WS #98 Universal Principles Local Realities Multistakeholder Pathways for DPI

Summary

This panel discussion, moderated by Sabhanaz Rashid Diya from the Tech Global Institute, focused on Digital Public Infrastructure (DPI) and its implementation across different countries, particularly in the Global South. The conversation brought together government officials, academics, and researchers from India, Brazil, Estonia, Dominican Republic, and South Africa to examine both the promises and challenges of DPI systems.

The panelists presented varied experiences with DPI implementation. Luca Belli from Brazil highlighted the success of PIX, Brazil’s digital payment system, which broke the Visa-MasterCard duopoly and demonstrated how DPI can enhance competition while keeping data governance local. In contrast, Keith Breckenridge from South Africa offered a more cautionary perspective, noting that while DPI represents an improvement over previous systems, financial inclusion hasn’t delivered promised benefits and has created new vulnerabilities, particularly for elderly populations susceptible to online fraud and gambling.

Sheo Bhadra Singh from India’s Telecom Regulatory Authority emphasized how India’s DPI journey began with addressing basic service delivery problems, leading to the creation of Aadhaar digital identity system that now serves 1.33 billion Indians. He stressed that DPI enabled massive financial inclusion through Jan Dhan accounts and UPI payments, processing billions of transactions monthly. However, Smriti Parsheera and Bidisha Chaudhury raised concerns about the strong state-market alliance in DPI development, questioning whether digitalization truly equals inclusion and highlighting the risk of creating “alt big tech” monopolies.

Armando Manzueta from Dominican Republic shared his country’s approach to building DPI as a “nation-building tool” grounded in rights and trust, emphasizing the importance of legal frameworks and civil society oversight. Rasmus Lumi from Estonia outlined the Freedom Online Coalition’s rights-respecting DPI principles, including human-centered design, inclusivity, transparency, and interoperability. The discussion concluded with agreement that successful DPI must be citizen-centric, ethical, and accountable, with proper safeguards embedded in law and meaningful participation from civil society to ensure these systems truly serve the public interest.

Keypoints

## Overall Purpose/Goal

This panel discussion aimed to examine Digital Public Infrastructure (DPI) from multiple perspectives, focusing on governance models, implementation experiences across different countries, and the balance between technological innovation and human rights protection. The conversation sought to identify universal principles for DPI while acknowledging that national contexts require different approaches.

## Major Discussion Points

– **Defining DPI and National Implementation Models**: Panelists explored various definitions and approaches to DPI, with examples from India (Aadhaar, UPI), Brazil (PIX payment system), South Africa, Estonia (X-Road), and Dominican Republic. The discussion highlighted how different countries have developed DPI solutions based on their specific problem statements – from financial inclusion to government service delivery.

– **Ownership, Governance, and Public vs. Private Control**: A central debate emerged around who should own and control DPI systems. The conversation contrasted Brazil’s fully public approach (enabling Freedom of Information requests) with India’s foundation-based model, and examined concerns about creating new monopolies while trying to break existing ones, particularly regarding big tech influence.

– **Rights, Privacy, and Accountability in DPI Systems**: Panelists discussed the importance of embedding human rights principles into DPI architecture, including data protection, transparency, and citizen control over personal data. Estonia’s approach of allowing citizens to track who accesses their data was highlighted as a best practice, while concerns were raised about exclusion and the risks of digital systems.

– **Financial Inclusion Benefits and Risks**: The discussion examined both the promise and perils of DPI-enabled financial inclusion. While acknowledging successes in expanding access to banking and payments, panelists raised concerns about predatory lending, online gambling, fraud vulnerability (especially for elderly users), and whether digital access truly equals meaningful inclusion.

– **Stakeholder Collaboration and Civil Society Role**: The conversation addressed the need for multi-stakeholder approaches to DPI development, questioning whether civil society has adequate voice in predominantly state-market partnerships. Panelists emphasized the importance of public participation, regulatory oversight, and the need for trusted intermediaries to help citizens navigate complex digital systems.

## Overall Tone

The discussion maintained a balanced, analytical tone throughout, combining optimism about DPI’s potential with realistic assessments of its challenges. Panelists were respectful but candid about both successes and failures in their respective contexts. The tone was collaborative rather than confrontational, with speakers building on each other’s points and acknowledging the complexity of implementing people-centered DPI. The conversation remained academically rigorous while staying grounded in practical implementation experiences.

