Building Indias Digital and Industrial Future with AI
20 Feb 2026 11:00h - 12:00h
Building Indias Digital and Industrial Future with AI
Summary
The panel opened by GSMA highlighted the convergence of AI, telecom and data sovereignty around digital public infrastructure, emphasizing the need for intelligent, programmable networks as core national assets [1][8-10]. Julian Gorman argued that mobile networks are shifting from mere connectivity to trusted layers that shape AI model performance, enable real-time services such as fraud prevention and digital identity, and therefore must be integral to future digital infrastructure [14-16][18-20]. He further stressed that digital sovereignty now requires strategic control over standards and intelligence, not just data localisation, and that global interoperable standards are essential to avoid fragmentation and keep countries connected to the global digital economy [21-24][25-27].
Debashish reinforced this by noting that networks have evolved from voice to intelligent platforms that embed AI for identity verification, fraud mitigation and sovereign data decision-making, and that ensuring trust, interoperability and global compatibility is a key challenge [37-41]. Rahul Vatts illustrated India’s experience, citing the processing of 28 lakh crore rupees through UPI on a billion users supported by a massive connectivity layer of over a million BTSs and 500 lakh km of fiber, and described how OTP/SMS, Aadhaar-enabled payments and telecom-derived credit scores create trust for billions of transactions [51-58][61-62][65-67][69]. A telecom-service-provider representative added that the DPI ecosystem adds contextual enrichment to raw data, offering open APIs that allow banks and other entities to access real-time fraud and authentication signals, thereby turning the network into a governance and resilience layer [82-88][100-108][110-112].
Deepak Maheshwari outlined three tiers of data sovereignty – physical/administrative control for critical state data, citizen-controlled data that may flow internationally, and business data that balances India’s outsourcing role with the need for two-way data flows – and called for active participation in international standard-setting rather than unilateral control [141-148][155-162][164-170][172-179]. Mansi Kedia warned that building public digital infrastructure and private capabilities in silos creates inefficiencies, stifles innovation and weakens trust, and advocated for flexible blueprints and standards that can be adapted across contexts while preserving interoperability [204-209][218-224][226-232]. Rahul then broke down practical sovereignty into four slices – data residency, control-plane localisation, operational control of network software, and jurisdictional exposure to foreign laws – and described Airtel’s sovereign-cloud offering that keeps data within its own network to meet these criteria [235-242][245-252][255-262].
Martin (representing Vodafone Idea) highlighted emerging regulatory frictions such as the need for AI explainability, accountability and clear standards for digital intermediaries, suggesting that industry-wide playbooks and reference frameworks are required to balance security with innovation [279-286][288-295][300-307][310-314]. Deepak emphasized that India’s DPI model is open, royalty-free and supported by diplomatic and research institutions, making it a scalable, adaptable framework for Global South countries without imposing monetisation barriers [318-326][330-338][340-347]. Mansi concluded that India’s extensive DPI experience, from UPI to mobile-data-driven credit scoring, provides concrete evidence for other emerging economies and that ongoing collaboration with multilateral bodies will help evolve standards and blueprints for inclusive digital development [353-360][363-368].
The discussion ended with consensus that achieving AI-enabled, sovereign yet interoperable digital infrastructure will depend on open standards, contextual data enrichment, and coordinated policy frameworks that align national priorities with global innovation ecosystems [22-24][218-224][279-286].
Keypoints
Major discussion points
– Telecom networks are evolving from simple connectivity providers to intelligent, AI-enabled platforms that underpin Digital Public Infrastructure (DPI).
Julian highlighted that “today’s mobile networks are becoming intelligent, programmable and trusted layers of the national infrastructure” and are now part of AI model decision-making, fraud prevention and digital identity security [13-18]. Debashish reinforced this evolution, noting that networks are “no longer passive carriers of data” but “intelligent platforms where AI is deployed” [37-40]. Rahul gave concrete examples, describing how OTP/SMS, Aadhaar-enabled payments and real-time fraud indicators rely on the connectivity layer [51-63].
– Digital sovereignty now means more than data localisation; it requires strategic control over standards, AI models, and the underlying infrastructure.
Julian framed sovereignty as “strategic control over the infrastructure… the standards, and increasingly, the intelligence that underpins the national digital system” [18-22]. Deepak expanded the concept, distinguishing physical data location, local contextual relevance, and the need for participation in global standard-making rather than unilateral control [140-176]. Rahul added practical dimensions, separating data residency, control-plane localisation, operational sovereignty (software patches) and jurisdictional sovereignty (e.g., US CLOUD Act) [235-242].
– Avoiding fragmented or parallel digital infrastructures requires interoperable open standards, APIs and collaborative blueprints.
Julian warned that “fragmentation… slows down” and that “open APIs, harmonised frameworks” are essential for scaling [23-25]. Martin’s question about “parallel digital infrastructure structures” and the risk of duplication underscored this concern [76-78]. Speaker 1 described how TSPs collaborate via open APIs (e.g., FRI, Digital Intelligence Platform) to provide contextual data to banks and other entities [118-124][127-130]. Mansi emphasized that “systems coming together build trust, efficiency and innovation” and that the World Bank favours flexible blueprints that incorporate best-practice standards [204-210][220-226].
– Concrete use-cases illustrate how AI-enhanced networks deliver trust, fraud mitigation and financial inclusion.
Rahul detailed Airtel’s role in securing UPI transactions, OTP delivery within 2 ms, and AI-driven spam/ scam detection that adds friction to fraudulent calls [51-63][66-70]. Speaker 1 explained the enrichment of raw call data with contextual signals (e.g., Aadhaar verification vs. call location) to enable real-time fraud decisions [102-110]. These examples show the network acting as a “contributor to governance, resilience and trust” [17].
– India’s open, scalable DPI model is positioned as a template for the Global South, combining open protocols with diplomatic outreach.
Deepak argued that India offers an “open protocol” DPI framework without licensing fees, enabling other countries to adopt and adapt it [318-327]. He also highlighted diplomatic mechanisms (e.g., Ministry of External Affairs think-tanks) that support capacity-building [332-334]. Mansi noted that India’s experience with UPI, NPCI and the “Finternet” concept provides evidence for other emerging economies and that multilateral collaborations are already shaping global adoption [353-360].
Overall purpose / goal of the discussion
The session aimed to move beyond high-level rhetoric about AI, telecom and digital sovereignty and to “translate them into direction… identify practical next steps… create space for collaboration” (Julian [28-31]). Participants sought to share India’s DPI experience, surface policy and technical challenges, and outline how global standards and open blueprints can enable inclusive, secure, and interoperable digital ecosystems for both advanced and developing economies.
Overall tone and its evolution
– The conversation opened with a formal, visionary tone-Julian’s welcoming remarks framed the topic as a strategic crossroads [5-9].
– It then shifted to a technical and evidential tone, with detailed statistics, product descriptions, and operational insights from Rahul, the TSP speaker, and Deepak [51-63][95-108][140-150].
– Mid-session, the tone became collaborative and policy-focused, emphasizing standards, avoidance of fragmentation, and the need for multistakeholder governance [23-25][76-78][204-210].
– Towards the end, the tone turned optimistic and diplomatic, highlighting India’s role as a model for the Global South and the potential for international cooperation [318-327][353-360].
– The closing remarks returned to a courteous, appreciative tone, thanking speakers and the audience [377].
Overall, the discussion maintained a constructive and forward-looking atmosphere, moving from high-level vision to concrete examples, then to policy coordination, and finally to global outreach.
