these models will be used for so thinking about hardening digital systems against increased cyber attacks just sort of admitting the reality of the situation in many ways and adapting to it rather than trying to prevent all harmful uses in the first place I think we need a variety of approaches across all these different actors

Yeah. And just to use an analogy for how broad the group of stakeholders is, if you think about a ride hailing service, a taxi service like Uber, you have the automobile manufacturers who have to make sure that this is a solid car design that was manufactured safely and appropriately to specification. Then you have Uber, where in some countries Uber owns the car, and so they’re responsible for ensuring that it gets maintenance appropriately. And then you have the driver who’s responsible for ensuring that they are actually following the law and driving the car safely. And if you apply that analogy to AI, you have the model developer, then you might have the sort of business use case deployer, which could be a bank, a medical device company.

Who? A financial institution, whoever. And then you finally have the end customer who’s receiving those services and making sure that they’re using them appropriately. And so. If you think about that sort of different body of use cases, as I said before, the capabilities are not symmetric across all of those. But there are sort of obligations. And so, Shana, I want to come back to you and ask this model that you’re proposing, what exactly does it mean for the different stakeholders in the ecosystem? How does their life change if we adopt the system that you’re in favor of?

Yeah, I mean, the overarching answer is we create trust throughout the system, which is the missing piece here. I think there are a couple of pieces that I would pull out. You had mentioned liability earlier, and let me talk about that a little bit. What this system does, it does not assign liability. It doesn’t say, you know, deployers, developer, it’s you, it’s you, it’s you. We’re seeing, at least in America, courts move their way through this. Sister. court cases move their ways through the court system and we’ll see where that is but where that ends up being but what is really missing is a standard of care and this is I think one of the real advantages that this system has so right now at least how it works in our current tort system is that if you’re Waymo kill someone someone can sue and then a judge and a jury has to figure out so again we’re not answering who should be sued but let’s say that the family of someone who got hurt or killed is suing Waymo what happens is that the jury has to decide whether whether the person who was sued did the right thing and if you are not technical that is the hardest thing even if you are technical and maybe even Waymo doesn’t know So what this system would do is confer, if you are verified, it would confer, the verification would confer a rebuttal presumption of having met a heightened standard of care.

So what we’re doing is clarifying and defining up front before an actual harm happens what a deployer or whoever is sued is actually supposed to do instead of having this very, very messy system where someone after the fact has to figure out what went wrong and who’s responsible for that. I can talk about other layers of this back here, but I think the liability piece is really key. I mean, we just see this. I think it’s a reflection of the trust problem here where when you’re a deployer, I mean, God, I think everyone that I talk to, you know, again, hospital systems, retail, banks, anyone who needs to be consumer facing is really worried about this problem.

I mean, when I get sued, what do I do? And maybe there’ll be. a populist backlash and everyone will hate everyone who’s using AI systems. And it’s much better to, ahead of something like that, ahead of that happening, have that standard of care defined up front and have that seal of approval conferred.

And Hiroki, as you think about the different stakeholders in the system and especially the idea of auditors, which now there are a number of organizations being founded, it seems like almost every day, who are proposing to provide external evaluation services that can help companies understand, as Shane has said, this product or this service or this company meets the seal of approval and we vouch for it as an independent entity. What kind of momentum do you see for this independent assessment part of the story across regulatory frameworks?

Independent evaluation. Independent evaluation is essential given that we are all using AI systems for all different situations, starting from language models to healthcare systems to car driving. But it would be not easy to persuade corporate executives to use the independent audit without clear economic incentives. For example, if you get the certification for autonomous driving, then you can sell the car to the big market. Then, of course, you pay for the audit. But if you take this audit for this language model, then you can prove that this language model is relatively safer than the other models. But it doesn’t necessarily make enough incentive for model developers to conduct the audit or evaluation systems, independent evaluation, because there is no clear financial incentives.

Actually, could I? I can ask you to elaborate on that. So where might these financial incentives come from? You mentioned one, which is the regulators force you to do it. That’s one. Maybe insurance is one. another like where where might these incentives come from

I think it should start from the regulated areas such as cars health care systems finance systems or infrastructures because everybody needs a strong requires strong trust on those systems if it doesn’t work well then somebody might have a baby kills that’s a big problem and maybe you could say hey but in the end if you are killed you can be compensated but it’s not the end of the story while if the damage could be compensated by money by the company and stakeholders are okay with that maybe companies like to just run the system go and and compensate to the victims for example if the language model says something discriminated the company can just say hey we’re very sorry we introduce better guardrails and we pay for that if you want compensation

in terms of what is possible, what interventions work, what the risks are. But I want to ask about how we go from that degree of consensus to something that might be more of like a standard around procedural implementation. You know, Shana’s term of art is standard of care, which matters a lot in the American legal system. I’m sure it matters a lot in other legal systems. I’m just ignorant about, you know, how and where. And so I’m curious, you know, what do you see as the gap? If these independent evaluators, these independent auditing organizations are emerging, how do they go from we think we’re good at this to, no, this is the accepted best practice?

You know, we have accepted consensus on the risks and the interventions, but, like, how do you turn that into a procedure? Just to give an example to the folks in the audience, I used to work at a rocket company, and the safety standard in the American aerospace industry is AS9100. And in the history, of our company, there’s kind of like a before AS9100 moment, and then there’s an after AS9100 moment. And everything changed for our company, you know, after we got that third -party audit evaluation. A lot of our customers, you know, just said, we do not sign checks for companies that are not AS9100 certified. So, you know, you are deeply steeped in where we are today on the consensus, but how far are we from converting that into standards and procedures for third -party evaluation?

Yeah. I’ll also say one follow -up to Hiroki’s point, too, about auditing. Not only is there sort of a lack of incentives to conduct audits voluntarily now, but there might even be disincentives where one is it’s costly, and it slows you down, and there’s very intense competitive pressures to release faster. And there’s also potentially… like, information or security risks to sharing. You spent hundreds of millions, maybe billions of dollars developing a model, and then you have to share it with an external party before deployment. Like, serious risks to, or perceived risks, at least, to having that information leak or… So I think, yeah, there’s some serious challenges there. I guess there’s one other potential part of the story, which is sometimes you see companies want to be willfully blind, right?

If they have a report that says my product is not safe, well, now they know they’re going to lose the lawsuit. Whereas if they never commission the report, maybe they’ll win the lawsuit. So, Shana, what do you see as meaningful interventions that can help address this problem, both the cost side that Stephen mentioned and the other parts of the incentive structure?

Yeah, let me make a couple of points. I mean, I think we’re talking about the cost of audits, and I think this… this is a big issue that we think about a lot. This system will not work if everyone, if there’s a flat fee, everyone is paying a ton. I mean, we are really, we think that an unsuccessful, there are many ways that a system looks unsuccessful, and one of those ways is if it is just protecting incumbents. And we’re thinking, we envision the system as something that works for, you could verify a general purpose LLM, you could also have narrow AI, you could have a tiny little tool, a little chatbot that is used in schools.

Those three different products should not be audited, not only at the same cost, but in the same way. I mean, compliance isn’t just the check that you’re writing, it is how much of a pain in the butt is it? How many lawyers do you need? How long will this take? So the great thing about this being a marketplace is that the system is right -sized to risk type, to size of these products. and again instead of having just a one size fits all this is what you have to do to comply because I think that that is a real issue it really quickly I just want to go back to you know the question that you asked Hiroki about incentives I mean you can imagine a system where this is mandatory and maybe in some areas you can imagine that but I think that there are three real real carrots for wanting to get verified we talked a little bit about liability so obviously the liability clarity that this is a big carrot I think the insurance piece the insurance piece is real right now we are seeing the big insurers saying we’re not going to touch this we’re not going to insure any AI products because we have no idea what’s inside of them at least in America the way that life insurance works is if you want insurance you have to have a lot of money and a lot of money and a lot of money and a lot of money you have to jump on a scale and tell someone how healthy you are and what are the things that you do and the insurer decides okay are you worthy of being insured and at what premium I think that’s actually a pretty direct analog for what we’re trying to do here where the books are opened and an insurer can look at whether they don’t have to do the testing themselves, but they can look at whether the system has been verified and say, okay, we will actually insure you or we will insure you at a more affordable premium.

I think the third thing is just straight -up market competitive advantage. If I’m a school superintendent and I am choosing between two learning chatbots to put in my schools, I’m not going to choose the one that has not been verified. I want the one that has been verified, that is safest. Yes, because I’m worried about getting sued, but because I want my kids to be safe. And you can imagine a situation much like Underwriters Lab in the United States where basically all consumer products like light bulbs, toothbrushes, basic things that you buy in a store like Walmart, all have the UL seal of approval, and those are the ones that get sold in stores. They have a huge market advantage.

They pay a little bit, but not very much. And in exchange for doing that, they go to market in a way that, or they compete in a market in a way that the ones that don’t go through verification. do. I’m so sorry, Gary, you asked me an actual question and I just answered everyone else’s question and probably not my own.

It’s okay. You get out of jail free card because you mentioned insurance, which is something I’m deeply interested in right now. I mean, in that space orbital launch vehicle example that I just mentioned, you can’t get insurance for space launches of satellites until you’re AS9100 certified. And that is 10 % of the cost of getting a satellite into space is just the insurance on the rocket. And so basically companies that can’t get insurance can’t compete in the market. And as Shana mentioned, and I think this is a super undercovered story, there are now many of the major insurers in the United States at least are saying, for your enterprise risk policy, AI is not included. So if you are a major bank and you are doing big, important financial transactions, as soon as you start using AI, you’ve lost all your insurance.

And I think the Trump administration in the United States has a very light -touch regulatory approach. And my concern there is that, well, just because the government is not doing anything big and bold on regulation doesn’t mean there will be no regulation. The insurers will step in. And if the insurers exit the market, maybe not in legal terms, but in economic outcome terms, that could be very similar to draconian regulation. So, Shana, you’re mentioning the Underwriters Lab, which is an organization that writes standards that are relied upon by underwriters, the people who are issuing insurance. This is a huge part of the regulatory and governance ecosystem that I think is really important. And so now I’m hoping, Stephen, that you’re going to tell me, that you’ve been reached out to by a bunch of insurance companies, and they’re all reading your report eagerly and thinking about this.

But maybe, maybe not. What’s the case?

Not yet, but it’s a really long report. 312 pages, but it goes like that. Maybe I can come back to the best practices point a little bit, because I think we’re talking about auditing here, and at least I know there’s a lot of steps involved, I’m sure, but at least at the technical level, the main tool we have right now to audit the capabilities of the RISC -MD AI model are evaluations. And although in my opening I sort of talked about, oh, it’s great we have this toolkit that’s emerging and it’s strengthening, and that is true, I think on evaluations in particular, as far as like, okay, let’s say we have auditors that are looking at these companies, looking at models, what are they actually looking at to audit or evaluate the models?

I think we actually have a big gap here, a big evaluation gap in terms of, well, how are we actually assessing? So if we’re moving towards best practices, not only do I think we don’t have a sense of the best practices right now, but if we did, they’d be different in a year, because the capabilities are moving too quickly for these technical tools to be in date, for very long. So for example, you’ll have, you know, these evaluations often look like a set of questions related to a certain topic, and you ask the model, so you have a bunch of questions about biosecurity or a bunch of questions about cybersecurity. And if it’s above, if it scores high enough on the test, you say, whoa, this is a dangerous capability, and we need to implement more safeguards or something.

And as far as what’s best practice or safe risk management for a company, we evaluate in terms of, well, does it seem like the safeguards apply proportionately to the risk that you’ve assessed? But I think in many cases, these evaluations we’re using are already not super informative about real -world risk because they’re too narrow. Because you have to build a set of questions that gives you some information about the vast range of use cases in the real world. And as models have become more capable and general and adopted more widely, this has become much more difficult. And I don’t think there’s very many actors out there that are constantly thinking about new ways to evaluate the capabilities.