Speakers

**Speakers from the provided list:**

– **Sabhanaz Rashid Diya** – Moderator, Tech Global Institute (not-for-profit focusing on global south digital rights and policy issues), Advisory network member of Freedom Online Coalition

– **Rasmus Lumi** – Director General of the Department of International Organizations and Human Rights at the Ministry of Foreign Affairs of Estonia

– **Luca Belli** – Professor at FGV Law School, Rio de Janeiro, and Director of the Center for Technology and Society

– **Smriti Parsheera** – Research Fellow at the Interledger Foundation, co-convening the panel with the Freedom Online Coalition

– **Sheo Bhadra Singh** – Principal Advisor for the Telecom Regulatory Authority of India

– **Armando Manzueta** – Vice Minister for Public Innovation and Technology at the Ministry of Public Administration of the Dominican Republic (online participant)

– **Keith Breckenridge** – Standard Bank Chair in African Trust Infrastructures at the University of Witwatersrand, South Africa (online participant)

– **Bidisha Chaudhury** – Assistant Professor of Government, Information, Cultures, and Digital Citizenship at the University of Amsterdam (online participant)

– **Audience** – Audience member asking questions (identified as Elizabeth, working with mobile industry association, previously worked with government of Nigeria on digital policy reforms)

**Additional speakers:**

None identified beyond those in the speakers names list.

# Digital Public Infrastructure: Governance, Implementation, and Human Rights – Panel Discussion Report

## Introduction and Context

This panel discussion, moderated by Sabhanaz Rashid Diya from the Tech Global Institute, examined Digital Public Infrastructure (DPI) from multiple perspectives, bringing together government officials, academics, and researchers from India, Brazil, Estonia, Dominican Republic, and South Africa. The conversation emerged from India’s G20 presidency promotion of DPI and has become a significant topic across global processes including Brazil, South Africa, the Freedom Online Coalition, the Global Digital Compact, and various UN bodies.

The discussion addressed a fundamental tension: whilst DPI represents deeply national technology for citizen databases and processes, it requires global dialogue and common language for interoperability. As Diya noted, this creates both opportunities and challenges for developing universal principles whilst respecting national contexts and sovereignty.

## Defining Digital Public Infrastructure

The panellists offered varied perspectives on what constitutes DPI. Armando Manzueta from the Dominican Republic provided a holistic definition, describing DPI as “not just a set of technologies but a nation-building tool that must be grounded in rights, resilience, and trust.” This framing moved beyond technical specifications to encompass broader governance and social objectives.

The discussion revealed that DPI implementations vary significantly across countries based on their specific problem statements. India’s journey began with addressing service delivery challenges for excluded populations. Brazil focused on breaking payment monopolies. Estonia built comprehensive digital government services over 25 years. Each approach reflected national priorities and contexts whilst contributing to a broader understanding of DPI possibilities.

## National Implementation Models and Experiences

### India’s Comprehensive Ecosystem

Sheo Bhadra Singh from India’s Telecom Regulatory Authority presented the most comprehensive DPI implementation, beginning with Aadhaar in 2009-2010 to provide public services access to large excluded populations. The system now serves 1.33 billion Indians with digital identities, enabling 550 million zero-balance bank accounts and processing 18.77 billion UPI transactions worth $290 billion in a single month of 2025.

Singh highlighted several major platforms beyond identity and payments: GEM (Government E-Market Place) with 11,000 products, 336 services, 164,000 buyers, and $625 billion in procurement; ONDC (Open Network of Digital Commerce) for democratizing e-commerce; and a health platform with 787 million registrations. He emphasised that India’s success required the state to play three key roles: creating institutional space through the National Payments Corporation of India (NPCI), scaling through policy mandates, and providing regulatory oversight of the multi-stakeholder ecosystem.

Singh stressed that DPI must be “user-centric and citizen-centric with the goal of serving society as the basic principle from which everything else can be built.”

### Brazil’s Competition-Focused Approach

Luca Belli from FGV Law School highlighted Brazil’s PIX payment system as exemplifying “good digital sovereignty” that fosters competition, domestic innovation, and people empowerment. PIX broke the Visa-MasterCard duopoly, reducing transaction costs from 3-5% to zero whilst distributing data collection and maintaining public accountability through freedom of information laws.