Speakers
– Rahul Vatts – Chief Regulatory Officer, Airtel; expert on telecom infrastructure, digital payments, fraud mitigation, and AI-enabled services[S1]
– Speaker 1 – Unspecified role (appears to be a telecom service-provider/TSP representative discussing DPI context, enrichment, and open APIs)[S2]
– Julian Gorman – Head of APAC, GSMA; specialist in intelligent networks, digital public infrastructure and global standards[S5][S6]
– Debashish Chakraborty – Moderator, GSMA representative; focuses on convergence of AI, telecom and data sovereignty[S7]
– Deepak Maheshwari – Representative, Center for Social and Economic Progress (CSCP); expertise in data sovereignty, AI policy, and digital public infrastructure[S8][S9][S10]
– Mansi Kedia – Representative, World Bank; works on digital public infrastructure blueprints, standards and development policy[S11]
– Audience – Various participants (e.g., Vijay Agarwal – jewelry manufacturer interested in KYC-embedded wearables; Yuv from Senegal; Professor Charu, Indian Institute of Public Administration; Dr. Nazar)[S12][S13][S14]
Additional speakers:
– Martin (Martin Schroeter) – Representative of Vodafone Idea; contributes perspectives on regulatory challenges, AI-driven networks and open-gateway APIs[S2]
– Matan – Speaker referenced in the discussion on DPI and data context; specific role or affiliation not provided in the transcript or sources.
– Vijay Agarwal – Audience member, jewelry manufacturer proposing KYC-embedded ring concept; raised a question on data embassies.
– Mike – Mentioned briefly in the audience Q&A; no further details available.
– Ambika – Named in the moderator’s prompt but did not speak; no role or contribution recorded.
The session opened with Debashish Chakraborty introducing the theme of “convergence of AI, telecom and data sovereignty all weaved around the digital public infrastructure” and handing over to Julian Gorman, head of APAC GSMA, for the keynote address [1-4]. Julian welcomed the participants and positioned GSMA as the global body that unites the mobile economy to “unlock the power of connectivity so industry and society thrive” [5-7]. He explained that today’s mobile networks are evolving from simple connectivity providers into intelligent, programmable layers that shape AI model performance, enable real-time services and embed governance functions, thereby redefining digital sovereignty as strategic control over infrastructure, standards and the underlying intelligence [13-22]. Julian concluded by stating the session’s purpose: to translate these themes into concrete directions, practical next steps and collaborative space [28-31], before handing back to Debashish [32].
Debashish reinforced Julian’s vision, observing that telecom networks have moved from voice-only services to “the trusted digital infrastructure that we use today underpinning the modern economies” and are now “intelligent platforms where AI is deployed either as an add-on or embedded already into the network” [37-40]. He highlighted that this shift enables digital-identity authentication, fraud mitigation and the exercise of data sovereignty [41-42].
Rahul Vatts, Chief Regulatory Officer of Airtel, illustrated the scale of India’s Digital Public Infrastructure (DPI). In a single month the UPI system processed 28 lakh crore rupees across a billion users, supported by a connectivity layer of over a million base-transceiver stations and 5 million km of fibre [51-54]. He explained that every transaction relies on the network for OTP/SMS delivery within two milliseconds, providing the “layer of trust” essential for Aadhaar-enabled payments and other citizen-centric services [55-63]. Rahul also described Airtel’s AI-driven anti-spam and fraud solutions that add friction to suspicious calls, thereby reinforcing trust in the ecosystem [66-70]. He noted that quantum-secure techniques are already being explored for Aadhaar, signalling early work on next-generation data protection [??].
Debashish then asked Martin (Vodafone Idea) how to ensure that the DPI trust layers being built by mobile network operators complement rather than duplicate the GSMA Open Gateway APIs [76-78]. Martin (Speaker 1) answered that the TSP community already collaborates with the GSMA team on Open Gateway APIs, several of which have now been certified [98-108][110-112][118-124]. He added that TSPs expose contextual enrichment-such as Aadhaar verification versus call-location data-through open APIs like the Fraud-Risk-Indicator (FRI) and the Digital Intelligence Platform, enabling banks and other institutions to generate risk scores in milliseconds for micro-loans [102-110][124-126].
Deepak Maheshwari expanded the discussion on digital sovereignty. He argued that sovereignty in an AI-driven world is not only about data residency but also about “having strategic control over the infrastructure” and the standards that govern it [18-22]. He distinguished three tiers of sovereignty: (i) physical/administrative control for critical state data, (ii) citizen-controlled data that may flow internationally, and (iii) business data that must balance India’s outsourcing role with two-way data flows [140-162]. Deepak warned against “walls” that block two-way traffic and advocated active participation in multistakeholder standard-setting bodies (GSMA, ISO, ITU, IEEE) rather than unilateral control [173-179]. He concluded by invoking the historic “3C” framework-carriage, content and conduct-as a lens to view today’s AI-driven DPI [??].
Mansi Kedia (World Bank) distinguished between “standards” (prescriptive, commercialised) and “blueprints” (flexible, adaptable best-practice collections), arguing that the latter better suit emerging economies while still ensuring interoperability [219-226][227-233]. She referenced the “Finternet” concept being explored with the BIS as evidence of scalable, interoperable digital rails [353-360][363-368].
Rahul returned to the sovereignty theme with a “four-slice” model-data residency, control-plane localisation, operational sovereignty (software patches) and jurisdictional sovereignty (e.g., exposure to the US CLOUD Act)-asserting that true sovereignty requires all four dimensions [235-250]. He advocated a “selective residency” approach: keep critical public-interest data (KYC, health, defence) under domestic control while leveraging hyperscaler efficiencies for non-critical workloads [245-252][255-262].
Martin (Vodafone Idea) raised regulatory frictions, calling for industry-wide referenceable standards or playbooks that would make AI decisions explainable and ensure compliance with emerging digital-intermediary rules [279-286][288-295][300-307][310-314]. He warned that private sovereign-cloud offerings could create parallel DPI layers, fragmenting the ecosystem [239-244][262-270].
Deepak stressed that India’s DPI model is an “open protocol” with no royalty or IP fees, allowing other countries to adopt and adapt it freely [318-327]. He noted that diplomatic channels-such as the Ministry of External Affairs think-tank and the Indian Council of World Affairs-provide soft-power support for capacity-building abroad [332-334].
An audience member, Vijay Agarwal, proposed a novel “ring-based” KYC and medical data vault that would store personal data on the body, encrypt it, and use blockchain-based consent, effectively creating a personal data-embassy [371-374]. Deepak responded affirmatively to the idea of reciprocal data-embassy arrangements [375], while Rahul pointed out that Aadhaar already incorporates masking and that data-embassy concepts are being explored at the governmental level [376-380].
After the Q&A, Debashish recalled that, until about 15 years ago, the IRCTC registration form required a compulsory marital-status field that offered no clear benefit, underscoring how data-collection norms have evolved and informing current debates on data sovereignty [??].
The panel broadly agreed that (i) telecom networks are now AI-enabled public layers delivering trust, fraud mitigation and digital-identity services; (ii) digital sovereignty must extend beyond data localisation to include control-plane, operational and jurisdictional dimensions; (iii) open, interoperable standards or flexible blueprints are essential to avoid duplicated DPI stacks; and (iv) India’s open, scalable DPI model offers a credible template for the Global South [13-22][23-25][318-327][355-362]. Diverging views emerged around the preferred normative instrument (hard global standards versus adaptable blueprints) [22-25][219-226] and the relative priority of AI explainability versus technical-jurisdictional aspects of sovereignty [279-286][236-250].