And so I think this… This is like an important gap in terms of our toolkit. that is, again, quite urgent because these models have been released and we’re using our current evaluations, which are already, in many cases, out of date and not super informative about real -world risk. Shannon, do you want to jump in here?

Yeah. Stephen, I agree with you so much. I mean, all of us are obsessed with benchmarks because that’s kind of all we have, and they’re just so narrow. I spend a lot of time with organizations that we think will become these IVOs, and testing is so, so hard. I mean, think about this. We have a fundamentally stochastic system, so I can ask something 10 times, system 10 times, and I’m going to get 10 different answers. So what does that mean in a safety context? Another problem that we have, what a model outputs is not the same thing as what someone does with it. So think about in the context of mental health. Maybe the model says to 10 different people different versions of, I think you should kill yourself.

Nine times, maybe for nine of those users, that’s fine, they will laugh it out. But for one of those users, there’s going to be a real problem here. and also the multi -turn nature of AI. I mean, you build relationships with these systems and you ask long queries and the stuff just gets really complicated really quickly as technical minds could explain far better than I could. So what we’re trying to do here is incentivize better testing because right now the only people creating evals or eval organizations or doing God’s work, doing awesome stuff, but what does it mean? You’re the best meter out there. I mean, there’s not an incentive to go from good to the best.

And the other actor working, of course, are the labs. And I think many of the labs are actually attempting to be responsible actors here, but again, there’s an incentive gap. I think the only way you’re going to solve this is to have an ecosystem where all of the actors are competing to have the best services, to have the best evaluations, to have the best feedback, to have the best feedback, to have the best feedback, And we hope one day one of these IVOs says, I’ve developed a new type of testing that figures out this kid safety thing that no one has ever thought about. And then the next day someone says, well, we have to be better because then everyone will want to be verified from that organization.

So you are incentivizing ever better testing. And as Stephen says, I think that just given how quickly and dramatically the capabilities and the risks of these systems are increasing, we need really good testing and tooling that can keep up with that. And the only way to do that is to incentivize

So, Stephen, if I could come to you about what Shana just said. You pointed out how the state of the art in evaluations and assessment is constantly shifting as the capabilities are shifting. I sometimes hear the frontier labs say, yes, and that’s why we’re the only ones who can do the testing, because we’re the ones out there on the frontier. But Shana is making this point about misaligned incentives, which I think we saw. In a conversation you and I had a couple weeks ago in the XAI Grok undressing children kind of example, there’s perverse incentives sometimes at work here in terms of the companies evaluating themselves. So how do you reconcile that gap between the frontier AI labs often do have a unique perspective and a unique understanding, but also it’s really hard to see how we could ever be comfortable with them being the only ones assessing themselves?

Well, I can talk about a bit in the context of the report where we try to work with everybody to get the state of the science across the whole landscape. And there I think it is true that there’s this big information asymmetry between the people in the labs who both have the most technical capacity and also the most access to leading models and all of the information about testing and development and all of the information about the technology that’s being used in the lab. And if you don’t draw on that knowledge, you can’t really do anything about it. you’re not going to be able to understand what’s actually going on in the AI world but then I think we brought in a lot of perspectives from academia and society and government feedback to sort of get a full perspective of the landscape as far as what to do going forward to deal with this I think probably it looks something like this with partnerships that are aiming to draw on that knowledge but then aiming for transparency and information sharing that gives third parties and external actors a better understanding of what’s actually going on because it’s true like even writing the report we were reliant on these papers that labs will occasionally publish and drop with like very useful data on how people are using the models or adoption rates but we’re kind of reliant on these like ad hoc publications and then that leaves a lot of gaps across the landscape and different risks and so we you know constantly had the word uncertainty or unknowns in the report because we lack that data outside of the labs

And do you think that that’s likely to remain the case, or do you think that that could change over time? As we’ve seen, literally, the safety staff of some of these labs quit and start their own auditing companies. So are they likely to have their skills atrophy as they get farther from the development process, or do you think it’s credible that these third -party organizations can build, the word that comes to mind is like economies of scale that are relevant to be able to continue advancing the state -of -the -art of safety and governance, even as the technology keeps evolving?

I’m not sure, but I think what we can do is sort of look at the trend, and the trend is towards, I think, a stronger ecosystem around AI labs. As more people, as these problems of lack of data and lack of independent verification are identified more, there’s more people working on it. And then I think we’ve seen some movement towards greater transparency with AI labs as well. So frontier safety frameworks are now a governance mechanism that’s in the EU. AI are in the code of practice, and it’s become institutionalized. It started as a voluntary, anthropic, just published, a responsible scaling policy. And so you see these movements towards sharing more information in more structured ways.

I think also, yesterday, there were the new commitments from the companies at the summit, which were related to sharing data about usage. So I think as a broader set of actors in society are paying attention to AI, because, again, we’re feeling the effects more clearly. It’s becoming more of an economic priority. We’ll see more demand from outside the labs to share this information, and maybe that will lead to some changes.

Hiroki, you’ve written a ton about AI, but in your capacity as a lawyer, you also have a lot of understanding of many different industries. Are there any lessons learned from other industries here that have solved this sort of technical expertise exists here, but the need for independence exists here? What kind of precedents do you see that we can learn from?

Okay, so before that, let me add one more incentive, which is public procurement. If the government says, we recognize this is a very important issue, and we recognize this is a very important issue, and we recognize this is a very important issue, and we recognize this is a very important issue, and we recognize this is a very important issue, and we recognize this is a very important issue, LLM or model is safe and then government procures this standard then it will be a big incentive for developers so that is one thing and when I try to answer your questions I think democratic debate is necessary as to what kind of risk level is acceptable and also what kind of test measures are good because there is any single specific answer as to this is acceptable level of perspective.

For example in Japan every year more than 2 ,000 people were killed by a human driven car and the question is what kind of safety would we require for the autonomous vehicles? Is it okay if the kill number is less than 2 ,000 or would we like to require more safety than human drivers? If so, what would be the level? There is no single answer to that kind of question so we need to debate. in a democratic manner as to what is our acceptable goal. And also about the test measures. For example, we can just simply compare the number of rates per kilometers, but if you test in a very safe straight highway, of course it’s easier to get to safety.

While if you try to drive in a pretty complex city, it’s gonna be very difficult. So how to measure how to define the test method is another question. And I don’t go into the details, but the thing that discussion has been done in a lot of industries, car industries, or finance industries, or aerospace industries, we can certainly do a lot of lessons learned from the existing.

Yeah, one analogy that as you were talking, you jogged my memory, is the National Highway Transportation Safety Administration in the United States, which actually industry begged for this organization to be created. They did. in the 60s and 70s because they said, look, all of us are going to claim that we have safe cars, but only some of us are making big investments in becoming safe, and we want to reward the people whose good behavior is making big safety investments. And so they created this new organization which would give cars a safety rating on one to five, five star or one star. And so now the companies can only get a five star rating if they’re actually doing what it takes to be safe.

And consumers, you know, they’re not always qualified to rip open their car’s engine and see what it looks like under the hood, what’s safe, but they can interpret that five star rating. And so my idea was to ask you, Shana, to elaborate on this in the context of your model, but I’m now scared of the beeper, which is quite loud and scary. So please join me in thanking our terrific panel. Thank you.

Session started. Thank you. the Netherlands, and Indonesia, as you’ll see reflected on the panel. And together with our distinguished panelists, we’ll explore how artificial intelligence can support the transition towards food systems that are more transparent, responsible, and inclusive. So this session is bringing together leaders from government, industry, academia, and international organizations to examine both opportunities and the practical challenges ahead from data sharing and infrastructure to governance frameworks and the partnerships needed to ensure that AI benefits are broadly shared. And before we begin the panel discussion, it’s my honor to invite His Excellency, Ambassador Fawai, Ambassador -at -Large and Special Envoy for AI of the Kingdom of the Netherlands, who will deliver welcome remarks. Welcome, Ambassador.

Thank you, Sarah. Is this working? Yeah. Thank you all for sharing this wonderful moment for me because we’re here with Madam Gorshan and Admiral Samari from Indonesia. Together we formed the chair and co -chair of the sixth working group on economic growth and social good in preparation for the summit. And I just wanted to say how much I was impressed with you, Madam Gorshan, how you managed the working group and how the outcomes were drafted and delivered, especially also delivered in the plenary. It’s not up to me, but I say well done. Really great. But thank you very much. It was really a wonderful journey with you. So, ladies and gentlemen, the use of digitalization and artificial intelligence in agriculture is developing rapidly.

It offers enormous opportunities to increase the productivity and sustainability of local food production. It offers opportunities to improve nature conservation and to foster a sustainable foster climate resilience in an inclusive and sustainable way. When this is all – when this – it also contributes to the autonomy and stability of countries. For the Netherlands, strengthening global food security is a strategic priority. Reliable, sustainable, and affordable food systems are essential for societal stability, economic development, and particularly in vulnerable regions. The ambitions in our digitalization agenda for agriculture, nature conservation, and food are to connect digitalization to the transition of agriculture needed for more food security, reduction of environmental impact, and climate resilience via public and private investments. Our primary focus on increasing productivity with lower environmental impact and improving climate adaptation, strengthening the resilience of food systems through response.

use of AI and digital technologies. Concerning today’s topic, the Dutch ambition is to enhance food security by making food systems more resilient and sustainable for all stakeholders. In my vision, digitalization and AI are powerful tools for that. They have already proven that they can significantly increase food productivity and reduce food losses. In addition, AI solutions can enhance the efficiency and resilience of food systems by supporting farmers to respond to sustainability requirements, make risk assessments, implement sustainable farming practices, and enable them to provide trustworthy and quality data sets about those efforts to be shared throughout the supply chain. The Netherlands has a strong AI ecosystem. Thanks to our technical universities and partners, we have a strong ecosystem of AI and companies like ASML, NXP, and Philips.

Despite its relatively small size, the Netherlands is not only a huge trader in agricultural produce, but also a global key player in agro -innovation and technology development due to the interaction between plant and animal science and technological knowledge systems in the Netherlands. Companies, science and government invest mutually in solutions for societal challenges. Examples include precision farming with AI, such as water savings of up to 90 % through smart irrigation, optimal crop yields with minimal input, and predictive models for disease control. To support digitalization in the agricultural sector in low – and middle -income countries, the Netherlands facilitates Dutch ICT agribusinesses to collaborate with businesses and startups there. And as you are… We are aware in the Netherlands that strong ICT ecosystems and highly innovative agricultural ecosystems come together.

ICT agricultural solutions combine the in -depth agricultural knowledge and advanced technology development in my country. Examples are applications for early warning of pests and diseases, optimization of water use and optimized plant breeding processes. Dutch companies and knowledge institutions are open to co -work on tailor -made solutions. Every country has its own typical local challenges and requires tailor -made solutions. Today special attention will be drawn to AI -powered solutions for small farmers and SMEs in producing countries in order to enhance their access to global agricultural supply chains while protecting their data. Our goal is to improve the ICT ecosystem and improve the ICT ecosystem in our country. We are committed to work together on this through knowledge sharing, co -operation and collaboration.

creation and capacity building so that AI solutions are locally relevant, inclusive and accessible to farmers. The need for an inclusive AI has also been central to our discussions in the working group of the Economic Growth and Social Group leading up to the summit. It fits well the summit motto, people, planet and progress. So I would like to thank India for its leadership in focusing on an inclusive AI future and underline that the Netherlands stands ready to contribute by forging concrete partnerships, sharing knowledge and technology while striving for measurable results in order to ensure that AI serves all of humanity. And I recall the Honourable Prime Minister’s speech in Flendry to which he alluded as well.