Belli provided important context about international payment system development, explaining how Russia developed the MIR system after 2014 sanctions, India innovated with UPI, and Brazil enhanced the model with PIX. He argued that PIX demonstrates how DPI can simultaneously increase customer welfare, enhance competition, produce innovation, and distribute data governance. Crucially, local entrepreneurs who innovate using this infrastructure pay taxes in Brazil, creating “a very different dynamic of empowerment through DPIs” compared to foreign-controlled systems.

Belli also emphasised the crucial role of India’s 2016 net neutrality regulations, which prohibited zero rating and dropped connectivity prices by 90%, enabling an innovation boom by making all services equally accessible.

### Estonia’s Long-term Digital Governance

Rasmus Lumi from Estonia’s Ministry of Foreign Affairs outlined a 25-year journey of building digital public infrastructure starting around 2000, achieving nearly 100% online public services access. Estonia’s model ensures every citizen owns and controls their data, with the ability to verify which officials accessed their information, providing what Lumi termed “preventive accountability.”

Lumi noted that “I think divorce is now, I think, the only thing that we cannot do online yet” in Estonia. He stressed that “trust is the most important but uncodifiable element for DPI success – people can be forced to use systems but won’t maximise benefits without trust.”

Under Estonia’s chairship, the Freedom Online Coalition is developing rights-respecting DPI principles including human-centred solutions, inclusivity, transparency, privacy, security, sustainability, data-driven approaches, and technology neutrality.

### Dominican Republic’s Rights-Based Framework

Manzueta shared the Dominican Republic’s approach to DPI adoption beginning in 2020, focusing on data interoperability using Estonia’s X-Road platform. With 71-72 institutions currently using the ecosystem for data exchange, the country emphasises shared responsibility and oversight from civil society and private sector in DPI governance and implementation.

The Dominican Republic faces significant financial inclusion challenges, with roughly half the population financially excluded and only 30% of the included receiving broader financial services. This context shapes their approach to building DPI as a foundation for broader development goals.

### South Africa’s Cautionary Experience

Keith Breckenridge from the University of Witwatersrand provided the most critical perspective, noting that whilst South Africa has had digital welfare distribution since the early 1990s and biometric population registers for 20 years, financial inclusion hasn’t delivered promised benefits and has created new vulnerabilities.

Breckenridge highlighted alarming statistics: one trillion Rand was spent on online gambling in 2023 – equivalent to one-third of South Africa’s GDP. He warned that “we’re building an infrastructure that’s going to make a huge population newly vulnerable to really serious crimes, and there’s no discussion of what we should be doing about that.”

He also raised concerns about the Trump administration’s withdrawal from regulatory frameworks that could affect DPI oversight and accountability mechanisms.

## Key Debates and Disagreements

### Public versus Quasi-Public Ownership

A significant disagreement emerged between Belli and Smriti Parsheera regarding ownership models. Belli argued that Brazil’s PIX is “truly public because it’s delivered by government (Central Bank), ensuring freedom of information access, unlike India’s foundation-based model that can avoid transparency requirements.”

Parsheera countered with concerns about the “strong state-market alliance” in DPI implementation, warning of risks of creating “alt big tech that pursues scale recklessly while being blind to due process.” She questioned whether DPI initiatives were truly different from private tech monopolies in their approach to scaling without adequate safeguards.

### Persistent Big Tech Influence

Bidisha Chaudhury from the University of Amsterdam raised a crucial point about the persistence of big tech influence even in state-owned DPI systems. She noted that despite UPI being state-owned infrastructure, users primarily identify it with Google Pay rather than recognising it as public infrastructure.

Chaudhury posed a fundamental challenge: “How do we actually circumvent the influence of big tech when we rely on an infrastructure where big tech interests are so much more embedded already?” This question highlighted the gap between policy intentions and ground reality in DPI implementation.

### Financial Inclusion: Benefits versus Risks

Singh celebrated massive scale achievements in financial inclusion, whilst Breckenridge provided a stark counternarrative about how financial inclusion through DPI can enable harmful behaviours. Beyond gambling statistics, he highlighted risks of usurious lending at 1% daily interest and vulnerability to fraud, particularly affecting elderly populations who lack digital skills.

Parsheera raised important questions about how financial inclusion is defined, asking whether it should encompass only payments and remittances or include broader financial services like insurance. She also questioned whether biometric-based ID systems are necessary for achieving financial inclusion goals.

## Governance Challenges and Civil Society Participation

Parsheera highlighted that “civil society participation is limited in DPI deployment, which follows top-down approaches without robust transparency and accountability culture like regulatory impact assessments.” This observation revealed a significant gap between DPI rhetoric about multi-stakeholder governance and implementation reality.