Key take-aways included: (a) continued development of open APIs that enrich raw network data with contextual signals; (b) regulatory frameworks that address AI explainability, digital-intermediary responsibilities and jurisdictional exposure; (c) promotion of India’s open DPI protocol through diplomatic and multilateral channels; (d) exploration of data-embassy and personal-device data-vault concepts; and (e) creation of sector-specific playbooks that balance security with innovation. Unresolved issues highlighted were the concrete design of jurisdiction-safe sovereign clouds, mechanisms to prevent parallel DPI infrastructures across operators, and detailed governance models for data-embassy arrangements [285-293][255-262][371-380].
In closing, Debashish thanked all speakers and the audience, underscoring the collaborative spirit of the session and the shared commitment to building AI-enabled, sovereign yet interoperable digital public infrastructure [377].
convergence of AI, telecom, and data sovereignty all weaved around the digital public infrastructure. I’m Devish. I represent GSMA. I’ll request Julian Gorman, head of APAC GSMA, to give his keynote address and then we start with the panel discussion. Julian.
Good morning, everyone. Warm welcome, distinguished guests, colleagues and partners and speakers who have joined us today. It’s a great honour to actually open this session for GSMA. GSMA, for those who don’t know, is the global organisation uniting the mobile economy, that means mobile operators and the ecosystem, to unlock the power of connectivity so industry and society thrive. And this session really goes to the core of that around intelligent networks, intelligent telecom networks for digital public infrastructure, a topic that sits right at the intersection of where the telecom industry is heading and where national digital public infrastructure is heading. And that’s where we’re being built. Of course. India is really at a pivotal point in its digital journey and a key player in this space.
They’ve been on the digital public infrastructure journey for a lot longer than the rest of us, but over the last decade, we’ve really seen the rise of digital public infrastructure recognised from identity and payments to digital commerce and data empowerment and has shown the world what is possible when scale, innovation and public purpose come together as delivered inclusion, trust, economic impact at a level few countries have achieved. But as we enter this next phase, which is shaped by AI, real -time data and increasingly autonomous systems, we need to ask a fundamental question, and that is what are the role the telecom networks play in this new digital infrastructure? For years, networks were viewed simply as connectivity providers and that view is changing.
Today’s mobile networks are becoming intelligent, programmable and trusted layers of the national infrastructure. and they’re shaping how AI models perform and will perform and how services are optimised at the edge, how fraud is stopped before it happens and how digital identity remains secure in a world of growing complexity. In India, networks already support core DPI functions, identity verification, payments, emergency response and major public service platforms. As AI becomes embedded in these systems, the networks don’t sit back anymore in the background it becomes part of the decision -making fabric providing context and priority for tokens or the critical elements of data which digital public infrastructure information is the predecessor of. Through this, the network becomes a contributor to governance, resilience and trust.
And that brings us to the second major theme of the day, digital sovereignty. In an AI -driven world, sovereignty is no longer just about where the data is stored, it’s about having strategic control over the infrastructure. The key to this is the ability to manage the infrastructure the standards, and increasingly, the intelligence that underpins the national digital system. Countries want to know, how do we build AI -enabled public infrastructure that is safe, interoperable, and aligned with national priorities, while still remaining connected and interoperable with global markets and innovation? This is exactly where global standards matter. Fragmentation, whether technical, regulatory, or geopolitical, slows down. Interoperability, open APIs, harmonized frameworks, help countries scale confidently, while staying part of the global digital economy.
India is uniquely positioned to show how this balance can be achieved. Open, yet sovereign. Scalable, yet secure. National in ambition, but global in design. And our goal today is not just to talk about these themes, it is to translate them into direction. To identify practical next steps. To create space for collaboration. and to learn from India’s experience in ways that matter for economies that are at every stage of digital development. So I’m looking forward to the discussion and to the concrete actions we can shape together and I look forward to very big contributions from the panel today and also to hear more from the audience later. So thank you. Debashish, I hand over to you.
Thank you, Julian. Thanks for the opening remarks. Am I audible? Looks like yes. So let’s begin. We have a fantastic panel here of experts. So let’s start with this discussion. So what we have seen over the past few decades that telecom networks have evolved. They have evolved a lot from just enabling voice to powering mobile broadband to becoming the trusted digital infrastructure that we use today underpinning the modern economies, right? So today’s network are no longer passive carriers of data. They are becoming intelligent platforms where AI is deployed either as an add -on or embedded already into the network, where digital identity is authenticated, where fraud is mitigated, where sovereignty over data and decision -making is increasingly exercised.
As India advances in digital public infrastructure and its AI ambitions, the key is how we ensure these systems remain trusted, interoperable, and globally compatible while avoiding fragmentation and duplication. And that is the conversation which we aim to explore today. Let me start with Rahul, who is the Chief Regulatory Officer for Airtel. Rahul, we often talk about digital public infrastructure as applications and platforms, but at the foundation sits the network which you drive. So from Airtel’s perspective, what makes the telecom networks uniquely positioned in the digital world? It is as India’s trusted infrastructure layer. beyond just connectivity.
Thank you, Devashish and GSMA for this particular session. It’s a session of particular interest to me as a user in the digital ecosystem and of course to the entire digital fraternity because if there’s one thing which India is doing great, it’s really the digital public infrastructure to the extent that President Marcon yesterday actually mentioned about it. It’s the biggest export which India has done across the globe. So let’s talk about what’s really happening today. If you look at the data of January alone, India transacted 28 lakh crores rupees of money through its UPI infrastructure. It was spread across a billion people and all this is happening on what? On what is the foundation layer? It is the foundation layer, is the connectivity layer.
and so for us at Airtel this is not just a plumbing job it’s the very heart of the foundation we are laying for trust how are people transacting this much amount of money because they trust the ecosystem to which they really want to do this and so beneath this layer is really the connectivity which has powered the country look at the numbers of connectivity in a country like ours we have got more than a million BTSs powering the entire country we have got more than 500 lakh kilometers of fiber running across in various shapes and forms across the country we have got as an industry more than a thousand edge and large hyperscale data centers now can you imagine each of the mobile switching center carries a load of at least 30 to 50 million people sometimes or even larger at times so this is the scale with which the infrastructure is becoming the layer with which we are operating what is all this enabling let’s look at that that.
What it is enabling is every transaction you do, there is a OTP or SMS which is coming out, right? So this OTP and this SMS is what? It’s a layer of trust that people are trusting the message which they are trying to get on their system. Let’s look at the Aadhaar enabled payment system. More than 500 million rupees done on that alone. And how is that enabled? Through a connectivity which is happening in less than 2 milliseconds. So this again is an example of that same ecosystem. Let’s go further. What’s really happening and how are we doing? I don’t know how many of you actually visited the Airtel stall. We have got solutions where banks can use the telco indicators to make a smart choice about giving you loans, right?
We rank a person’s history based on a low risk or a high risk which enables the bank to be able to take a smart decision in a matter of milliseconds. Remember, in India, it’s not the large loans that matter. A lot of loans which are happening in the ecosystem are less than 200 lakh rupees, right? Just 2 lakh rupees or below are also a large amount of loans which happen. there is a financial risk fraud indicator which the department has created banks can dip into that risk indicator and also get a score out of that to say okay what is it that we are really you know trying to get out of this all this is what the layer is let’s look at what vs telcos are doing vs telcos are giving you trust to say that the call you are giving call you’re receiving is spam free or not right we have got a at least three products launch over last one year we first launch our you know solution which warned you about a suspected spam right then we went ahead we started blocking fraudulent links you know basis the large database we created with you know global players like google and open fish and mavener at the third stage we just launched around two weeks back a very powerful product you know one of the reasons for spam is urgency that i’m calling you please share your otp urgently right and to remove that now we have created a friction you know one of the reasons for spam is urgency that i’m calling you please share your otp urgently right if you are on a call you get a flash message saying please be careful you are on a call you’re receiving otp this may be spam so it creates a friction for those 30 seconds to say do you want to really do this or not all this is what this is uh reinforcement of the trust we want to create in the ecosystem let me go a little larger uh we are operating in large countries uh uh you know across the globe and one of the things we have been doing wonderfully well in africa is to really take the digital public infrastructure blueprints from india and take them to africa uh so it’s all about identity it’s about payments you know it’s about how they are able to transact and we have got a solution called dpi inbox right which we are in conversation with a lot of african leaders to be able to transplant the india stack onto the african ecosystem and how do we do that we are giving a bundle of hardware and a software we are giving a very air -capped cloud you to do that and we are creating the entire ecosystem for them so that they are able to implement a digital public infrastructure stack in their countries.