Ladies and gentlemen, we are honored to organize this important event together with the OECD, the go -to organization when it comes to AI governance, and to discuss the opportunities for international knowledge sharing and cooperation with FAO, the Wageningen University in the Netherlands, and the distinguished co -chairs of the Working Group on Economic Growth and Social Growth, India and Indonesia. We warmly thank India for hosting this summit and look forward to continuing and strengthening our cooperation in the field of AI and agriculture, both bilaterally and within the global partnership on AI. We also thank our co -chair Indonesia for continuing cooperation and we would like to highlight our appreciation and firm support of Indonesia’s ambition to join the OECD and its commitment to global standards and evidence -based policymaking.

International knowledge sharing and cooperation is needed to accelerate the development and application of new technologies. With the help of trustworthy AI. Having AI. And agricultural ecosystems on the agenda in this important AI summit is extremely valuable and a. forward in order to make a positive impact for all stakeholders. I wish you a fruitful meeting and look forward to our conclusions, and thank you for this opportunity to listen. So the floor is now Sarah.

Thank you, Ambassador. And on behalf of the OECD, I just want to thank once again the Netherlands for the leadership in convening this timely discussion. And as was just reflected in the Ambassador’s remarks, the Netherlands is obviously a pioneer in advancing food and agriculture innovation, and we are so delighted to have them as co -chairs as well of the OECD FAO Advisory Group on Responsible Agricultural Supply Chains. From the OECD’s perspective, we clearly see this dynamic of agriculture and food systems today operating in an increasingly volatile environment, and farmers face a wide variety of shocks, from droughts, floods, pests, to conflicts and economic crises. With growing frequency and severe… and so therefore strengthening resilience while also ensuring inclusion, as was also stressed by Ambassador Federe, is really an urgent global priority that I hope we can talk about today.

AI in this regard offers significant potential. We’re seeing AI systems and tools being applied to optimize the use of critical resources, as was already mentioned, such as water, fertilizer, and pesticides, and also to reduce environmental pressure while enhancing productivity. The OECD and JPEI, which also met today in a ministerial session, have been examining AI use cases in agriculture with a focus on the EU and on Southeast Asia, and we continue these dialogues. And what we’re seeing there is that the evidence from real -world deployment is really, really promising. So, for example, AI -enabled precision spraying has reduced pesticide use by up to 30 percent, and this is actually without compromising yield. while computer vision green on brown systems can cut herbicide used by up to half by targeting only the weeds that require the treatment and thus not the crops.

And in addition, we’re seeing how forecasting, monitoring, and early detection of climatic and biological threads means that AI systems can strengthen our capacity to respond to crises before they even escalate, so some degree of preemption. AI is also revolutionizing agricultural innovation itself and supporting more efficient plant breeding that can develop climate -adaptive variety in a fraction of the traditional time. And here we also have some interesting data seeing in Central Europe that researchers have identified drought -tolerant traits in crops such as sorghum and chickpea that boost yields by up to 25 % during end -season drought. And in Asia, meanwhile, we’re also seeing global AI hybrid rice platform demonstrating how AI can shorten breeding cycles by predicting optimal parent combinations and enhancing resilience in one of the world’s most vital staple crops.

Beyond the farm gate, AI is also reinforcing the resilience of our entire food supply chains. And AI -enabled traceability, market transparency, and smart logistics can reduce losses, improve compliance, and strengthen food safety systems. Evidence from these digital traceability initiatives across the OECD members demonstrates a growing maturity of exactly these systems, so something really to look out for. But technology alone, as we know, does not ensure impact, and so adoption is where we’re really looking now, and that remains quite uneven still. And this is obviously why we’re all here in Delhi. So while we’re seeing in Australia that 96 % of farmers are using digital tools, the same number for Chile is just 12%. And this is highlighting a digital divide that could deepen existing inequalities if we don’t look to address it.

There’s also important challenges in the use of AI, and this goes back to sort of the core work of the OECD, looking not just at the benefits but also the challenges associated with AI. Farmers and regulators need transparency in how AI systems make their decisions, but at the same time fragmented data governance frameworks introduce complexity to the use of AI tools that support the trade, traceability, and resilient food supply chains across the border. And this highlights the need for greater interoperability, which is also a theme at this summit. So structural barriers including high cost, limited digital skills, and lack of trust. These are some of the things that continue to slow the uptake of AI.

So bridging these gaps, which should be a priority for all of us, requires investment in connectivity and other digital infrastructure, in skills and affordable solutions. So smallholders, women, farmers in remote areas who play a critical role in enhancing global food security, they’re able to also benefit from AI’s potential. And farmers must be able that their data is collected, shared, and used responsibly. So in this area, the OECD is working to help countries put in place policies that promote these objectives through an AI policy toolkit. And this toolkit will provide practical, context -specific guidance to countries. The toolkit builds on our policy navigator. If you haven’t already visited it, it’s on osd .ai. And it so far covers more than 2 ,000 policies across 80 jurisdictions.

So this is where you can find examples. Examples of national AI strategies, but also in specific sectors. And we continue to update this, and for anyone in this room representing a country not represented, we encourage you to visit and to also contribute your policies. We’re also advancing work on digital governance in agriculture. This is within GPAY that I mentioned earlier, a priority there, where we examine governance models across countries and their applications for responsible digital transformation more broadly. We also see strong complementarities with the global AI impact comments, which is a key deliverable of this summit, and which shares concrete use cases of AI with known impact and scaling potential. So for the OECD advancing trustworthy AI consistent with our OECD AI principles requires a strong enabling ecosystem alongside technological progress.

And what we’re seeing is that if we succeed, we’re really in a position to raise productivity. sustainably and also strengthen resilience in agricultural supply chains, including by ensuring that the benefits of innovation are widely shared and existing divides are not deepened in the process. So I really look forward to this panel’s insights to help us take this conversation forward, looking at practical pathways to achieve this vision. And with this, it’s my pleasure to introduce our esteemed panel. Many have traveled far to be here. So first, I would like to introduce Professor Arvind Sumari, who is an Indonesian Air Force officer and professor at the State Polytechnic of Malang. Welcome. And also we have with us, next to Professor Sumari, we have Mr.

Dayan Jakoblevich. He’s Chief Information Officer and Director of Digital FAO and Agroinformatics Division at FAO of the United Nations, based in Rome. We also have with us… We have with us today the pleasure of having Debjani Ghosh, Ms. Debjani Ghosh. Distinguished Fellow and Chief Architect of NITI Frontier Tech Hub. And finally, it’s my pleasure to introduce Dr. Arun Pratihast, Senior Researcher at Wageningen University Environmental Research. So welcome to this session. And what we will see today is each of our speakers bringing a unique perspective on how AI can help build food systems that are resilient and inclusive, which is the topic of the session. And after the panel discussion, I will also be giving the floor to anyone in the room who might have questions.

So now let’s begin. I’ll hand the floor over to Dan, who will set the scene for the conversation. Dan, you have the floor.

Thank you very much. And I would like to welcome everyone on behalf of the Food and Agriculture Organization. I thank you to our hosts here. The summit from India, but also ECD and. government of the Netherlands ambassador thank you when we look at agri -food I heard in the interventions before about the agriculture and the food we look at agri -food systems from the FAO perspective why because the food itself as if we look at the agriculture food is one product but not only one so there is a whole ecosystem behind agriculture of products that are not necessarily food and they are equally important when we make considerations when we look at for example at the water use transport and many others so in from agri -food systems perspective AI brings us fantastic opportunities and if we look at our topic today in terms of inclusiveness and resilience and inclusion and inclusion and inclusion and resilience and inclusion and resilience and inclusion and inclusion and inclusion and inclusiveness and resilience and resilience and resilience inclusiveness is still a big issue if we just think back back maybe two, three years before the, let’s say, chat GPT came out, the inclusiveness and the digital divide was still strong and present.

And the key issue is that it used to be possible to exist outside of the digital ecosystem. We all know we could maybe go to the bank, but nowadays it’s not. So if a farmer or communities are outside of the digital ecosystem, they suddenly are outside of any ecosystem almost. And now with the AI, it makes it even worse. So this is something we need to continue to press on and jointly in making sure that everybody has equal opportunity within the digital ecosystems. And on the positive, let’s say, note, on the positive, let’s say, note, on the AI when it comes to inclusiveness. We see very encouraging opportunities with AI. What I mean by that is we can, in fact, lower the entry barrier to knowledge.

Just two days ago, I’ve seen here actually this opportunity at the event, great advancements, the new tool that was produced by government of India where farmers can, with a phone call, as not everybody has a smartphone, can get advisory in the area of agriculture, from shrimp cultivation to pest diseases and similar. So this is great. The service can be in many languages. So this is a fantastic opportunity example where AI can help us actually lower the entry point to the AI. In the same time, for governments, it’s even more so difficult. to have the capacity to build the AI infrastructure to provide such services. So this is, again, I think one area, and forums like this help us consider what it takes to build it.

When we look at the resilience specifically, I was very happy to hear in the previous openings you mentioned resilience in terms of, Jeff and from Ambassador, we heard on anticipation. So I would say this is the key word. The key word is anticipation. So anticipate the shocks to the agri -food systems that impacts the food security. We know we have natural disasters. We know we have also conflicts. We have many different factors that impact agri -food systems. So building the systems that are capable of absorbing the shocks of these situations and anticipating. Anticipatory actions to when the shocks happen, what can be done to kind of. go over these shocks. So this is where AI can be a great enabler, where we can then, with new capabilities, anticipate these shocks, and with the help of data and our joint work, really, put together decision -making tools, anticipatory tools, situation rooms, to be able to quickly not only anticipate, but when something happens, we don’t really improvise, but we have tools in hand to address these situations.

We still have about 700 million people without food on the table today. So from this perspective at FAO, and I’m sure we shared the same sense of urgency to actually do something. So I wanted to say from this perspective, we are very grateful to be part of this conversation and thank you for your time. And we can work together in finding the new solution. So I thank you for that. and I’m looking forward to our panel. Thank you.

intelligence research group and are the co -inventor of the Knowledge Growing System, a cognitive artificial intelligence framework designed to enable adaptive and evolving decision making. So from Indonesia’s vantage point, we’d be interested to hear where you see the most significant AI capability gaps across the agricultural system and where you see the greatest opportunities at the same time for AI to make food supply chains more efficient and resilient, something we also heard as a priority. And we also know that Indonesia is one of the countries advancing an ambitious AI agenda. So if you could briefly outline also the key pillars of Indonesia’s AI roadmap, this is of interest and to explain how you are balancing horizontal AI governance with more sector -specific regulation in agriculture.

Over to you. Thank you.

Thank you, Sarah. First, I would like to deliver my appreciation and congratulations to the host, India, and also my chair, Ms. Goss, and also my dear colleagues from the land ambassador harry first letter for coaching our working group together and also other speakers and Sarah thank you and our audience regarding your question about the artificial intelligence for Indonesia as we already know together that Indonesia is not only the agriculture but also maritime nation we we were self -sufficient in in rice about 20 30 years ago and then it wasn’t a I for making our country had sufficient in in rice but nobody I is something that that can make our program to be to become a self -sufficient country in right can be achieved.

We are much aware that the ideology is developing very fast, not only in America or Europe, but also in Asia, especially in Indonesia. This rapid and democratic application across all agricultural potential areas presents significant challenges, especially given the potential location which are separated by ocean. And you already know that Indonesia has 17 ,000 islands separated by ocean. We only have 36 % of land, 64 % of water, and 100 % of air. And this is a challenge for us. If you don’t believe me, you can count the numbers of our islands. And this is a challenge for us. And we also have another challenge. We are living above the ring of fire. There are also other challenges for our people of Indonesia. And as I mentioned previously, this gap is further widened by lack of democratically supporting AI infrastructure, such as telecommunication.

We have three different times region, the west region, center region, and eastern region. And each one has different one hour, one to another. And also, there is a problem with unequal distribution of AI talent. I think the problem is not only in Indonesia, but also all over the world. In terms of the biggest opportunities for utilization. AI in the food supply chain, especially in agriculture country like Indonesia. efforts to do such as like we can use AI for prediction of soil condition and nutrition before opening new land for agriculture. Our president has a program to open almost 1 million hectares of new rice files. 1 ,000 hectares in some big island of Indonesia in order to get the safe efficiency in the next five years.