Belli emphasised that successful DPI implementation requires “independent, well-resourced institutions that understand the technology and can work with stakeholders, plus systemic vision considering connectivity and affordability.” This institutional capacity extends beyond technical expertise to include governance capabilities and stakeholder coordination skills.

## Accountability and Protection Mechanisms

### Metrics of Exclusion

Breckenridge introduced the crucial concept of “metrics of exclusion,” arguing that DPI systems should make visible “the number of people whose identity numbers are at the moment disabled and how long it’s taken to answer queries around those numbers.” He cited South Africa’s example where two million identity numbers were disabled by simply editing a comment field, preventing access to all digital services.

### Liability Frameworks

Breckenridge called for liability frameworks where financial institutions bear responsibility for fraud losses, referencing Britain’s model of bank liability. This approach would shift responsibility from individual citizens to institutions better equipped to prevent and address fraud.

### Privacy and Data Control

Lumi outlined the complexity of privacy in DPI systems, identifying two dimensions: external privacy from outside users and internal privacy with checks on government access to citizen data. Estonia’s approach allows citizens to track who accesses their data, providing transparency and control over personal information.

## Technical Vulnerabilities and Risks

The discussion revealed multiple layers of risk in DPI systems. Breckenridge highlighted technical vulnerabilities including backend hacking where criminals steal encryption codes, plus social engineering attacks targeting vulnerable populations. The fragility of systems was demonstrated by South Africa’s experience where editing a comment field disabled two million identity numbers.

A recurring theme was the particular vulnerability of elderly and digitally unskilled populations to fraud and exploitation through DPI systems. The discussion revealed that whilst DPI can expand access, it can also create new forms of exclusion and harm for those unable to navigate digital complexity.

## Future Directions and Unresolved Questions

Several fundamental questions remained unresolved throughout the discussion:

1. **Digital Access versus Meaningful Inclusion**: Whether digital access automatically translates to meaningful social and economic inclusion remains contested.

2. **Governance Models**: How to balance state-market alliances without creating new monopolies whilst ensuring public accountability.

3. **Big Tech Influence**: How to achieve true digital sovereignty when private platforms control user interfaces and experiences.

4. **Vulnerability Protection**: How to protect elderly and digitally unskilled populations from exploitation whilst expanding digital access.

5. **Civil Society Participation**: How to move beyond token consultation to meaningful participation in DPI governance.

## Conclusion

This discussion revealed DPI as a complex phenomenon that defies simple categorisation as either beneficial or harmful. The panellists demonstrated that whilst DPI can deliver significant benefits in terms of financial inclusion, service delivery, and breaking monopolies, it also creates new risks and vulnerabilities that require careful attention.

The conversation highlighted the importance of context-specific approaches whilst identifying universal principles around transparency, accountability, and human-centred design. The tension between celebrating technical achievements and addressing unintended consequences suggests that DPI development requires more sophisticated evaluation frameworks and stronger governance mechanisms.

The diversity of perspectives represented in this panel – from government officials celebrating scale achievements to academics warning about systemic risks – reflects the broader global debate about digital governance and the role of technology in development. Key areas requiring continued attention include developing meaningful civil society participation mechanisms, creating robust accountability frameworks that make exclusion visible, and addressing the persistent influence of big tech platforms even within state-owned infrastructure.

Ultimately, the discussion reinforced that successful DPI implementation requires not just technical expertise but also institutional capacity, stakeholder engagement, and a commitment to protecting vulnerable populations whilst ensuring these systems truly serve the public interest.