So really, Devashish, it’s about trust which we try to create with infrastructure layer but we get smart and make people’s life easier and customers’ life easier is what we are
Thanks, Rahul. Those were very key messages which you gave in which the network is being used for citizen -centric services and that’s how the network has evolved the last few years. Coming to you, Martin. Martin represents Vodafone Idea. You heard Rahul speaking about how the network is being used for various citizen -centric services for fraud mitigation, for taking care of spams. A lot is being done by the mobile network operators, right? But my question here is there’s also a growing discussion globally today about avoiding parallel digital infrastructure. structures. India is building new DPI trust layers for authentication and fraud prevention. How do we ensure that the efforts which the MNOs, the mobile network operators are making adding layers, how do we ensure that there is no that these complement and not duplicate the operator -led capabilities like Open Gateway APIs that GSMA has?
So, in fact, I was part of one of the entity which set up and contributes to the largest DPI infrastructure today. I used to earlier be associated with the NPCI and then moved on here to Telco for the past five years now. So, the overall DPI infra, if I were to go by, I would want to answer this by bringing in four key words that I want to associate myself with in this. One, context and enrichment. And the second thing that I wanted to touch upon is serviceability and purpose. So when the entire DPI infrastructure evolved for the country, it evolved with two core purposes to be addressed with, right? So we were wanting to take the entire digital infrastructure to reach the last mile civilian.
We also had the objective of financial inclusion to be driven by the country. So the DPI framework was created to meet these two core objectives. The role of a TSP in this, by and large, was to ensure that the goal of digital India and financial inclusion landed up reaching the masses. That’s the role that TSP played. And with every net new tech evolution that has happened, there are various things that come in. Fraud evolved, so because banking happened in doorstep. fraud also started happening in doorstep. You don’t have to go and loot a bank today. You can loot thousands of individuals in the most easiest manner and fraud evolved that way. So in each of these contexts, while we realize the Digital India vision and the financial inclusion for the country as a whole, the DPI networks played a role, TSP’s played a role to ensure that these realizations come in handy.
Now, Rahul briefly touched upon a few of them. We are limited TSPs in the country, three, four of us comprehensively, who work in conjunction context. Amongst us three, we land up working together. So I still remember those days when, as from my previous entity going to TRAI, asking them to land up sitting up, how do I find out fraudulent mobile numbers yearly? Today, we look at it as FRI, which is exposed by, the DOT themselves today to multiple other financial institutes, which can go and look up into and then take a decision. decision. There is something called as digital intelligence platform, which again, amongst all three collaborated TSP data, which is converged and provided by the DOT themselves to rest of the financial institutions to look into.
Now, all of these, I will bring back my word around context, right? So these are information that multiple of us as TSPs are able to provide, provide, collate and make it available. Who can consume? Any of these providers, because fraud is not happening to me as TSP. For me, if there is a call that is connected between person A and a person B, it’s revenue to me. But for a bank, while in call, something else is going, that is a context. And this context is something that you can provide back to enrich the data. And with this enriched data, making a decision making for what do they want to. I see an Aadhar, verification happening live from a location called A.
while at the same time there is a call happening showing that the presence of the person is in B, it does not matter to a telco because for me both are actually revenue. But for an authentication entity versus an entity which is approving a financial transaction, they may consider them as a fraud. So the context and enrichment of the context associated with the data, TSP today has the ability to provide a large amount of context -driven information to these individual players whereby they can consume them for their own utilization and make active decisions. So that’s the way that I would want to try and comment. One good part is at least all three of us, four of us are operated in converged platform.
We have done the experience with DLT that we set up during the earlier days of spams. Now spams were those days only. The unwanted telemarketers messages that were coming, it has evolved to spam. Spam has become scam. So now we are working towards how do we overcome scam beyond scam, whatever comes. Now there are digital errors, humongous money being lost. So as TSPs, we work in conjunction, put them in order, collaborate with the likes of COI and DOT to set up infrastructure as open APIs and then allow these APIs as interfaceable for institutions who would want to take decisions appropriately. Rahul touched upon digital lending, right? So there is not only, if you look at countries serviced today by more than 1100 member banks across the country.
We might be knowing as sitting in metros, we might be remembering only few banks, but to service such a large nation, we have 1100 member banks. Imagine these guys don’t have to always go back to civil only and provide a lending. You may want to relate back by postpaid consumers, the quantum of money that they pay frequently, etc. It’s an inclusive decision. Those are open APIs we are able to set up. And India is. We have been forefront to set it up and we have operated it way too well already. is what I would want.
By the way, your team is also working extensively with the GSM team on the GSM OpenGate APIs. Many of them have been even certified now. I can tell you that. Thanks for that context in which you are talking about contextualization of data. That’s again a unique perspective that you’re talking about. Moving on to Deepak Maheshwari. Deepak represents CSCP, Center for Social and Economic Progress. Deepak, you have been attending and speaking in this conference for the last couple of days. Data sovereignty, I’m sure, is a term which you would have encountered several times. I want to ask you this, Deepak. How should India define data sovereignty in an AI -driven DPI era beyond just data localization and control?
But how should India define data sovereignty without control over standards, decision -making systems, and long -term strategy? strategic autonomy.
Thank you, GSMF, for having us here. When we are looking at this whole issue about digital sovereignty, data localization, etc., and data localization itself, we could look at it in different ways. For example, it could be about just the physical location of the data. That’s one. That’s a pretty obvious one. The second is also about data context, as Matan was just mentioning, in terms of what is the local context. So, for example, a lot of people think about data localization only in terms of local languages. But suppose you are seeing weather, and it shows you weather in Hindi here in Delhi, but of New York, probably it might not be that useful. So you also need local context.
And then beyond all these things, awfully what is happening is, and again, this is not such a new concept of sovereignty as such. So people have been talking about sovereignty. It’s been around for a fairly long time. Of course, the terminal of sovereignty is the fact that it’s not just about the data the lexicon has evolved but this whole notion has awfully become much more important for example even in India we had the digital, when we looked at the previous versions of the data production law if you look at the previous reports which never become the policy which is the non -personal data framework again in all those things we had this notion that India’s data should remain in India.
Another thing I mean in February 2019, 7 years back we had something called draft e -commerce policy. Now the tagline of that however was India’s data for India’s development. It was not about commerce. It was more about data. So from that perspective when we look at today and even when I was member of the METI’s committee in 2018 when first time the government set up a committee on AI, again this whole thing came up that okay what about data here. Now this is something that we need to look at in three different ways. One is yes Yes, there is some sort of data which India should have within its own physical as well as administrative control. So obviously things related to defence, national finances, etc., you would like to do that.