And then we also use the AI for prediction of the most appropriate food crops given the soil condition and nutrition of existing agricultural land. We have seven dozen islands and each island has different soil condition, different soil nutrition. And you can use AI to predict what kind of nutrition, what kind of soil condition, what kind of vitamin that belongs to that soil. So we can predict the proper crops, the proper plants that have to be planted in that area. The second one about optimizing the most optimal fertilizer content to produce the best harvest result as well as optimizing the volume of water required according to the type of fertilizer given. Some of my students, they did some experiment how to predict the percentage of fertilizer combined together to get the most optimum production of any kind of crops.

Even if it is corn, rice, or sweet potato. And then we also can use AI for intelligent farming. We don’t say smart farming. Smart is not really intelligent. Intelligent is different. There is knowledge that has to be grown in the system. So intelligent farming is just like a human. They grow their knowledge within their brain. By optimizing the seed planting in the land so that plants can grow and develop healthily to produce the best products to optimization of the harvest process until delivery to logistic warehouses. So it’s just like end -to -end mechanism. And then we also can predict the weather dynamics just as a short step of the flood and something like that. So we can predict the weather dynamics to obtain the right conditions.

So that’s the vision for planting seed and reducing the level of crop failures. The crop failures that… This often happens if the farmer, they fail to predict what kind of pest, what kind of, what type of the soil and everything. And then the last one, optimizing the logistic transportation route to reduce the operational and other unnecessary costs. You can count how much operational costs to deliver the crop production from one island to another island in Indonesia. The price in the eastern area can be double or triple times in eastern area. So if we buy rice in eastern area only $1, it can be $3. $5, $6 in. eastern area. So that’s why we need AI to optimize the transportation and logistic transportation routes.

Whether it is from water, from the ocean or sea, and also from the air. Regarding the policy and regulation, you asked about the air roadmap, right? And then about how to balance the horizontal AI government with sector -specific agriculture, right? Yeah, we are proud. UNESA is proud to be a leader in our region, exploring how AI policy and regulation can be powerful tools for promoting trustworthy AI, especially in critical verticals like the agricultural sector. This one. Agriculture is very important to UNESA because most of the people in Indonesia, they are farmers not only in Java Island but also in other big islands in Indonesia if you see, there are five big islands in Indonesia from western area like Sumatra and then Java and the southern area we have Borneo in the central, also Sulawesi, or Celebes and the biggest one in the eastern area is Papua Island still have so much area that can be explored to become a rice field our national AI roadmap is not merely a technological blueprint it is a strategic framework designed to create an ecosystem that harnesses AI for inclusive and resilient system, including food system, so there are two keywords in here inclusive and resilient inclusive means it must be transparent AI must be transparent, AI must be explainable.

We’ve been having problems with the neural network -based system that the black box cannot be explained in plain. And then the second was Sicilian. This is very important for agricultural -based nation. So the implementation of AI needs a strong and sustainable national ecosystem, like my dear colleague, Ambassador, first of all mentioned about ecosystem. The AI cannot be implemented, cannot be applied without a strong and sustainable ecosystem that collaborate all stakeholders, not only government, but also business, industries, communities, media, and also academia. so we have a concept of helix maybe you ever heard about quad helix, five helix, six helix that’s very important so when we are developing the ANS roadmap the government in this case Ministry of Digital Information and Communication and Digital Affairs is open a voluntary contribution from all stakeholders not only the government but also from industry, academia media and communities so our roadmap has seven pillars that include AI regulation AI ethics, that’s important the third one is investment like it was mentioned before about financing when I was working the attending the US forum in AI export they mentioned about financing financing is very important, without that there is no AI ecosystem financing and investment and then the third AI data, the fifth one AI innovation and then the next one AI talent development the last one is AI use case so because we embrace all stakeholders so we assure there is no one left behind.

Thank you.

Thank you very much professor and we can come back to those in more detail later perhaps in the Q &A but I really want to thank you for sharing the promising use cases from Indonesia, very instructive I think for this discussion and now I would like to turn over you talked about the helix and how we work together to have the industry perspective from Ms. Ghosh India as we mentioned also co -chairs the summit working group and so I’d be interested to hear now that we’re seeing AI as quickly becoming foundational to agricultural productivity and food security but the big question now is whether as we mentioned it will deepen inequalities or indeed democratize the opportunity so from your vantage point Ms.

Ghosh what practical steps are needed to broaden access to AI capabilities so that emerging economies and smallholder farmers can also benefit and fully participate and as adoption accelerates hopefully broadly how should public -private partnerships evolve to scale responsible AI deployment and prevent the AI divide? Thank you.

It’s a very long answer question. I’ll try and keep my answer very short. But before I do that I have to acknowledge the presence of Yeah, okay. But before I do that I have to acknowledge the presence of I think one of the The biggest experts in this field of agriculture in this room, Professor Ramesh Chand, who’s also a very esteemed member of NITI Aayog. And I requested him not to come for this session. I’m going to be too nervous if you’re going to be sitting right in front of me. But yeah, let’s see if we live up to his expectations or not. You know, the biggest problem with AI today is that we throw AI at every problem that exists.

And we expect that something will happen out of it. Right. And as a result, we generalize the technology a bit too much. See, the thing with AI is if you really want to unlock the technology, you have to know what exactly are you solving for? What problems? And then you have to go deep because there are so much that has to come together for AI to work. For example, is the data in place? How good is the quality? Is the ecosystem in place? Are capabilities in place? So AI requires investments. And AI is a pretty deep investment overall. Right. So it’s very important to understand what problems do you want to solve with AI. And I think that’s one of the biggest issues today because we are not taking the time to think through it.

We keep saying AI is the magical world for everything. Right. So now let’s look at the food system. And I hope I’m correct. Professor Chan, I’ve learned this a bit from you also. But I think the biggest issue today is while the world is producing enough food to feed, I think, 8 billion people. But there are still millions and millions. Who are hungry. So there’s a paradox. And I think when you start breaking it down further to understand the exact problems as to why this exists. distribution? The entire access to food, do you have access to food so there is surplus and there is deficiency and then you don’t have a bridge to ensure that there is distribution happening at real time that is needed.

And what this results in is tremendous amount of food shortage, food wastage. And some of the culprits when we think of it, of course geopolitical wars are a big culprit, conflicts are a big culprit but climate is another big culprit. So this is how you sort of at least how I, because I look at everything from a tech lens, I’m by no means an expert in the domain but when I look at it from a technology lens and I say how do I best apply the technology to this problem, this is the domain that we have to play with. So now when you look at it if I have to say where do I want to go deep the problem to solve for at least when I look at all of this is the biggest problem to suffer in the food supply chain according to me right now just purely looking at it from a tech lens is the wastage.

How do I bring down food wastage? What role can AI play to bring down food wastage? So then you start looking at logistics, you start looking at supply, the cold chains that exist globally or not. You start looking at trade, you start looking at geopolitical agreements because all of that will come into mind. Now in terms of industry coming together to solve for AI again if you want the best out of industry you have to ensure that there is alignment on the problem statement you want to solve. Otherwise everyone will come and everyone will do the same pilot everywhere. That’s what’s happening today. When you look at AI executions around India and around the world, and because of the AI commons that we have built, every country is trying out the same thing, farmer advisory, right?

Every country is trying it out, but why is it not scaling? Why are we not solving for other problems? So again, it’s very important to identify what is the problem statement? How do you ensure that when industry gets involved, there is a route to market? And there is a route to commercialization because that becomes very important for industry. And one of the things that we advocate is coming up with maybe a center of excellence, a center of innovation that is identified to solve specific problems. I think one of the problems today we have with COEs are you have AI COEs, you have blockchain COEs. I really don’t understand what that means. But what if we had a COE to say that how do we ensure that the cold chain problem is solved across the country?

How do we have a COE that ensures that climate resilient crops in XYZ areas can be grown, right? And then bringing the industry together to say that how do we collaborate to create, I think gives you the right kind of outcomes. Thank you

very much, Mishkosh. And this is a perfect segue, I think, to our next speaker, turning to the research community and how to really bridge research into advanced AI to more practical tools. So, Dr. Pratihast, I would like to turn to you now for, you know, some examples of, you know, how these advanced AI tools can really be made to good use in more low -tech farming environments. And maybe you can give us some concrete examples, what distinguishes those who succeed from those who don’t. Maybe speaking also to some of the points that Mishkosh raised. Thank you. Thank

you. Thank you for invitation. It’s very timely discussion. And of course, always when we talk about AI, we often talk about the technology, how fast the model are, how big the data set they can handle, what are the parameters. That we always talk. But if you think about the food system, and of course, Terry mentioned that, you know, food system have different layers. And bottom of this layer is basically a smallholder farmers. And that farmer operate in a different environment. If you look at it last year, there’s billions of euro investment has been done in the tech industry to build more models. Is the same thing happen to the smallholder farmers? No. So there is often there is a problem that what we want to solve in the server room or computer, it doesn’t work in the field.

Right. So we. Really need to think how. the AI or model which we are really developing that is applicable to the grassroots level. And so within the Wakeningen and personally, I have been working in Asia, Africa and Latin America. And one of the problems, basically, there is three problems we are facing in this whole AI domain nowadays. First is really data scarcity. Still, there is not enough data. The data is not shared. As you mentioned, there is no ecosystem. There is no fair infrastructure where data can be shared. And that hinders the model. The model works on a global scale, but when you want to work on the local scale, it doesn’t work. It doesn’t provide the input that is expected for the smallholder farmers.

Second is the trust. Often, the farmers don’t own, and then, of course, the… the model and the farmer’s expectation is different and then there’s often not much trust how to apply this in the local level. That’s why most of this advisory is failed. Farmer doesn’t follow the advisory because it doesn’t make sense. And then third thing is scalability. Often we think that scale is not only the technical scale. Like you process something fast doesn’t mean that it can apply the same way. So we need to really think differently. And that’s why like we started I give a couple of three concrete examples. One example about food security. We need to understand what is the map.

Where are the crops? There is no global map that is accurate enough. So with the help of European Space Agency four years ago we started the World Cereal Project where we try to map the global crop length. So we started the World Cereal Project the World Cereal Project and we started still the maps are not perfect because India, China, many countries they don’t share their data so there is no data and if there is no data we have fantastic model we have built very nice geo -embedding with NASA harvest but applicability of this model is still very low in this country second thing is about high tech solution in low tech environment for example chocolate industry cocoa, agroforestry is really suffering from the climate change and we have established many advisory services but not from the researcher or tech perspective but engaging farmer perspective and that works we build basically chatbot with their language that really understand what they need and how we can translate their problem they know which disease are coming so we are using computer vision from their lens and then we are training and that works So there are a couple of things which we really see that if you really want to make these things working, you need to make sure that these solutions should work in low -tech environment.

Most of the things, connectivity has gone up. People are on social media, but still data is not there. Data infrastructure is not there. And always tech industry or like we as a modeler, whatever we call. So we see always data as the input and output. Data should be as the infrastructure. We should engage farmers in that infrastructure. And then only we can achieve the

Thank you very much. And I think with this, unfortunately, we’re coming to a close on time. I think maybe the speakers can be kind enough to stay a little bit after if there are questions. We won’t have much time for Q &A. but just to thank you all for really providing a diverse set of perspectives for the timely discussions to the ambassador of the Netherlands for framing this important discussion and I think some of the key takeaways perhaps is that there is vast potential and we saw the Indonesian perspective of all these very concrete examples also Dejan talking about potential for anticipatory action and we heard about this global and even domestic paradox of food insecurity when there really is enough food but it may not be distributed enough or properly and also I think importantly that to have impact with AI we need to make sure that it is problem driven that it is driven by the local context and the farmers who need to use it and maybe lastly a very important point which is exactly core to the work we do at the OECD that to drive this adoption we also need to ensure that there is trust in what is produced.