Sabhanaz Rashid Diya: Okay. Let’s get started. Hello and good morning. I am very happy to see that there are still here people on Friday. So fantastic. And I’m very excited about our panel today. For those of you who don’t know me yet, I think I’ve met many of the people here already, but I am with the tech global institute. We are a not for profit that focuses on global south digital rights and policy issues and I’m very excited to be moderating this panel and being joined by a fantastic set of panelists both in person and online. We have about 90 minutes together and we’re going to divide this time up with some conversations with the panelists as well as an opportunity for the audience here as well as online to ask us some questions. So on that note, maybe I can start by introducing our panelists and then moving on to kind of talking a bit more about the discussion topic today, which is something that’s top of mind for all of us, which is digital public infrastructure. A topic that has been actually started by the Indian presidency of the G20 and has really moved on from there to Brazil and now South Africa is now also a major topic of the Freedom Online Coalition, which is where we are. I’m an advisory network member. A number of other global processes where DPI is increasingly coming up, whether that’s the GDC, whether that is some of the work from UNDP and a number of other UN bodies and UN processes. So this is an important topic, especially for those of us who are in the global majority. And I’m sure many of our panelists will have lots to say about that as well. So on that note, I am going to start with my panelists here. From the extreme left of me is Rasmus Lumi, who is the Director General of the Department of International Organizations and Human Rights at the Ministry of Foreign Affairs of Estonia. Then we have Luca Belli, who is a professor at FGV Law School, Rio de Janeiro, and Director of the Center for Technology and Society. We then have Smriti Parsheera, Research Fellow at the Interledger Foundation, who is also co-convening this panel with the Freedom Online Coalition. We then have Sheo Bhadra Singh, who is the Principal Advisor for the Telecom Regulatory Authority of India. That’s our in-person panelists. And online, we are being joined by Mr. Armando Manzueta, who is the Vice Minister for Public Innovation and Technology at the Ministry of Public Administration of the Dominican Republic. Keith Breckenridge, who is the Standard Bank Chair in African Trust Infrastructures at the University of Witwatersrand, South Africa. I feel like I butchered that name, but hopefully Keith will correct me. As well as Bidisha Chaudhury, who is the Assistant Professor of Government, Information, Cultures, and Digital Citizenship at the University of Amsterdam. So a star-studded panel, as we say, but I’ll probably get started on that. So as I was saying, digital public infrastructure, and I was actually just in a panel right before, where we also quite talked about how do we think about DPI, not just within the context of specific processes, but perhaps all processes. And we were discussing that in the context of the G7 presidency as well, where it hasn’t really come up as much as a priority. But there is real opportunity for that, particularly because this is such an important issue for the global majority. But often as we think about. regional public infrastructure, there’s a bit of a tension, because in many ways, this is a technology that’s deeply national in many ways, right? Because it is about building for your national citizen databases, looking into various national processes, and really strengthening the role of the public sector in delivering services and opportunities for its citizens. But increasingly, these national processes are also becoming, because they’re part of these global dialogues and global conversations right now, there is an increasing need to find a common language around DPI. DPIs can be of many shapes, forms, we still are lacking a very universal definition of DPI. I think we are maybe at the tail end of that debate now, which is great, but that was a major issue of discussion for several years. But because so many countries are finding their own adoption models, thinking through their own public infrastructures, there is increasingly a need for thinking, what is the universal element of it? Are there certain principles that ties DPIs across different kinds of countries? How do we think about governance in this space? And are these governance approaches cross-border? Should they be cross-border? Can they be cross-border? And what are some of the political contexts, the social contexts driving DPI? Because while we want DPIs to be interoperable within national boundaries, we also want them to have some level of interoperability within international borders as well, because the whole point is to really tackle the data fragmentation problem, tackle some of the EGOF fragmentation problems. So there is a lot of promise there. So those are some of the, I guess, topics that I hope we will be discussing today. But that is sort of the context and backdrop within which I am having this conversation. I would also like to flag that as part of the Freedom Online Coalition, Estonia has been, during their chairship, leading on the rights-respecting DPI principles effort as well that Rasmus will speak about, which is also one way of really helping us think through those universal principles. So on that note, maybe I will start with Arma. with Luca, actually, who’s here with us today, maybe share a little bit more about, you know, what are, what do you think are some of the social, political, economic incentives driving DPI adoption in countries? And perhaps, you know, what, what’s, what’s some of the consideration that governments and other stakeholders should be having when they’re thinking about implementing DPI?