Second is as far as citizens’ data is concerned, some of that data, yes, so UIDI, voter database, etc., obviously that type of thing, yes. But there is other type of data for which citizens themselves may like to exercise their choice and may like to exercise their own agency in terms of using that data not only in India but also outside India. For example, if I apply for a visa to another country, I will have to provide my data to that country. So there is no way that it can happen without that. And then the third thing is in terms of business aspect when we look at it. Now in terms of businesses, on one hand we are seeing in India, and we are very proud of it, that for the past, three decades, we have emerged as a global outsourcing hub.
are the global hub for data coming from all over the world and which is being processed here. But at the same time, if we try to create these walls around us, that okay, India’s data cannot go outside, but we expect that outside data should continue to come in. I think there’s a challenge in that. There’s a dilemma in that. There’s a dacotomy. Because these are walls. If we create these walls, and these are not walls, because in fluid dynamics, if we go back to our school physics, the walls are something that do allow one -way traffic, not two -way traffic. But walls are two -way isolations. So that’s another thing that we should keep in mind.
So when we’re talking about digital sovereignty within the context of AI, yes, obviously, there are things that we do want to have here and we should continue to do that. But there are also things where we do need more collaboration. So for example, one of the terms that he used was about control. a school, and you’re talking about a school, and you’re talking about I would like to control the standards so much as contribute to those standards. So, for example, whether it is GSMA or CGPP, ISO, ITU, IEEE, et cetera, I mean, so many other standard organizations, whether they are plurilateral, whether they are multistakeholder in whichever form, they all have certain mechanisms of people and countries to participate in that decision making.
So rather than controlling that standard, the effort should be, the endeavor should be about contributing to that standard making as a participant, as a contributor, and then evolve it. Obviously, when you are contributing and you are collaborating, you won’t have everything your own way. There will always be inevitably some give and take because sovereignty by itself in a globalized world has a challenge because the moment we talk about any international organization, we are talking about international organizations. whether it is UN, whether it is WTO, whether it is ITU, whether it is an organization like GSMA, if we want to work there, we’ll have to give up something to get something. The important thing is how do we create an institutional mechanism that we have, are in a position that whatever we are giving, we believe that we are getting more than that.
So there should be some sort of incentives around that. And the last thing that I want to mention is that, yes, often we have been talking about that India’s digital public infrastructure itself is a massive digitalization which is happening, but actually it is not so new. It’s more than one and a half centuries old. Because the original telecom networks that came was in the telegraph era, and that was also in dots and dashes. So it was a binary world even at that time. And people may or may not believe it, but India got its first sub – cable in the same year that the US got. And that was in 1858. Just four years after the first submarine cable came up first time between UK and France.
India got its first law Vivek in 1854. The first Indian Telegraph Act came in 1854. I have written a lot about this in this report. I mean it is available online at CESAC website if people are interested. Using a 3C framework. So carriage, content and conduct. But what is more important is in this world of AI is not just the carriage which is of course fundamental as I mentioned. Carriage is fundamental because without that you just won’t be able to do anything. Content, what’s going through it. But more importantly in terms of
Beautiful insights. Thanks for taking us back to the concept of walls and walls. I like to come to Mansi now. Mansi sitting here is representing the World Bank. Manasi, from World Bank’s experience, we are talking about standards and we are talking about the DPI era. What are the risks you see when public digital infrastructure and private digital capabilities, Matan spoke about it briefly, when these two, the public digital infrastructure and private digital capabilities are built in silos, and why are global standards essential in accelerating inclusive digital outcomes?
and Raul spoke about a lot. So systems coming together help build trust and therefore having independent systems means there are more points of, more vulnerability in the system. So systems come together to build trust. Systems have to come together for efficiency. I think that’s the biggest economic argument against a lot of things that you were saying about why banks are coming together, why is data coming together. So that is the, efficiency is the other thing. And the third which was mentioned but again not articulated was innovation. So how mobile data is now becoming a source of data for lending. I mean why are we using that as understanding credit risk and fraud risk and not something else.
So there’s innovation happening on something that was never understood to be for that purpose. So systems that operate in silos, whether it is the public sector or the private sector. Close it. Sorry. Whether it’s the public sector. Maybe it’s off. Oh I didn’t have it on, I’m sorry. I have a loud voice, so I hope everyone was able to listen to me. So I think the risk of building systems in silos, whether it is the public sector or the private sector, is essentially missing out on efficiency capabilities, innovation capabilities, and building trusted ecosystems, which is actually nothing but the foundations of digital public infrastructure. You used standards. I think the World Bank works more towards the ideas of blueprints.
We have been doing a lot of work on trying to develop blueprints, which are slightly more flexible, adaptable, but bring together best practices from different countries and see how they can be made more adaptable to different contexts, something that Deepak sir was saying in his initial remarks, that you want to make systems that bring you the operational ideas and principles, but don’t necessarily require. They may be prescriptive in terms of how they need to do some. So when you have a standard, you know it’s prescriptive, and that’s how the networks are running. So for that, you need a standard. But when you’re building systems. I think the World Bank is approaching it more from a blueprint point of view.
So last year, the bank came up with a digital public infrastructure and development report where it articulated what it meant by digital public infrastructure. What are its principles? What are the objectives? What is DPI? What is not DPI? And I think that’s the way we are going to go ahead, even with AI, AI commons, building common infrastructure, to be able to determine the pathways for the future, which countries can adapt to in their ways. It need not necessarily become, I mean, I’m just trying to distinguish between standards and blueprints here, because standards then get into ideas of commercialization and, you know, there has to be a process around it and there’s a whole private sector play.
Here there’s a private sector play and a public sector play, but the idea is to work more on the approach than on a particular way of running something.
perspective back for data sovereignty. So I’d like to ask you as AI moves deeper into network operations, right, not just at the surface level, what does data sovereignty practically mean for an operator in terms of data storage and control, edge processing, cloud reliance, control of the AI models?
Yeah, thank you. I think one of the biggest misconceptions we all have today is, you know, what exactly is sovereignty? And a lot of people confuse to say that any hyperscale account, if it is housed in India, for example, or that country becomes a sovereign, you know, infrastructure. I think nothing can be away from growth than that statement. Why do I say that? I think if I have to define what is really sovereign for me, I will at least take three or four slices, you know, into it. first slice for me will be is the data residing in the country or not and the answer to that may be yes you know it may be residing in the country it’s not a big deal hyperscaler clouds do reside in the country the second indicator for me will be is there a digital sovereignty you know in that data and digital sovereignty for me will be is the control plane of that cloud within India or not in India right how are you really controlling that data and the cloud and the answer to that is not a single hyperscaler will have the control plane in this country that’s the fact the third indicator or a slice for me will be really about the operational sovereignty so you are saying that you want to upgrade the network you want to put a patch on the network right you want to put a software in the network where are you doing it from the fact is you are not doing it locally again most likely you are again doing it outside the fourth indicator for me and a very important one is the jurisdictional sovereignty right today under the US cloud act for example is it not true that if the US government so wants they can demand data now why should any other territorial power have a control on my data right so for me while the answer for data sovereignty may be it is locally residing but the fact is the control pane will not be in this country the fact is that we will not have even the patches coming up within this country and the fact is that we will be subject to jurisdictional controls so how are telcos you know getting aware about this only last week I read about DT you know Daoshi Telecom and they’ll just launch the sovereign cloud offering in Europe why did they launch and by the way six months ago Airtel launched its own sovereign cloud offering and the answer to us was very simple we were already managing data of nearly 500 million people and we were able to get a lot of data and we were able to get a lot of data in our network and we realized where is the data housed?