And this requires, obviously, a number of factors, such as explainability and transparency and so on, and also responsible data collection. But just with that, let me thank the panelists for their rich inputs. Please do stick around a little bit for some questions, maybe in the margin. And thanks again to the Kingdom of the Netherlands for co -hosting this event with the OECD. Thank you.

Thank you. Thank you.

Thank you. Thank you. across CPUs, GPUs, SoCs, and AI engines that power cutting -edge compute systems worldwide. She brings a rare combination of deep silicon expertise, global product leadership, and national ecosystem engagement. She is deeply committed to talent development in the ecosystem as well. Please join me in welcoming Jaya, who will be moderating our session. Our first panelist is Dr. Thomas Zakaria, Senior Vice President for Strategic Technical Partnerships and Public Policy at AMD, Inc. Dr. Zakaria previously led Oak Ridge National Laboratory, where he oversaw the deployment of multiple world -leading supercomputing systems, including Frontier, the first exascale supercomputer. His career spans scientific discovery, national compute infrastructure, public policy, and global partnerships. Please welcome Dr. Thomas Zakaria.

Joining us is Professor Vivek Kumar Singh, Senior… advisor on science and technology at NITI IO. Professor Singh plays a central role in shaping India’s science, technology and innovation architecture. From R &D governance to university industry collaboration and state level innovation ecosystems. With a background in computer science, data analytics and experience in academic leadership at leading institutions, he bridges research depth with national policy execution. Please welcome Professor Vivek Kumar Singh. My apologies. And finally, we have Mr. Rahul Garg, founder and CEO of Moglix. Rahul has built one of India’s leading industrial supply chain platforms and has expanded into manufacturing and industrial finance, navigating the realities of scale, capital and execution in India’s industrial ecosystem. Please welcome Mr.

Rahul Garg. We will now be beginning the discussion. Thank you so much for joining us.

All right. Good afternoon, everyone. And I would like to extend a very warm welcome to each one of you for this session. And thank you for taking time to be here with us. So we are meeting at a moment when AI is no longer a niche technology. And these conversations have become foundational. And there is a shift in shaping the entire economies. And that’s the global impact that this technology can have. And having spent about three decades in semiconductor industry doing design engineering, I have seen compute evolve. From a single threaded processor to massively parallel AI systems. And that’s. stupendous growth that we have seen and a transformation of technology. And honestly, AI is a technology that is probably the most transformational that we will be able to see in our lifetimes.

And true AI leadership is something globally there’s a contest. Every country wants to achieve self -reliance and, you know, leadership in AI. And that’s the importance of this technology that we are talking. But true AI leadership itself happens when silicon, software, systems, and policy, all of these aspects have to come together to achieve that leadership. No one aspect can really get us there. And that’s what truly excites me for today’s session. We have experts who have the knowledge. In each of these, many of these aspects, and we will be, you know, asking questions and they’ll be sharing their perspectives on this, which I’m sure all of us will enjoy listening to. So coming to India, from what I see, India is truly well poised for this technology shift.

And we bring together engineering talent, silicon design strength, and a growing ecosystem of system and infrastructure partners, including manufacturing. But what truly defines and makes this moment different is the scale and the speed at which we are moving. So we do see a strong commitment, but what is also important is collaboration. No one country or one organization can truly achieve the results or be successful at this, but we all need to collaborate. We all need to become very aware because this is not a simple thing. It has the potential to touch human lives and humanity. At a time when we are in a situation where we are in a situation where we are in a situation where we are in a situation where we are in a situation where we are in a situation where we are in a situation where we are in a situation where we are in a situation where we are in a situation where we are in a situation where we are in a situation where we are in a situation where we are in a situation where we are in a situation where we are in a situation where we are in a situation where we are in a situation through this panel, today I want to look at three perspectives.

First, how do we continue to build the intellectual foundation? Second, how do we build manufacturing depth and supply chain resilience through a sustained investment model? And third, how do we build a credible, sovereign AI capability? I will get to Vivek. I’ll

Thank you, Jaya. This is a very important thing, very important question. I think India has already taken a call to go in a big way in the whole deep tech domain. And a lot of changes that we see happening in terms of AI compute, then AI data centers and so on. Recently, we all heard about the tax holidays for data centers that are going to be created in India. Also, platforms like AI Coach, because that’s very, very important. If you want to create AI applications for India, you need AI data, which is centered in India, which is for the context of India. So what I believe, when you talk about credibility and how credible we are into this deep tech domain, comprising AI, semiconductor, biomanufacturing, even other, areas what is very very important is that credibility doesn’t come only from announcements so what we what we really need to know and what we really need to do is to go at a scale and fortunately a lot of positive changes are already happening we have india ai mission we all know about that 10 000 plus crores for five years and it’s a very systematic effort where we have almost all so all seven pillars address you know all kind of needs that we need for ai and similarly if you look at semiconductors we we all know about what is happening in fabs also you know we know that india has a very strong ecosystem of uh you know vlsi design semiconductor design and so on unfortunately most of that ip is not with india but you know there’s a time when is also going to happen that india would also be owning a lot of ip so credibility i think uh for india would be very very important and this is coming not only as part of announcement but it’s it’s coming you know it’s coming it’s coming it’s coming it’s coming it’s coming it’s coming it’s coming you know as part of commitment for scaled deployments, for scaled growth, accelerated growth.

And what we see now is something, you know, which nobody could have thought of 10 years back or 5 years back. So we, I believe we are on the track and we are very much into, you know, into the whole realm of AI and semiconductors. And a lot of push is there and the whole ecosystem is evolving and we all, as we move further, we all are going to work towards, you know, creating a very, very credible ecosystem for overall growth of the sector.

Now, great insights. Thank you, Vivek. Now, moving to Rahul. There’s clearly a growing momentum to strengthen manufacturing in India. Given your journey, you have expanded Moglex from digital marketplaces into manufacturing and industrial financing. Do you believe the Indian private sector is truly ready financially and have the mind? set to take on long term investments that are needed?

So firstly, thank you for having me. I think the question is very pertinent because again, pre -COVID there was a very different environment, both from a geopolitics perspective, supply chain perspective. And I think the supply chain as a word started to become popular in COVID times. So and I think I take some pride in the fact that at least as Moglex, we have been part of seeing the supply chain journey in the country as well as continuing to now see the manufacturing journey. On the specific point that you raised on both from a will perspective, capital perspective, demand perspective, if you look at these three aspects of it. So I think the demand in the country clearly is growing rapidly.

And one of the changes that has happened, obviously India becoming the larger in terms of GDP size, consumer demand, people expecting faster and faster products, people want variety. products so on so forth so i think demand discretionary spend is increasing the one significant change that has happened post covid we see is that while the demand is growing there is also an increasing uh appetite for people to start building more and more manufacturing to also start to look at many of those being localized rather than just depend on global supply chains because obviously we have gone through moments where like we may not have enough mask capacity in the country we may not have enough oxygen concentrated capacity in the country and we some of those shocks kind of uh got both the private and the public sector realizing that there is a bare minimum manufacturing that needs to happen in the country for it to be truly self reliant at the population scale that we are in so i think that will has gotten generated the capital is starting to flow in i think the question on whether the capital is large enough and long term enough i think we are seeing increase increasing trend that there are clearly government will whether it is in terms of the fund that we have seen of 1 lakh crore, now 1 .2 billion dollar for specific AI deep tech, things like that.

But also private capital, which within this week, the numbers that I’m hearing is more than 100 billion dollar plus commitment from the private capital companies saying that they are going to invest into data centers, localizing, so on, so forth. So I think the capital is happening today. Is it happening broad -based? The answer may be no. No. But has it started to happen? And has it started to go from like maybe few hundred crores to few billions of dollars? So that is happening. Can we execute at the same speed and scale? Only time will tell.

Sure. No, there’s definitely an increased momentum. But along with manufacturing, I mean, I’m also biased more towards the design front based on my experience. I do definitely want to see lot more local startups. And Vivek just mentioned, we don’t have the IPOs. I mean, having our own IPs is one of the key steps. we need to take. So moving on, question for Thomas. If advanced fabs remain limited globally, where should India focus on in the near future? Where can we realistically create value in the next three to seven years?

Thank you, Jaya, and I just want to echo the sentiments that my colleagues here on the panel have mentioned, so I’ll build on that. So I think the opportunity for India is to move from compute to capability, right? I mean, that’s really where we need to be. And I’ll pick a couple of areas. So sovereignty and resilience gets intermingled. So I’m going to sort of keep those two things separate. Sovereignty is one where you are really trying to make sure that your data and your application or use cases are resident in country and it’s relevant to country. And that’s an area that is uniquely India to lead because no one else is going to do that.

It has to be done and you already mentioned the opportunity, we were with the CEO of Medi today talking about 50 ,000 startups. I don’t know how to get my head wrapped around 50 ,000 startups so I asked him, can you tell me who the top 50 are so that perhaps a company like AMD can partner with them and try to help them to mature. So that is on the sovereignty side. On the resiliency side the reality is that clearly India needs any sovereign country expects to have resiliency create their own IP and India should have the same aspiration given the scale of ambition and scale of population and here I think while we certainly should have an ambition to go up the development cycle to the leading edge of chip design.

I think there is an opportunity to also look at being part of the supply chain for leading edge deployment. So you don’t necessarily have to be at the two nanometer scale for GPUs or CPUs. There are critical technology in the deployment at scale of AI infrastructure where India can play a role. For instance, we know that the entire ecosystem is going to be driven to optics as interconnect technology, co -package optics. And there are clear supply chain that is not available globally. That is something that is being considered today. And leading candidates obviously today I would say are U .S., Japan, Malaysia. But those are the kind of niche areas where India can stab the jib.

And that is the journey where you are. really contributing to the first -of -a -kind or the nth -of -a -kind leading -edge technology. So that’s the way I would approach it.

Great insights. Thank you, Thomas. Now, continuing, today AI leadership is ultimately limited not by ambition, but by access to secure scalable computing resources. So, Thomas, continuing with you, you have led exascale class systems and now you’re working on sovereign AI partnerships globally. In the U .S., programs such as Genesys and broader national compute initiatives have attempted to systematically align infrastructure, research, and industrial capacity. So what lessons from these models are actually applicable to India?

So I think this is a great area for public -private partnership, in my view. The public part of it is a uniquely government function. Government brings both policy as well as the demand signal, particularly in the area of science and innovation, critical infrastructure, whether it is energy sector or national security, as well as uniquely government missions. And the opportunity here is to, I mean, India has supercomputing mission. I think there is an opportunity, and I think India is already thinking about deploying this national supercomputing mission and national scientific infrastructure that is on a trajectory to be at global leading scale. So today, countries like U .S. China. China is a particularly interesting example. China developed the intellectual ecosystem around HPC, which then translated to AI, over a period of 20 to 25 years.

It was intentional. And if you look at where the AI penetration, AI adoption, AI infrastructure resides globally, you can directly trace that to investments in sort of supercomputing mission that built the underlying infrastructure. So I think that is a great opportunity. Already plans are there. But it’s not a static view. So one of the things that I would encourage as we plan for the future is not plan based on where things are, but plan on where things will be by the time we deploy this kind of infrastructure.

That’s great. A future -looking planning is what we… Thank you, Thomas. Vivek, moving to you, from a policy standpoint, how do we balance national security concerns with openness and global collaboration?

Vivek Murthy Well, it’s a very tricky question, I would say. You know, for a country like India, we all know, you know, I mean, the kind of culture that we have in India is we have always believed in the fact that knowledge is a common good. And that is how, you know, our whole innovation ecosystem has been operating. Our universities have been creating a lot of knowledge and we all, you know, researchers, R &D persons, they have been trained with the fact that whatever you create should be for a, you know, for a common good. There were never efforts to productize them, to convert them into socioeconomic goods, to, you know, protect them with, you know, excess rights and so on.