Luca Belli: Thank you very much for inviting me here and for organizing this very timely discussion. And let me also clarify that most of the thoughts I’m going to share are the result of our empirical research in one of our flagship projects at the center, which is called CyberBRICS, that analyzes the digital policies of the BRICS grouping. And so we have been analyzing data governance, cybersecurity, digital transformation, AI governance over the past seven years in the various phases. And a parallel theme that has emerged is the theme of digital sovereignty, which is what has led us to analyze even more in-depth DPIs. And let me tell you why. So we have recently published a book, which is online in open access on the site of the editor of Cambridge University Press on digital sovereignty in the BRICS. And by analyzing how the BRICS grouping has shaped the narrative, the initiative on digital sovereignty, you immediately understand that there is an ample spectrum of initiatives. Some of them are those initiatives that we analyze, but we do not recommend to reproduce, that flirt with authoritarianism or protectionism. Some others are some very interesting initiatives that I define as good digital sovereignty, that basically utilize the construction of technology to foster competition, to foster domestic innovation, and to empower people. And one of the very good examples of this is a digital public infrastructure we use in Brazil for online payments, PIX. I’m not sure if everyone knows it, but the So, I’m going to talk a little bit about the digital public infrastructure, and why this is a very good example of digital public infrastructure, and why this is a very good example of good digital sovereignty, and why this is something that automatically triggers from our experience user empowerment, innovation, and competition, and also taxation of the innovation that is happening in the world. So, the first way we have in Brazil, we had in Brazil to perform electronic payments was through Visa and MasterCard. So, we were totally dependent on a duopoly, two companies that are foreign companies, nothing against, I have literally zero against them, but they charge between 3% and 5% for every transaction. So, in every country, the consumers pay 3% for every transaction, and the consumers pay 5% for every transaction to Visa and MasterCard if you are dependent on them. So, since the Brazilian central bank that has developed the PIX with a multi-stakeholder effort with specialists and also with banks and financial intermediaries, and also implements it through a multi-stakeholder process, because there is something called the PIX forum, where all the implementers, the financial sector representatives can participate in the process, and the investors can also participate in the process. So, since this has been implemented, not only the duopoly has been broken, which is something that we should really celebrate, and our antitrust authority, through conventional means, would have never been able to do this, the duopoly has been broken by creating an alternative. And that is extremely powerful. So, the duopoly has been created, and it has dramatically reverted in the pockets of cost of customers, 3% to 5%, which is already something stellar. but also that this has distributed data collection. So people do not realize that over the past 10 years, Visa and MasterCard have become big data companies. So the main part of their revenue is not the three to 5% they collect per transaction, is the data they collect and they process, and thanks to which they generate AI and new services that usually also is the intellectual property of which is usually then delocalized to fiscal havens like the Cayman Islands and the Virgin Islands. So what is interesting here is not only that the customer welfare is increased, but also that competition is automatically enhanced, innovation is produced, and data governance is distributed. And the local entrepreneurs that innovate thanks to this, then they have to pay taxes in Brazil. So we have a very different dynamic here of empowerment through DPIs. So that is why we really see this as an incredible example of good digital sovereignty that actually Brazil has to some extent copied from India. India has UPI, and actually it’s an untold story that we recount in our book that India to some extent took inspiration from Russia, another BRICS country that in 2014, after he invaded for the first time Ukraine and next in Crimea, was blocked from the first sanction was the blockage of Visa and MasterCard. From overnight, they had to create a new system, it’s called MIR, like the space station, but it is based on a card. So it’s a very 20th century conception of it. The great innovation of the Indians has been leveraging their skills and do it a software-based digital public infrastructure. And I think to some extent, Brazil has enhanced it because if you look at the Indian example, that is something that in our observation makes it much easier to implement DPIs to delegate them. to the implementation, at least, to a non-public stakeholder. In India is the National Payment Corporation of India, which is a foundation, which is much more agile in implementing things than bureaucracies, and everyone knows the limits of government and bureaucracy in this sense. But I think that what is the problem here is that if you want to file a freedom of access to information in India, so let’s say that you are a user and you want to know which data it’s collected about you and which are the data security measures implemented, they can reply to you, we are sorry, this is not a public organ, this is a foundation, so we are not bound by freedom of access to information. And I think that this could be easily improved in India, but so far, I see, I have this criticism, the only criticism I actually have on the UPI in India. In Brazil, this has been improved, because it is a public, it’s a truly public service, it’s not something public only because it is publicly accessible, it’s public because it’s the government that delivers it. And so if I file a freedom of access to information request to the Brazilian Central Bank, they have to reply me, otherwise I sue them. And they have to, at that point, they will have to reply me. So if I, as a Brazilian citizen, I want to know which kind of data are collected and what are the measures, they need to tell me. So it’s much more performing in terms of accountability. So I think that for all these reasons, I think that there is a lot of room for considering DPR something very good for digital sovereignty, for good digital sovereignty, but also there are some limits, and so we should not praise them as if it was a religion, but be very pragmatic and analyze the fact. Thank you.