We said within our own networks. So we really have the capabilities to manage that complex data set. Then why is it that I cannot offer the same thing to my customers? And that’s where this whole, and that’s why telcos are having a renewed interest into getting into the sovereign situation. Why is it important? And let me be very selective about this. Do we need hyperscaler clouds in the country? I’m saying yes, we do need. Because if there are efficiencies of scale, if there are better products to be used, why not? But tell me, why should a KYC data of my customer be sitting outside with somebody? Why should the health record of citizens in this country be sitting outside this country?
Why should any critical data set which relates to defense or security agencies sitting outside this country? I think we have to get selective. We use the efficiencies of scale to the best party who is available to give that solution. But we should get selective. Get selective on what data? should reside and remain in control within this jurisdiction. I think that is an important part and that I think is a discussion we need to do. If I go to the market today, there are a lot of players selling Sovereign Cloud. But really, I mean, there is no sovereignty which is involved. But I think AI rests on data, right? And we cannot take the right decisions on data if we cannot really control it in the proper sense.
Hence, we require dynamism in our regulations and policies, but we also require sovereignty to be practiced in real sense for us to be able to do that. Airtel Cloud, which we made, we do around 140 crore transactions per second. That’s the bandwidth we have built. It was very interesting that day when the Prime Minister came to Airtel stall, he was asking, Rahul, what is the capacity of the thing you have created? And I told him, you tell me, sir, what is the capacity you want us to create, right? It’s really up to you. You have to guide us and say, we want to have these multiple use cases. Thank you. lining up the country and we are most happy to do that.
So I think we are in a very good place. We have got very robust infrastructure. And how do we now navigate this world of AI and provide a real opportunity and sense to our players within the ecosystem is what we are really looking forward to.
You reminded me from this conversation which we were having just a couple of days ago when someone was talking about data sovereignty and he said, it’s so utopian to talk about data sovereignty where if we slice and dice, then you realize where is the sovereignty. And you touched on that. Thanks for that point. Martin, I’ll come back to you. This was actually meant for Ambika, but you have to deal with this. So from Vodafone Ideas regulatory lens, what are the biggest policy frictions emerging as networks become AI -driven platforms? If you see any regulatory challenges, how can these be met with data sovereignty? slowing innovation?
So I’ll try and answer them in two perspectives. We heard our Honorable PM mentioning AI being responsible and reasonable. The word he used was reasonable in nature, multiple location, right? So it brings in, and there are multiple other contexts with reasonability that comes our way, one being explainability, another being accountability, and so on and so forth. So today, if you look at we as TSPs, TSPs are governed under the ambit of what we want to call ourselves with unified license, which is narrated by DOT. In some of these examples that we, with Rahul touched about, I touched about, and whatever World Bank team as well related back, we are able to see that our portfolio has expanded beyond the conventional TSP governed under the US.
license and today looking at the expanded approach that we are offering to market whether monetization not monetization thank god at least the data privacy is enacted now apparently i’m also the dpo for the firm so by virtue of which when we touched upon this area called data localization or what we would say is data sovereignty i think we largely misinterpreted is my personal view around that data privacy the dpdp at least clarifies that data collected has to be defined with a purpose we put in with a purpose now thankfully although i’m a tsp base is we falling under the ambit of a significant data fiduciary most likely we will be also governed by the data privacy laws of the country So there are regulations which are governing us possibly properly well.
So if I narrate this in three or four broader perspective of looking at accountability and explainability, when we leverage AI, we would want the AI to come and explain. Now, is it covered under the ambit of UL or in the data privacy? Maybe no at all, right? So we would want somewhere, Mansi actually narrated it very well, which is we would want somewhere a referenceable standard coming our way, where all of us can relate back easily and apply back. It could be blueprint, it could be playbooks, it could be. So such framework, does it exist for easily adaptable manner? The larger entities like us, we will be the first one possibly to invent the way to do through, make it as a playbook.
Related back to somebody who can make it as a blueprint and make it as a standard, then apply back to. the rest of the industry as a whole. So that’s the first and foremost. So the role of a TSP also is changing today, right? So from a conventional telecom provider, today we are talking about the previous example that I highlighted as an intermediary providing additional data inside. Now there is a law for digital intermediaries. Now the purpose for which a civilian has shared the data to me is for some other purpose. But the purpose beyond the purpose that he has shared to me, if I have put it to from a monetization standpoint, do I apply the ambit of digital intermediary also on me?
That’s a, that’s a, I wouldn’t want to comment as a, should my regulator look upon that and then put that also as applicable to me. But those are evolving space that we are looking at. And the last very famous topic amongst telcos that is floating around is on the spam and the scam protection, right? So here, let’s look at from again, Honorable PM, perspective, which is reasonable AI. Most of us associate reasonable AI back to explain. Now, imagine we have deployed scam solution which auto blocks things and we would want that AI to explain. Why did I block you? If it were to be blocked, then what am I looking at? I’m actually advancing the ability of scamster to know why am I blocking him so that he refines himself to not get blocked.
So that comes in the context of security. Do I do I make a framework? Do I make a guideline to tell here I would not want to have an explainability where security becomes a far more important element as compared to. So frameworks have to evolve. We need to have standards, but standards do not have the ability to make it universally applicable in all possible manner. So standards are taken, applied back as per individual enterprises and the context that we have to put them to use and then make it work. So I look. Look forward. Regulators will be innovative in allowing us to make the choices as appropriately while regulations can continue to evolve appropriately.
Thanks, Martin. I’ll take this conversation slightly global with my attempt, Deepak. How do you think India can leverage its DPI and telecom -led digital architecture to provide a credible, scalable model for the global south, particularly countries seeking digital sovereignty without technological isolation?
Okay. So when we are looking at somebody offering any technical solution to someone else, typically it comes with certain – It often comes with certain intellectual IPRs, intellectual property rights. So, for example, somebody is using a particular technology, so there could be patterns, there could be copyright, et cetera. Now, when India is offering its DPI -led model, nothing of that sort is going. Okay. So countries are able to adopt. It’s a framework. It’s a philosophy. And there’s an open protocol. So they can adopt it. They can adopt it. and they can change it the way they wish. So it is really open in that sense. So that’s one very important difference compared to let’s say some other country or company offering some particular technology but then it also involves certain type of monetization in terms of this is what you continue to pay us if you are scaling it to let’s say 1 million population, this is what you will pay us if you are doing it for 10 million or 100 million, this is what you will do.
India doesn’t ask for that type of thing. So that’s one very strong distinction. The second thing is in terms of the enablement. The enablement is also happening not just in terms of offering this as a technical sort of assistance, it is also happening through multiple other organizations. So for example, we have a research and information systems think tank under the Ministry of External Affairs and others is the Indian Council of World Affairs. So they are also doing a lot of work in terms of developing intellectual frameworks and capacity to do this as a matter of diplomacy itself. so that’s another dimension which is not often seen but it’s again a matter of soft diplomacy so for example three years back in 23 at ICW again I had proposed a framework called EOSS which was again basically about taking DPI in India I mean you can of course create a different acronym etc globally and again the focus was more around interoperability security etc there the other aspect is about standards so Mansi did distinguish between standards and systems or blueprints as she mentioned but one very important document I would again refer to a World Bank only so of course she did mention about the DPI report but even more recent document which has just come up a couple of months back from World Bank is the World Standard Development Report on Standards okay so I mean we all you look at traffic lights you look at traffic lights and you look at the traffic lights and you look at the traffic lights okay the three red amber green And this traffic light, the current traffic light standard came up only in 1968.
It’s not very old, okay? But it did happen. And this has become globally acceptable. But the way the design is, yes, you can put it vertical, you can put it horizontal, and there are other variations. So this is what it is doing there. So I think the way India is doing this is something that we are doing a lot of enablement across the global south. In fact, I just published a policy brief called Global South’s AI Pivot by CG of Canada just last Friday. Again, it talks about three things, equity, ethics, and ecology. So India is not only talking about things like, okay, it should be reasonable, it should be responsible, it should be accessible, it should be inclusive, accessible, all of that.