So that was the common thing that we have been doing earlier. But what is happening now is that we are not doing it in a way that is, you know, we are not doing it in a way that is, you know, we are not doing it in a way that is, you know, we are not doing it in a way that is, you know, we are not doing it in a way that is, you know, we are not doing it in a way that is, you know, we are not doing it in a way that is, you know, we are not doing it in a way that is, you know, we are not doing it in a way that is, you know, we are not doing it in a way that is, you know, we are not doing it in a way that is, you know, we are not doing it in a way that is, you know, we are not doing it in a way that is, you know, we are not doing it in a way that is, you know, we are not doing it in a way that is, you know, we are not doing it in is a completely different word.

And that is where our academia, our R &D institutions are also being asked to, you know, change the complete quotes. So it’s not only that researchers, you know, faculty members in universities, they should end up with research publication, that’s all. So it’s very, very important that you productize also. Now what is happening, see, if you talk about the culture of innovation and how you see in terms of the global world that we are in, particularly for sectors like AI and semiconductor, I think what we need to do is we need to go for a, you know, a strategic decision -making in the sense that what is it that we want to do? So, for example, there are certain sectors where, you know, the setup that we are using has certain components which may be used in some critical deployment.

So in those cases, what we need is a set of clarity of rules. What is it that we would like to indigenize? What is it? that we would like to have built on our own? And what is it that we can keep open for rest of the world, for collaboration and so on? So I think what we need is, I would say two words would be important is strategic autonomy. Autonomy in the sense that autonomy where it is needed, but at all other places where we can collaborate with the world, where we can contribute in terms of collective knowledge creation, India can always play a role and India is playing a role

Great. Rahul, question to you. As AI infrastructure scales, demand patterns for chips and hardware will shift. How should Indian manufacturers position themselves early? And secondly, where are the first mover advantages?

I think. we are kind of late to the party in some sense in the semiconductor and chips some say it’s two decades three decades late to the party right so and then there are couple of countries which have a disproportionate advantage not just in terms of what is more popularly known as the 2nm and two three companies dominating that but also in terms of the entire ecosystem that is required around all of those factories and chipsets and systems so on so forth so I think for us I think the India journey will be its own unique path so that’s one thing that I’ve always at least over the last 20 years I’ve seen that if you were to wait for landline to become 10 % of the population we would have not had the mobile revolution if you had to wait for credit cards then I mean like it would not have happened right so I think the in this new era that we are living I think the manufacturers will have to find few spots which may not be as obvious which given the conventional way the countries and ecosystems are built and I think one of the good advantages of events like this is you start to have a very large population and smart and talented people throw darts at hundreds of problems simultaneously and maybe five years later we will say like okay we knew that these are the three things which will work or all of that kind of thing right but I don’t think there is like a today unique path I mean definitely does seem that we need to start building capabilities and capabilities need to be built design we have capabilities we don’t have the productize capability so that is one capability which needs to be built the manufacturing capability while we are starting with some of the fabs which are in the mid zone but the entire ecosystem of chemical suppliers like clean room suppliers utility suppliers.

How do you make sure that. there is enough packaging verification and many of that ecosystem getting developed. So all of those are going to happen simultaneously. So I think the opportunity remains in all of the areas. And I think therefore, at least my encouragement to even my management and the way we are looking at Moglex and so on and so forth is, I mean, you try 10 things. Do not be scared to try one thing or two things and then you fail. And conventionally also while in the Western and so world, there have been horizontal capabilities companies have built and scaled. In India, historically over the last 15 years, every startup, every large company has built vertical stacks of companies.

So they are doing an integrated. They may be chip designed to manufacture, to systems, to product. I mean, like that’s how just the model has evolved so far. So. I think that’s what vertical stack manufacturing all. parts of the ecosystem will have to give a shot and maybe over time will become horizontal.

That’s great. Thank you. So, you know, I do see quite a few students in the audience. So one thing that we are now facing is kind of with this technology. What is knowledge? How do we acquire knowledge? I mean, traditionally, we go to schools, universities for that. But today it’s at your fingertips. And with that advancement of AI, it’s just going to get better. You want to learn about something, you always have it on your fingertips. So what really do we need? How do we prepare the next generations to solve the problems of the future is the question. I mean, we cannot just stick around with our traditional ways of learning. And we have to scale and adapt to the newer ways.

So question for you, Vivek, how can we prepare ourselves and equip ourselves for this next phase that’s coming?

well I would say efforts have already started so that’s the best thing and as you rightly said this is the best time to be a student you know if you take yourselves 20 years back you will always be constrained with resources the best that you have is lot of books you will have to go to a library there are books that you can’t afford and so on and books are also not on time so you have later editions and so on so what is happening now is you know with lot and lot of information information which is which is you know can be customized for you specifically then you have lot of recommended systems you have retrieval augmented generation systems you know all of this with generative and what is happening so the best part is that you have plenty of information you want to learn anything you want to acquire a skill you always have resources and most of the time these resources you really don’t have to pay for that because this lot of material that you can access for for free.

The programs, particularly for India, we have something called NASCOM’s Future Skill Prime, where you can, you know, which is an aggregator for a lot of online courses. Similarly, there are platforms across the world that you can use. Now, what is happening is that essentially what we have been doing in our universities and generic colleges and, you know, other institutions earlier is that it was largely a kind of memory -based learning where we were acquiring knowledge, we were memorizing things. But now, over a period of time, it’s a more synthetic perspective which is being, you know, percolated across institutions. So, students now are going more into that creative aspect where they’re able to create solutions for certain problems.

And with the whole ecosystem around startups, we all know India is the third largest startup ecosystem of the world. With a lot of support system that we have, most of our universities have incubators. You know, other support systems. So, this is… best time and that is why I said this is the best time to be a student if you want to do anything if you have a creative idea you always find support and there are lot of skilling programs from government of India from many other organizations many you know philanthropic supports are there so even lot of organizations which have their own products they are you know offering free of cost training to students so this is very good but what is important is largely also due to the fact that you know we keep on hearing that AI is going to cause a number of you know disruption in terms of jobs that are there because lot of jobs which were there in areas like software testing and customer support and all of that is gone but at the same time these technologies are also creating new jobs and for that you need to prepare yourself and fortunately the best part is that we have enough material enough resources enough support system that we can use to create a new job and for that you need to prepare yourself for the new kind of jobs that are going to come so this The whole revolution that we see in front of us will require massive skilling, a bit of reskilling also.

So many of my batchmates, 25 years into the careers, and they now feel that they have to reskill themselves with many, many new things. Life was very good somewhere in Silicon Valley, 25 years, a lot of money, but now they feel threatened. And that is the beauty of startups and all these new ideas that are there. So I would simply end by saying that this is the best time to be

Absolutely, totally agree. You know, I have to share this thing. I was actually conducting a panel discussion within AMD with the senior execs. And one of the fun questions was, if there is a machine or an equipment that you want to invent, what would it be? And the unanimous answer is, I would love to have a machine that can make me 20 years old. 20 years younger, right? So, you know, you guys are extremely lucky, make use of this opportunity to the maximum. All right. So as we come to the final leg of this discussion, India’s opportunity in AI and semiconductors is very real. But it’s also time bound. Momentum alone will not be enough. Sequencing, capital discipline, institutional alignment and infrastructure depth truly matters.

And all these areas have to work in complete alignment with each other. So, you know, let me close the session by asking each of you one more question. First one is for Rahul.

was if there is a machine or an equipment that you want to invent, what would it be? And the unanimous answer is I would love to have a machine that can make me 20 years younger, right? So, you know, you guys are extremely lucky, make use of this opportunity to the maximum. All right. So as we come to the final leg of this discussion, India’s opportunity in AI and semiconductors is very real, but it’s also time bound. Momentum alone will not be enough. Sequencing, capital discipline, institutional alignment and infrastructure depth truly matters. And all these areas have to work in complete alignment with each other. So, you know, let me close the session by asking each of you one more question.

First. First one is for Rahul. In the global race where others are moving fast, what is the one move India must execute flawlessly to stay competitive?

I think like many other things I think it’s not one move so maybe we do everything as a Bollywood dance move right so they’re like 10 moves to everything but I think the one of the things which has happened at least from my vantage in the startup ecosystem is over the last 15 years we have become extremely good at being fast followers like maybe 15 years back if there was a product or a service in US or in Europe it would take three to five year lag to come to India and now maybe that lag is like one month 15 days I mean like so probably chat GPT within the first one month the maximum number of users are coming from India right so we have become extreme fast thanks to technology that we are fast followers the number of apps that might be built in India might be higher than most countries combined together maybe US China might be the only ones but otherwise I think India would be in the top three in terms of building all the apps in the world I think the move that needs to happen is the scale of ambition beyond India into the global platform because most of this effort that has happened in last 15 years are around kind of dominating the Indian consumer businesses applications so on so forth I think we need to up the game on global and we would require a significant amount of public private upping the game because most of the countries capital pools that we are fighting cannot be only attracted by the private players so I mean if someone is raising 100 billion dollar 200 billion dollar we need to at least start the race with 10 billion 15 billion 20 billion right which is not possible today completely in the private so I think how do we bring this and push the you capital bar, global bar together as a government and as private players I think that’s one thing I would love to see

That’s a very valid statement Right Next question to Thomas Thomas, if we had to place one strategic bet that defines India’s position in AI and semiconductors by 2030 what should it be?

So I’m going to repeat what Rahul said there is, you know, the one I don’t know much about Bollywood dance moves but I would say one move is certainly ambitious I’m going to sort of regress back to a few previous questions since we have a few minutes, I thought I will sort of start with public -private alignment. Rahul mentioned that it is very, very hard for private sector to to to raise the kind of capital that’s being raised elsewhere in India. And that’s part of it is, you know, so one of the important things that government can do is to de -risk that enterprise. Now, I don’t believe that government should de -risk a private sector’s business venture by investing in that effort.

But there are unique places where government can de -risk through public -private partnerships that would enable this ecosystem to develop so that additional ventures can be taken up by private sector on their own. Because I don’t think that my taxpayer money should be used to subsidize. I mean, look, there is a role. So you mentioned Genesis. I did not describe Genesis. I don’t know whether… of you in the audience know what genesis is so I’ll take a couple of minutes to just discuss that as an opportunity to think about how to frame public private partnership so today United States spends a trillion dollars a year in R &D expenditure and roughly about 20 to 25 percent of that is government the rest is private sector now if you look at the R &D spend in the United States it’s been steadily growing at about keeping up with inflation maybe slightly above inflation 2 to 3 percent year over year but if you look at innovation output it’s been flat lined part of it is because the problems are getting more and more complicated discovering new materials cure for cancer all All those things are increasingly, significantly impactful for society, but also significantly challenging.

So the goal of Genesis Project is to really, one, align public and private partnership, two, invest government resources to bring academia, national laboratories, and private sector to identify what they call lighthouse problems, so you can call it grand challenge problems, that are relevant, that is likely to move the needle across these areas. And the government is then investing substantial resources for compute infrastructure, software stack, partnering with private sector in these important problems. Because it is being done in a… open, collaborative framework, private… This work is, in my view, appropriate for government to invest because the government is not investing directly in any particular business, but the business is able to take the fruits of this collaboration to drive innovation in their own sector.

So I think that is a really good model. I think it was already alluded to. If you are a fast follower or if you follow anybody, the danger is that it may be appropriate for a business, but as a nation, anytime you follow somebody, and if that is your ambition, you are destined at best to be number two, at best, because there is always somebody ahead of you. So I think for a country with the history of India, the ambition of India, the talent of India, and now the will of India, there is nothing wrong with aspiring to be, strategically deciding where India can go. can be world leading in part of this, I mean, no country is going to dominate every aspect of this ecosystem, identifying strategically where one can be that leader globally.

And I would say there are, at least if I can speak to AMD, just as an example, we were discussing about Helios and how it is based on open standards. There are many components. It may not be GPU that you start, but there are many components there where a private sector in India can aspire to be a leading provider based on open standards so that a business like AMD or a public private sector would say, well, I can get a better product, better total cost of ownership if I can plug into that. And one last thing, I cannot let you get away with, you know, just this time. I’m being great for the youngsters. Ray Kurzweil said that today, for each of us in this room, we age only eight months for every chronological year because of advances in medical care.