Sabhanaz Rashid Diya: Thanks, Luca, that was very helpful. And I think perhaps that’s where I would love to kind of bring in Keith from online. In terms of, you know, Luca talked a little bit about, sort of, you know, both the promise of DPR, but also the pragmatism associated with it, and the fact that. you know, there’s a very positive example for Brazil, but I know from your research in South Africa, you know, there has also been a different set of results. So, I’m curious, Keith, if you could share a little bit more about what your experience has been, you know, from the old age of IDs, digital IDs, and now, you know, the new shiny DPI, what some of those learnings and experiences has been from South Africa, and perhaps what are some of the political and socioeconomic contexts within which DPIs are existing in the global majority?

Keith Breckenridge: Thank you. Yeah, it is very interesting. Thanks very much for the invitation. Yeah, so DPI has come to South Africa through the G20 and the Gates Foundation and Operation Bulentlela. It’s coming in from the top of the bureaucracy, but many of the things that people think about as being kind of benefits of DPI have existed in South Africa for a long time, as they have in other places. I’m not suggesting we’re the only ones who’ve had that. So, we’ve had a digital system for welfare distribution since the early 1990s, in fact, a little bit older than that, and a biometric population register that covers everybody, at least, I can say it’s been in place for about 20 years, and both of these things interact with the payment system. They do not work in real time. There are all sorts of kind of clunky bits and pieces of it that don’t work well. I will say DPI is a big improvement on the plan that was in place in South Africa three or four years ago, which was one aimed at moving the population register into the security cluster. So, it was all about basically targeting people who were, in the officials’ minds, were illegally in South Africa. So, migrants, in particular, were going to be hunted down using high-resolution ID photographs and a CCTV network, all of this done with the support of the contracting companies. I won’t name them here, but they were selling that as a kind of resource to the bureaucracy. So, DPI, comparatively, wonderful. I mean, it’s a much… better system, a much better plan. And I think there are lots of things that DPI can do that would be excellent in the plans at least especially the UPI plans. We see space for enabling intermediaries for example letting people act as what we would call fiduciaries. And there’s possibility to build that out so that in fact you can. It’s not a question of very under equipped poor people confronting a powerful state and not being able to get any answers or remedies at the moment. But but the big kind of takeaway benefit that people keep pushing for DPI which is really about access and financial inclusion. I mean those things have been in place in South Africa for a long time for generations. There’s no big payoff from that area. I’m really skeptical that there are benefits for poor people a real time settlement. I think there’s actually this risk there. I don’t think it’s a benefit. And then so. And I think we really have to ask some hard questions about financial inclusion. I’m sorry. I can see. I mean I do understand. If you look at my work you’ll see a lot of benefits to this. But we have to ask what’s happening now. And financial inclusion is not doing what people promise for it. It is again something that we’ve seen for a while. We’re at least a decade into mass financial inclusion. It’s encouraging usurious lending. There’s no asset creation. People are borrowing at 1 percent a day and they’re being kind of pushed into a blacklisting engine that says if you don’t pay we’re going to block you out of ever accessing the formal credit system. The biggest problem in South Africa people don’t want to talk about is online gambling. So now people accessing the payment system using the gambling apps. It’s truly astonishing. One trillion Rand was spent on online gambling systems in 2023. That’s a third of GDP one third of GDP spent on gambling. And the tax collector is kind of complicit in this process because it’s like a big chunk of that and put it into. So it’s like a hidden digital tax. And there are lots of risks for families here, too, that people, well, you know, but people don’t want to talk about it while we’re designing and talking about designing a more and more expansive system to make people visible on these networks. So there’s backend hacking, people breaking into, this happens in South Africa. I’ve yet to see what it looks like anywhere else. So when people break in, they steal the encryption codes and they can essentially help themselves to people’s online accounts. Lots of examples of that. We have lots of examples, as everybody else does, of what we call pig butchering, people being kind of, you know, seduced into giving away their money to, sometimes members of the family, but sometimes people they meet online. And all of that worry, that kind of fear of the dangers of moving the poor into the payment system was before the Trump administration, right? Which has now seen a wholesale withdrawal from the regulatory state. And you know what this looks like, I’m not going to repeat it, but, you know, the key institutions, the SEC, the Consumer Financial Protection Bureau, and probably more important than anything else for us was this Foreign Corrupt Practices Act, which made companies that behave badly, literally hand themselves into the Department of Justice in the United States. And that’s a wide range. There are dozens of companies that did this, including some famous examples from Brazil. And that kind of regulatory, whatever one calls it, context, has disappeared. All right, so if we don’t ignore all of these problems, what should we be doing? That’s, I suppose, the question. And the fundamental issue really has to be making what we call metrics of exclusion much more visible. So while we’re talking about DPI, what happens with DPI is we make credit scores visible. That’s the big driver. So companies can now build either their own system or they can share data. And it puts individuals in a, you know, truthfully, I’m not sure they’re the beneficiaries of this, I’m not sure. sure that financial inclusion is not much evidence. There’s a lot of good evidence that in fact the overwhelming effect of a kind of mass digital lending system is is actually increasing poverty. It’s not fixing the problem. So exclusion and metrics and publishing that should be immediately on the Web site. It should be. This is the number of people whose identity numbers are at the moment disabled. And this is how long it’s taken for us to basically answer queries around those numbers. I can tell you what this looks like in South Africa. There was a court case last year in January where the Department of Home Affairs admitted that two million identity numbers had been turned off turning off. And in our current system just involves editing literally a comment field in on the third page of the population register. And that disables the identity number from functioning. It means you cannot access the birth of your child. Can’t bury your parents on access to bank accounts. You cannot function digitally. All these Estonian capacities people like to talk about. They’re very easily switched off in the current system. And that needs to be much more carefully regulated. So officials who edit the population register they need to be subject to exactly the same kind of oversight as the as the people who are being essentially identified by. So and then I’d say the third big issue is making space for what we call trusted fiduciaries. I think it has to be more than lawyers. These have to be officials almost or semi-private officials who are self-regulated in the first instance who can interact with this DPI engine. The nicest example of how this actually works in South Africa are the undertakers. So that you know we have excellent death registration and very quick well handled death documentation that when someone dies they get very quickly documented. And the reason is that undertakers earn a fortune from putting bodies in the ground and they’ve been given granted access to the population register in order. it to speed up that whole process. Now we need something like that that helps when you know you go to get an identity documents and you see that the population registers misspelled your name mischaracterize your gender. You know there are so many mistakes. Once you start once you’ve lived with a big powerful system like we have for generations the areas become they sort of lurk in the databases. And as soon as you build a nice clever joined up into operable system you begin to see the dangers of that for all sorts of things. You now can’t access health care because you somehow fall in foul of a maternity suit or paternity suit in the in another area of the state’s databases.