But also looking at things from an efficiency perspective. Efficiency is not just financial efficiency. Here we are talking about resource efficiency. So how do we manage these things with minimum? footprint of material, of energy, water, things like that. So, and this again goes back to something like the Prime Minister keeps on talking about this life, which is lifestyle for environment. Now this whole philosophy of
Thanks, Deepak. I’m conscious of time. Mansi, last one to you. You know, India’s approach to the DPI built on open, interoperable and scalable digital rails is increasingly influencing the global conversations. How do you see India’s DPI model shaping digital development strategies across emerging economies?
Thank you. I’ll keep it really short. I think at the bank we started working on ID for development and G2P and fast payments even before this whole big DPI push happened in India and particularly that became more socialized through the G20 process. and many other actors came across foundations, think tanks, technology companies, and started to socialize the idea of DPI and the DPI approach to digital transformation. India, surely for the vast amount of experience and scale and heterogeneity that it has, offers excellent evidence on what works and what doesn’t work. And it’s really great that a lot of the people who were part of the foundation and building of the DPI have now gone ahead and tried to take this to other countries in a way that is adaptable to them.
And there are so many organizations, without taking names, lest I miss out on other important ones, I don’t want to take that chance, but there are several organizations who are doing a fabulous job of doing that. And the government itself, so whether actively or indirectly, they are also trying to talk to the world about how the DPI approach works. And more actively, you know, in UPI and NPCI, as Martin was mentioning, there’s active collaboration on making these fast -paced… and systems work in collaboration with BIS to see can we actually think of the Finternet, the idea of the Finternet that came up with BIS. So I don’t see this dying down. I think we have a lot of, like I said, evidence of the foundations as well as now sectoral applications.
So there are just particularly because this is GSMA session and mobile, I don’t want to forget mentioning this really important part about how the Department of Telecom has begun to think about utilizing mobile data while the telcos are thinking from credit perspective and fraud management. They’re also thinking of it very actively in terms of using it for planning and mobility, which I think is really fabulous. It’s not as if other countries haven’t done it, but the DPI approach that they are taking towards it to scale the access to data, to make models available, to provide compute, and build that whole stack is not something that has happened. And obviously it’s going to evolve. I don’t think it’s perfect.
feel the pressure of making it perfect at go but this learning experiences will surely inform how other countries can do it. Some of the things that we are trying to do it at population scale. Yes exactly.
So I think if I can just have one question from the I can see three hands already how much time do we have? Do we have a question for two one question gentlemen please state your name and to whom do you want to address this question
Mike I am Vijay Agarwal and I am interested in AI by profession I am manufacturer of jewelry so what I wanted to propose was why don’t we have a product like a ring kind of product where the privacy data the KYC data resides on that physically only on that item which is on the body and then we can if it leaves the body it leaves in an encrypted form only and it can only be collated with another key for the purpose for which it has been consent has been given and there is a blockchain record to it.
You mean in the form of a jewelry?
Yeah, so we have Adha ring for every Indian and it will store the KYC record, the medical record which could be accessed in case of emergency but there should, all these control layers that you are talking about could be in the form of cryptography. The concept of data embassies as part of the discussion on data sovereignty, so is there a good case for maybe India to offer data embassies? obviously it will be on a multilateral but any thoughts on that
I would say yes if it is on reciprocal basis
let me try and address the first part which you were trying to say I think today it’s not the problem of your data being insecure with Aadhaar I think it’s very secure right there are lot of things which Aadhaar does there is also the masking which they have started so the leakage of data or private data is really not the issue out here the data going out has got various other forms and factors particularly the way the government is taking the data from users it is the government which has to really start looking at for example telcos are required to share the subscriber data every month in physical copies why would you do that right so it is not really the digital aspect which is a problem it is really how you are managing the data is a problem and I think quantum work has already started sir I think Aadhaar itself is working on that on data embassies Vikram I think I completely endorse you know Deepak it cannot be just me right look around and have it and so let’s play it right but you cannot expect the world’s largest data creator and consumer to be the ones to start offering this first it is a two way street right for too long I think as a country we have been you know in a sphere where we are supposed to give and we are not supposed to take anything that has to change
the organizer is already standing on my head just wanted to say one thing only mentioned in terms of government taking data so about 20 not now of course now IRCT doesn’t do it but till about 15 years back or so if you are creating an IRCTC ID for first time it used to ask even your marital status and there were apparently no benefits or disadvantages and it was a compulsory field by the way I would like to thank each of the speakers here to make it a very engaging conversation, thank you Mansi Rahul, Deepak, Matan for your time and to have this session, thank you very much audience thank you Thank you.
Thank you, Julian. Thanks for the opening remarks. Am I audible? Looks like yes. So let’s begin. We have a fantastic panel here of experts. So let’s start with this discussion. So what we have seen ov…
Event“We’re talking of AI being a possible DPI, a digital public infrastructure.”<a href=”https://dig.watch/event/india-ai-impact-summit-2026/building-indias-digital-and-industrial-future-with-ai?diplo-dee…
EventDPI was portrayed not just as infrastructure but as part of an inclusive ecosystem involving legal, financial, and societal frameworks that enable technology to serve the public interest effectively. …
EventThe discussion revealed that data sovereignty encompasses more than simple data localization. As Pramod noted, true sovereignty requires control over legal frameworks governing data access, encryption…
EventThis paradoxical statement challenges the typical understanding of digital sovereignty as protectionist or isolationist. It introduces a nuanced view that true sovereignty requires strategic collabora…
EventMarie Ndé Sene Ahouantchede explains that ECOWAS views public digital infrastructure as built on three pillars: payment systems, data management, and digital identity. These infrastructures are design…
EventIn aletter dated 1 September 2023and transmitted to all permanent representatives and permanent observers to the UN in New York, Rwanda and Sweden – as co-facilitators of the Global Digital Compact (G…
UpdatesMaintaining an open, secure, and interoperable internet while avoiding fragmentation was another key point of discussion. The participants acknowledged that the Internet Corporation for Assigned Names…
EventInteroperability is key for enabling cross-border use of DPI and preventing fragmentation.