And that is true because longevity, people are living longer because of better drugs, better health, living, etc. So AI has the added advantage of providing greater solutions. So it’s not just the youngsters, there is hope for us if

Absolutely. So we are all lucky to be here at this age of AI. We are truly lucky to be in this. No, that was very insightful. Thank you, Thomas. So Vivek, a question for you. I’m going to say what is the one bold decision, but I’m going to change that to what are some of the bold decisions. We must take to ensure we don’t look back and regret five years from today.

well i think uh what the the biggest advantage that india has is of course you know a huge pool of talent so that is something that we all need to rely on that’s that’s uh that’s the most important thing for a country like india and this essentially see india has an inherent culture of innovation so it’s not that you know we’re always following or we’re looking at technologies and so on the fact is the ecosystems where we have been living in they were not you know uh geared up they were not situated in the context that we were creating products so the culture of transforming that innovation to products has not been there unfortunately for a long period of time things are changing now and probably what we need to do is to invest more in our youth to invest more in skilling to invest uh more in how do we convert the knowledge that we generate generate in our universities, in our R &D labs into actual usable products which have, you know, socio -economic impact.

So that’s the most important thing that I believe we should be looking at. Of course, given the fact that we also have an advantage that we have the advantage of scale also. So, you know, a lot of things that we have done and we have proved it in terms of the digital public infra that we have created at a population scale of size India, it matters a lot. If you go to any part of the world, particularly anywhere in Europe, if you identify yourselves from India, you know, and you are in some discussion related to IT and so on, so you will always be regarded with, you know, a lot of depth in the sense that everybody believes that India is an IT superpower largely because of the talent that we have.

So this is something that we should leverage on and we should do it. And something that we really need to invest in heavily to see what is going to come for the next generation and to provide an environment and our prime minister keeps on saying ease of doing business so that is something that we really need to look into to to enable and to create an environment where we are able to transform the knowledge that we create into usable products

no absolutely right i mean talent skilling and ease of doing business i mean all of these are coming together for india and in fact i led the committee for future skills and i worked i got the opportunity to work with 13 other eminent leaders from industry academia across the board and one thing that stood out was if we can actually get our skilling right we can actually supply talent not just for india but globally you know that’s something that’s going to be very effective you if we can get our skilling actions right so thank you again Today’s conversation was truly insightful and inspiring. We touched upon many aspects of semiconductors and AI and the ecosystem and India’s potential as such.

And again, AI leadership will not really happen by accident. It will require a deliberate alignment across policy, industry, research, and infrastructure. And we have many strengths that we need to work on. We need to work on strengthening and leveraging for the growth that we are ambitious about. And truly what matters now is decisive execution, moving with clarity and with urgency. So it’s going to be a great journey. And I once again want to iterate that we are truly lucky to be here in this phase. And what a fantastic journey we have ahead of us. And let’s be committed to that journey of learning and advancement. Thank you also. much for attending this session. I appreciate your time.

Thank you. Do we have time for audience questions? We can take one question, one or two. Out of 500 out of 500 sessions here, this is one on semiconductor. I’m very glad that you guys organized it. Very, very insightful. A few amazing questions and a good response. Quickly to my question. I teach AI and sustainability at IIM and I cover the entire supply chain starting from the rear, going through the chip design, manufacturing, the semiconductor supply chain, essentially, and all the way to data centers and electronic use. So, sustainability is at the core of all design decisions in my class. And that’s what we are trying to teach the new management human resources in India. Your thoughts on having sustainability not as a trade -off but as a core design choice for every decision that is made either in India or some country.

So it’s a great question and I think every one of us, certainly I can speak for the company that I represent here and I must say I’m in India so I am going to give a shout out to the 10 ,000 AMDers in this country. AMD would not exist without you. We would not be able to do what we are doing without the contributions that they make every day. So India is already very much part of a global supply chain. Sustainability is very key. We design our products with a goal, explicit goal of flattening the energy curve because it’s easy to say we’re going to build megawatts and gigawatts, which we may because it is going to be a fundamental infrastructure in which society is going to progress.

But it’s incumbent on us to ensure that we are very, very thoughtful and committed to sustainability. I also would like to say that we have to be humble enough to know that we are not going to get everything right. I was at a U .S. National Academy meeting where Subhash Suresh, he was the president at the time, we had just rolled out the grand challenges for the 21st century. And he said, you know, if you look at it, the grand challenges of the 21st century. are attempting to solve the problems created by the solutions to the grand challenges of the 20th century. So the reality is that we don’t know what we don’t know. But yet, as long as we use sustainability as a core goal and be humble enough to know that we are not going to get it all right, then I think we cannot stop progress.

We need to continue to move forward. But know that we are not going to get everything right and course correct as we go along.

Okay, I’m told we are out of time. Yes, actually we are running out of time. And I really appreciate for joining us for this session. And I’m very much heartfelt thankful to our distinguished guests. So as a token from Mighty’s side, I would like to give a short, I mean cute memento. Pooja, you can also join. Thank you. Thank you. Thank you. Thank you.

precisely this, how do we sort of build capacity in order for this technology to be applied significantly better. And in the days to come, I would really love to see a day when India and the EU collaborate much more closely to make this happen, not just in India, but all over the global south. Thank you very much for having me. Thank you very much. Don’t go away, because now I’m going to call the panel. We have a distinguished panel today, but we would like to take a picture first. So if I can invite Vice President and Secretary Krishnan to stand here, and then I invite Arthur Mensch. He’s the co -founder and CEO of the European Union.

He’s the CEO of Mistral AI, if you can just stand next to the Secretary, which is a European company developing large language model, but also Jarek Kutuloski. who is the founder and CEO of a German company called DeepL, which is on language technologies. Matteo Valero, who is a professor of computer architecture at Technical University of Catalonia and the founding director of the Barcelona Supercomputing Center. And from the European Commission, I’m pleased to announce Roberto Viola. He’s the director general of DigiConnect. And he plays a pivotal role. He’s the director general for our digital policies. Okay, so as I said, it’s a very distinguished panel from the European Union. And I would like to thank all of you for being here to participate.

I’ll start with Artur from Mistral. I repeat that he comes from Mistral, which is a European model and one of the main large language models. In your opinion, how can LLMs or general purpose models in general reshape the public sector? And as a developer, how do you work with governments to apply it in the public sector?

I’m the co -founder of a company called Mistral and we effectively train language models and perception models and we then use them to create applications for businesses and for states typically the models is never enough to actually provide value for the states we work with we have a program called AI for citizens that have multiple pillars but when we work with states the first thing we work on is efficiency what generative AI allows you to do is to delegate tasks in general and to automate certain processes that can be fairly complex, that can be fragmented, that can involve multiple people, that can involve multiple tools that can deal with IT legacy and so a state is not different, an administration is not different from an enterprise in that respect, that they have IT problems, in that they have processes that are sometimes inefficient in that they have pressure on talents because there are a lot of people that are actually retiring, so knowledge is a very big problem and management of the ledges is a very big problem.

The kind of things we do is related to that. So we deploy our horizontal platform and we create use cases. We work backward from use cases that are around procurement, that are around writing reports on the, visible in that it can show to the citizens themselves is building public services on top of artificial intelligence. And so one example is we worked with France Travail which is an employment agency in France to actually help with the matching of job employers, of employers and of people seeking jobs. And often times people would just connect and they’re looking

Thanks a lot. I now turn to Yarek, founder of DeepL, which has been a very important part of the project. Yarek has been working since 2017 in AI -based translation tools. and so there is a lot of linguistic diversity in India as well as in the European Union and so how can the AI language models help to overcome this multilingualism issue I say of course we consider it also a benefit but in administration it can be sometimes be a challenge

I would definitely try to not characterize it as an issue I think it’s something that’s actually pretty beautiful about a lot of the countries that are so multilingual and there’s a lot of differences in how deeply multilingualism is embedded in different countries and in different societies I think here in India everybody understands it extremely well but it’s not the only country in the world and there’s countries like Canada, there’s countries like Switzerland whom we’re working a lot with the public sector that have this intrinsic necessity of being able to connect to their citizens in very many languages and where partially that communication is even embedded as a part of their constitution. And here, those countries have been struggling over the years, maybe as you have indicated, on how to actually make this happen.

And AI and those kinds of frontier models that we build and the applications on top of them that are specifically tailored to bridging this communications gap, they help a lot. Nowadays, not only in written language, but also in spoken language, enabling real -time conversations maybe with citizens in a setting when they come up into an office and want to get… certain service done. So a lot of options there, but also a lot of complexity as those use cases that governments have really differ very, very much based on what you’re doing. It’s another challenge to translate legislation into different languages. It’s another challenge if you want to enable those real -time conversations with citizens. Quite a lot of exciting problems to solve.

Thank you very much. Now I turn to Matteo Valero. You are also a professor, but you’re also the director of the Barcelona Supercomputing Center. So can you maybe explain, you’re also in an AI factory, what the AI factory does and how it can help the transformation of the public sector and of SMEs?

Thank you, Lucila. Good afternoon to everyone. It’s my pleasure to be in India once more. Sorry? Sure. My pleasure to be here. You have an incredible country, believe me. So, thank you for inviting me. And I am going to start 50 years ago when Seymour Cray produced the first supercomputer, no? And this supercomputer increased the speed from 10 to 10 until now, okay? With this computer, we did simulation and we produced better results in science and engineering. In Europe, every country was alone until, thanks to Roberto Viola, we created the EuroHPC. And then, because we had the EuroHPC, we have now a reasonable amount of power in the supercomputers. So, if you look at the top 500, probably out of the first 15, we have 6 in Europe.

And we do science. We do science and this is very good. So now because the data, because the computer, and especially because the research of these guys and many others, the AI is invading us. It’s changing any activity we have. In my field, I am changing the way we do high performance processor. In the supercomputing center, let me tell you that we are 1 ,400 and we have 500 people doing hardware software, using or designing in topics related with the AI. There is no question that now the data, the control, the data, the computers, and the algorithms are dominating the world. So what we could do in Europe, we have the supercomputer, but we need to devote more energy in order to…

get the AI distribute around any activities. So the idea of the European Commission was create the factories and now the gigafactory. The AI Factory is a platform, AI platform is hardware and software, but as important as that, it’s co -located where there are people with the skills in AI, there are people with experience in transferring technology to the society. So the idea of this AI is the service is free, the people is free, is to connect as much as we can with the society to make a better world. This is the target for us. Obviously, there are many, many possible contributions, and one of them is the administration, and obviously how we can make happy to the citizen.

If we make happy to the citizen, we are successful, okay? And we can make happy to the citizen if we provide them with personalized information, accurate and fast. After that, a second question. I will give example, but I think… So this is the target for the AI factories and the gigafactories is the same but competing with the data center. Because I forgot one thing. What Europe could do is just to use this data in the platform from outside or create our platform to use our platform using this data. I think this is the right way to go.

Thanks a lot. So we have talked about what the models can do, the computing capacity that is made available. Now, Roberto, I would like to ask you, since you have reflected and designed all of this, how would you now… Mention your words. I’m your boss. Yes, he is. How would you help now facilitate the uptake of AI? By the public sector, because if we have the models, we have the compute capacity or we are building it. and we’re also building more access and more availability of data sets. But how can we make sure that the public sector actually uses AI?

Thank you. Thank you, Lucille. Good afternoon, everyone. It’s really, for me, a pleasure to be here and to be together with Lucille, with the three crown jewels of Europe, which are all very much representing what is for us giving out to citizens, society, innovation. Because you can test the Mistral on the web for free. You can use DPL for free on the web and test it and enjoy it. Translation from all Indian languages to European languages, I dare to say, yes. You can test Destination Earth. Destination Earth is the most sophisticated climate digital twin of the world, AI digital twin. You can replay the climate of the past into the future. You can zoom in in certain areas.

You can have a resolution which was an error of 200, 100 meters, three weather events, because there are already two twins which are running, the twin of the climate and the twin of extreme weather events. Again, for free on the web. So this is the first point I want to make. There’s an economist, maybe you know the name, Mr. Solow, that he expressed with numbers and, I mean, evidence a paradox. The more people invest in IT and software and other infrastructure, the less the productivity. Actually, there’s no… There’s no productivity gain in doing that. So it’s called the solo paradox because it’s a paradox because you as a user, me as a user, experience a much better user experience to have a public administration which is more digitized or an hospital where everything is digital as well or a doctor which is savvy because it has an AI co -pilot.

But in terms of productivity gain, according to the solo paradox and the numbers that he has, in a compelling way, put in front of us, there’s no productivity gain. So many economies, and of course, who solves in this room this paradox is for Nobel Prize of Economy. So, I mean, the challenge is open. So I’m not going to solve it, but I try to answer the question of Lucy. The reason why many have observed this is that because normally, IT, and that includes also now AI, overlaps what exists. And of course then it becomes very intuitive. Imagine an hospital, I mean, having all the doctors still and nurses and everyone in traditional process. So doing a bit of paperwork as they do, but also doing it digitally.

And having two systems running in parallel, of course, I mean, you imagine that the productivity doesn’t move much. Now we have seen some changes during COVID. Why? Because people, I mean, were secluded and they were forced, I mean, to use only digital. So in certain areas, sadly, I mean, you saw the productivity was in a way more linearly linked with the use of technology. The European Investment Bank has published an econometric study that shows AI as a productivity increase of 4%. Which is not the stellar numbers, I mean, some of the vendors around say, but it’s… compared to the solo paradox is not zero. And I think this sign is because with AI, especially agentic AI, you see the change.

So you don’t see the overlap anymore, one process with the other, but you could see that there’s a new process, new way of man -machine interacting and working. But Arthur, before, said something which is the key of all of this. Because if people in the public sector are not empowered, they don’t understand it. They are not part of this change. The change will not lead to any productivity gain. Because you can have the most expensive and sophisticated AI software of the world, probably absolutely not needed by the private sector, because better to have bespoke models, open source, that serve the purpose. But even if you have the most sophisticated, you still get that one. if you have someone that refuses to embrace the technology or in any case you have an organization a process that is not ready not fit for it then there’s no productivity gain.

Maybe as a citizen you can see their wonders but in reality I mean the old system becomes two times more expensive and the adoption rate is low and this is really the real challenge of artificial intelligence as paradoxically it can be. I think we can proudly say that as I see it in India and I see it now in Europe, we are developing an ecosystem which is really brilliant, self -reliant, sufficient in terms of good company producing open source, producing language technology, producing advanced algorithms. We have supercomputing center offering capacity. All of this, I mean, goes in a completely different model compared to other models, and it’s all fine. But now, I mean, we really need to work with the people and with the public administration and to make sure that we

Okay. So, Arthur, if I were to ask you, how do you get the AI accepted by the citizens and also the public administration? What kind of tools? You already provide, of course, the chatbot Le Chat. But what are other tools that you think will be easily accepted?

Well, we’ve turned Le Chat into something that we call Vibe, actually, which is a product where we can delegate tasks. We can delegate tasks fully and delegate workflows. The challenge and the reason why you don’t see productivity gains when you deploy chatbots in enterprise is that basically you’re focusing on an individual productivity gain. so that’s the case in enterprise but it’s the same in administration and so if someone can actually write a mail faster it’s not actually changing the way your business is being run when the thing starts to change if you look at a full process let’s say procurement for instance which typically you entail like multiple touch points with multiple people and you ask the agent to actually run the process itself so you move from an individual productivity endeavor to a collective productivity endeavor and you move from being equipping ICs so individual contributors to equipping managers that are going to span the same way a manager will delegate sometimes to a human it can delegate sometimes to an AI process and there are two big challenges associated to that and that needs to be solved for product but also through human interaction I would say.

The first is that you need the process automation that we run and we design them we bring our engineers in, they work with subject matter experts and they design they write the code using our coding model and then they deploy the code that is going to run the automation and that’s going to ask questions, that’s going to interact with the tools. The way we design them is to try and get the humans out of the way because the problem is that the process only brings productivity gains if you’re not bottlenecked by the humans themselves, if you’re not interrupting them all the time. A good example is coding. If you want to code faster with AI, you need to give them tasks and then disappear and then you come back maybe one hour later and the task is done.

If the thing comes and nags you like five minutes after, maybe you’re doing something else and so the thing is actually not progressing as fast as it should. You need humans to actually get out of the way of the AI automation if you want this automation to work. And then the second thing that goes with it is that once you’ve done the automation, you need to rethink the organization because once you’ve automated your procurement process, well, suddenly the people that were actually running the analysis of the procurement needs to do something else. And that’s actually… Take… some thought around how you’re reorganizing people, how you’re rescaling people, how you’re turning individual contributors into people that will effectively manage AI -operated processes.

And so you need, and enterprises need, to actually turn people that were used to do menial work into people that are delegating those work. And as you know, and as every manager in the room knows, it’s actually fairly hard to learn how to delegate and to move from being an individual contributor to being a delegator. And because the only way AI actually brings you productivity gain is through strong delegation and long execution, well, every one of us needs to become a strong delegator. And so that takes some training. We are not trained to be delegators at school in Europe, I would say, at least in France. And that’s something we will need to learn. And so we’ll need to learn it early on.

And we need to reskill the people that are, that needs to learn a new way of doing things. A new way of working. And to rewire the brain. I think a very good example, I’ll stop with that. is that we have our coding tool called Mistral Vibe. And what we see is that if you take very young developers, they use it very quickly. And so they learn how to use it. They’re excited. The way they work is inherently wired to using AI because they are 23, and that’s everything they’ve ever done in coding is through AI. Then you have the very senior people that are like 35 years old, I guess, or 33 my age. And those are still much better than the agents themselves.

So they know how to design the architectures. They can give some very precise guidance on how the agents need to rewire the code or bring a new feature. And then the problem is the people in between. The people in between, well, they got very attached to writing code. And now they need to rewire. They were very good at writing code, but they need to become something else. else. And so that’s where reskilling is very, very important. And that applies to software engineering, but that will apply to everybody working on knowledge. And so that’s the three years ahead of us, and we need to work together to make it as smooth as possible.

Thanks. And indeed, Jarek, you have developed AI agents. So when you think of applying them to the public administration with all these caveats we just heard, what do you think the acceptance is going to be like?

Yeah, I think it’s not only about the individuals and how people reskill and how people adopt AI and how they learn to use AI, but it’s also about organizations really rethinking the way that they are working. Thinking about workflows, thinking about processes, whether that’s like general purpose agentic workflows, or whether that’s something that has language at its core, rethinking of how do we do these things. We’ve gone over a couple of decades now improving those processes and maybe putting parts of AI, especially in language processes, that’s been already happening over the last years quite significantly. But we haven’t yet rethought those whole processes. Like, do we need that human review step anymore in a particular use case?

Or is it just enough to use AI? We have organizations who are translating R &D documentation for drug discovery and submitting that to the local regulators, just purely translated by AI with the appropriate guardrails. We have organizations that are translating plain maintenance records and using them as the source of truth. So there is a lot of potential in using AI, but you have to think a little bit out of the box and really forget the old ways of doing things. And the same holds for agentic AI, and I think even more so, because the potential of AI is even bigger. So it’s… It’s both for the public sector and for businesses. It’s a big redesign of how work gets really done.

And the bigger the organization, the obviously bigger the inertia that is out there. And public sectors tends to be the largest organizations in any country. So the challenge is even bigger there.

Thanks. And so, Matteo, as in the Computing Center of Barcelona, it is quite specialized also in applications for public sector, for health care. So what do you see as the main applications that are being developed on the basis of demand?

teach to the young people to understand the problem to propose solution. Thank you.

So, Roberto, you heard the challenges in terms of acceptance and implementation in the public sector, which were sometimes maybe the skills are not very strong. So how do you think that policy can really help to enable this transformation?

I think policy needs to be tuned with the transformation. So in a way, as I was trying to say also before, if you invent a digital bureaucracy, it’s a bureaucracy. It’s digital, but still it’s a bureaucracy. And you have then a bureaucrat and a digital and an AI agent bureaucrat. I mean, It would be very simple for the geniuses in this panel to produce an AI bureaucrat. And I’m sure AI can do bureaucracy even better than us, much better than us. Regulation -generating bots, yes. That would be super useful. Or regulation -correcting bots, that would be good. So you see, I think we need to be also from the legislation side disruptors and look at things with completely different eyes.

And for this, let me say that there are one thing that really does a striking similarity between Europe and India is this idea of believing in the public stock. So the idea that you can actually be managing your identity, your attribution. your capacity to sign, to timestamp, to actually exchange these attributes in an open source and an open model. This is for people and for businesses. And then in this way, the state is in your hand. I mean, it is actually, you have the bureaucracy under control because the bureaucracy, it’s you. So, if you actually reverse the logic of the citizen going to an office, that’s what you refer to, to the office going to the citizen with, of course, all sorts of nice agents, push notifications, attestations, then, I mean, you re -engineer the state.

So, my point is, if we dare, and I dare to say in India you are daring, in Europe we are daring, you can actually redesign the paradigm and then if you do that then creativity is really at work because there can be many different agents many different ideas on how to improve processes Thanks

Now we only have very little time to go but before leaving since we have these four geniuses I would like to ask you maybe a very last thought that you have on innovation in the public sector and how you can contribute

I think public research is very important in particular I think partnerships between private companies and public efforts is actually something that works because doing research takes some infrastructure infrastructure takes some capital and so I think that’s the way we can actually accelerate together

I would say that the the the the the AI is a dual use technology and we need to look for the good use in this direction I think we can do a lot in Europe because as Roberto said we have the infrastructure and then we need a little more to invest a little more and then to define common projects because don’t forget that if you look at the power at the national level between the state and China they have more than 80 % of the computer more than 80 % of the people and more than 80 % of the investment so what we can do one possibility should be being in India alliance I think it would be very good to have an alliance between Europe and India in this topic we as BSE we have alliance with the SIDAC with the super competing centre and we have with the Institute of Science in Bengaluru.

And also financed by the Commission, we have a very good project that we are very happy to collaborate with you in the end. Thank you.

Now, time is up, but I would like to know very shortly, last thought from Roberto and Jarek.

Yeah, I think we can build and we will build from the commercial side, from the business side, amazing products that are driving a lot of value creation in the AI space. I think that that’s clear. And we’re going to be trying to do that in a way, of course, that our users and our customers can be really delighted by those products. But I think there is a lot of work that the public sector can do in terms of bringing this importance of adopting this technology into the broad base of population. I think both the German and also the European countries are going to be very happy and also from all of the conversations that we had here, the European and Indian governments do understand that, but we should not underestimate this challenge.

And I think there needs to be a very strong partnership between… businesses and the public sector on driving that. Thanks.

Thank you.

I am one of the few that has been in the three summits. I mean, Blanchley, this one, and last year in Paris. And of course the size already gives you an idea how things have changed. In the kind of discussion room at Blanchley Park we were 20, including the leaders. I mean, that gives you an idea. Now, the point is, I’m so happy to be here because what I always thought a little bit is true. There’s not one future for AI and technology. And it is not written. It is not written. The thousands and thousands of people that participated to the summit this year will write the future. So those that tell you there’s only one way, I mean, there’s only one scale.

And the rest of the world should watch and applaud. and I mean adapt to it absolutely I mean this summit shows and application of AI in public service what India is doing, what Europe is trying to do shows there are many futures and as I was trying to say before the future is in our hand

Thanks a lot and with these very intelligent and smart sentences tell me to thank the speakers and thanks a lot for your participation Thank you Thank you Thank you