EventConsumer centric safeguards obviously by way of transparent disclosure clear appeal processes and human intervention mechanism are critical to maintain public trust. These pathway ensure that governan…
EventWe joke that we shouldn’t worry about AI until we figure out AV. So I guess this is a perfect example of that. Thanks for the question, Arat. I think maybe the first thing to say about this, and this …
Event### India: Flexible Modular Architecture Brazil’s PIX payment system exemplifies successful regional innovation, now expanding internationally. Similarly, India’s UPI system is gaining traction in th…
EventI was just going to do one more thing, which is thank you, Shalini, and thank you to the panel for allowing us this small break. For those of you who don’t know, Secretary Krishnan from the Ministry o…
EventIt is being designed as a replicable public infrastructure model for India and the entire global south. In partnership with India AI Mission the government of Maharashtra, the World Bank and the Wadhw…
EventDPI offers various benefits for meeting the SDGs through effective data collection and utilisation. According to a policy brief by the UN Secretary, DPI has the potential to contribute to the SDGs by …
EventThe tone was consistently optimistic and collaborative throughout, with speakers expressing urgency about the civilizational importance of the work while celebrating partnerships and community-driven …
EventThe tone was consistently collaborative, optimistic, and mission-driven throughout the conversation. Speakers demonstrated mutual respect and shared commitment to inclusive AI development. The atmosph…
EventThe tone began very positively and constructively, with the Chair commending delegations for focused, specific interventions rather than general statements. Speakers expressed appreciation for the Cha…
EventThe discussion began with a collaborative and appreciative tone as various stakeholders shared their visions and commitments. However, the tone became increasingly tense and critical during the explan…
EventThe discussion maintained a consistently optimistic and collaborative tone throughout. Speakers expressed enthusiasm about AI’s potential benefits while acknowledging challenges responsibly. The tone …
EventVery low level of disagreement. The speakers were largely aligned on goals and strategies, with differences mainly in emphasis rather than fundamental opposition. The main disagreement was on whether …
EventThe discussion maintained a consistently collaborative and solution-oriented tone throughout. It began with an authoritative presentation of frameworks and evolved into a pragmatic exchange of real-wo…
EventThe discussion maintained an optimistic and collaborative tone throughout, with speakers consistently emphasizing human resilience and adaptability. While acknowledging legitimate concerns about AI’s …
EventThe tone was collaborative and solution-oriented throughout, with speakers building on each other’s insights rather than debating. It maintained a balance between technical expertise and practical imp…
Event-Internet Governance and Multistakeholder Approach: Strong support for strengthening the Internet Governance Forum (IGF) with a permanent mandate and sustainable funding, while reinforcing multistakeh…
EventThe discussion maintained a collaborative and constructive tone throughout, despite addressing complex and sometimes contentious issues. While there were moments of challenge and critique (particularl…
EventThe discussion maintained a consistently collaborative and constructive tone throughout. Panelists demonstrated remarkable consensus on the importance of standards, with no significant disagreements o…
EventThe discussion maintained a professional, collaborative tone throughout, characterized by constructive problem-solving rather than confrontational debate. Speakers acknowledged both the challenges and…
EventThe tone was constructive and solution-oriented throughout, with speakers building on each other’s points rather than debating. It began with high-level policy perspectives and gradually became more t…
EventAt the opening session of the Summit, the G20 agreed to admit by consensus the African Union (AU) as a permanent and equal member of the grouping. With this single decision, 55 developing countries of…
BlogThe tone was consistently optimistic and inspirational throughout. Sunak maintained an enthusiastic, forward-looking perspective that celebrated both technological progress and human potential. The sp…
EventIn conclusion, the delegate’s remarks highlighted the enduring spirit of solidarity and collaboration, while also conveying gratitude for the persistent contributions of the Chair and other key indivi…
EventThe tone was formal and optimistic throughout, maintaining a diplomatic and collaborative atmosphere. Speakers consistently emphasized hope and possibility while acknowledging challenges. The discussi…
EventThe discussion maintained a consistently professional and collaborative tone throughout. It began with formal introductions and technical explanations, evolved into an enthusiastic presentation of pra…
EventThe necessity of forging strategic alliances to transverse sectoral barriers and to embrace digitalisation in agriculture was emphasised. In a bid to close, the speaker noted with an apologetic tone t…
EventTheir call for consensus may be a strategic diplomatic move to present a united stance internationally, thereby circumventing potential disputes or misunderstandings stemming from any proposed edits. …
EventThe tone of the discussion was formal yet collaborative and appreciative. There was a sense of accomplishment for the work done during the track, gratitude expressed to organizers and hosts, and enthu…
EventGratitude was expressed towards both presenters and participants for engaging in the dialogue.
Event“The session opened with Debashish Chakraborty introducing the theme of “convergence of AI, telecom and data sovereignty all weaved around the digital public infrastructure” and handing over to Julian Gorman, head of APAC GSMA, for the keynote address.”
The knowledge base explicitly mentions the convergence of AI, telecom and data sovereignty around digital public infrastructure and references Julian Gorman as head of APAC GSMA being invited to give the keynote, confirming the report’s description.
“Julian Gorman positioned GSMA as the global body that unites the mobile economy to “unlock the power of connectivity so industry and society thrive”.”
GSMA’s broader mandate is described in the knowledge base as promoting digital, mobile‑based solutions worldwide and advancing innovative solutions that drive economic growth, providing context for its role as a unifying global body.
“Digital sovereignty is defined as strategic control over infrastructure, standards and the underlying intelligence.”
The knowledge base defines digital sovereignty as a nation’s ability to understand, develop and regulate digital technologies to maintain control and self‑determination, adding nuance to the report’s definition.
“Rahul Vatts highlighted the scale of India’s Digital Public Infrastructure (DPI), noting UPI’s massive transaction volume.”
The knowledge base notes that UPI is the world’s largest digital payment system, underscoring its massive scale and supporting the claim of a large‑volume DPI, though it does not provide the exact monetary figure cited.
“The discussion was part of a panel titled “AI in Digital Public Infrastructure (DPI) – India AI Impact Summit”.”
A panel discussion on AI in DPI at the India AI Impact Summit is recorded in the knowledge base, confirming the existence of such a session.
There is strong consensus that telecom networks are now AI‑enabled public infrastructure requiring trusted, interoperable, and standards‑driven operation. Participants agree that data sovereignty must extend beyond mere localisation to include control‑plane, operational and legal dimensions, and that open APIs, contextual enrichment and open‑protocol DPI models are essential to avoid duplication and to scale solutions globally, especially for the Global South.
High consensus across technical, regulatory and policy dimensions, indicating a shared vision that can underpin coordinated actions on standards, open‑source DPI frameworks and sovereign cloud strategies.
The discussion reveals broad consensus that telecom networks must become intelligent, trusted platforms for AI‑driven public services. However, speakers diverge on how to achieve data sovereignty, the appropriate normative tools for DPI (global standards vs. flexible blueprints), the risk of parallel private sovereign‑cloud solutions creating duplicate DPI layers, and the regulatory focus—whether on explainability and digital‑intermediary rules or on technical‑jurisdictional control. These disagreements are substantive but not antagonistic; they reflect different institutional perspectives (government, regulator, operator, multilateral development bank) rather than outright conflict.
Moderate. The disagreements are primarily about implementation pathways and normative instruments rather than fundamental goals. This suggests that while the community shares a common vision of secure, AI‑enabled digital public infrastructure, coordination will be needed to reconcile technical‑jurisdictional frameworks, standard‑blueprint choices, and the balance between private sovereign‑cloud offerings and open‑API collaboration. The implications are that policy harmonisation, joint standard‑development processes, and clear regulatory guidance will be essential to avoid fragmentation and to realise the trusted, interoperable network vision.
The discussion was steered by a series of pivotal insights that moved it from high‑level rhetoric to concrete, actionable ideas. Julian’s redefinition of sovereignty set the thematic foundation, which was then grounded by Rahul’s real‑world scale of UPI and the trust layer. The introduction of context‑enriched data and open APIs (Speaker 1) opened a technical pathway, while Deepak’s ‘walls’ metaphor and Rahul’s four‑slice sovereignty framework reframed the policy debate around collaboration versus isolation. Mansi’s clarification of standards versus blueprints and Deepak’s emphasis on an open, non‑IP model provided a pragmatic roadmap for exporting India’s DPI to the Global South. Finally, the audience’s wearable‑data concept illustrated how these ideas could manifest in future consumer products. Collectively, these comments shifted the conversation from abstract notions of digital public infrastructure to a nuanced, multi‑dimensional view that integrates technology, regulation, standards, and geopolitics, shaping a forward‑looking agenda for both India and emerging economies.
Disclaimer: This is not an official session record. DiploAI generates these resources from audiovisual recordings, and they are presented as-is, including potential errors. Due to logistical challenges, such as discrepancies in audio/video or transcripts, names may be misspelled. We strive for accuracy to the best of our ability.
Related event