Sabhanaz Rashid Diya: Thanks Keith. I think those are some interesting things I’ve thought for us. And I think lots for also the panel like you know I will definitely come back to you with some questions. Maybe I’ll try and turn to 3D now in terms of I think you know Luca and Keith gave very different starting points for DPI in terms of you know what the political social legacy set of context was in that space. And I’m curious. And I think some of the points that were raised really was around sovereignty around security and governance around transparency access to data in order to kind of build robust systems. So I’m curious from your research and your experience what do you think are some of the prerequisites as far as regulation and governance is concerned to really build a people centered DPI. Thanks.

Smriti Parsheera: They are such a pleasure to be a part of such a stellar panel. Let me just begin by introducing you know the organization I’m with Intelligent Foundation in case people don’t know of them. It’s a nonprofit that works on financial inclusion and open payments. But I should caveat the views I’m expressing here are my own as a researcher and civil society participant from India. And I’m going to try and address that and also speak to some of the points that were made earlier by you know first thing what is the good of the moment that we are in. This momentum that we are seeing around DPI as Keith said not all of it is new. Not all these interventions are new. But what is new is the firepower that’s going. I think that’s a good direction to be going in, but I want to add some observations around, you know, what could be done better, and I think the first thing I want to bring in is, you know, if you think about the participants in this ecosystem broadly for the state, the markets, the civil society, and the actual public, who is the, you know, the beneficiary of this, you know, this ecosystem, who is the, you know, the driver of this ecosystem, who is the, you know, the, you know, the, you know, the, you know, the, you know, the, you know, the beneficiary, ultimately, of, supposed to be the beneficiary of all of these developments, and there is this very strong state-market alliance that we are seeing, both in terms of the implementation of DPI, in terms of how it’s being developed, how it’s being rolled out, but also in terms of the philosophy behind DPI, right? So if, in the Indian context, we often see the entities which are developing DPI being called as a startup in the government, right? So there is this valorization of the idea of startup, of the idea of markets, Aadhaar, which is our digital ID program, has been called as a poverty killer app, you know? So a lot of phrases, a lot of thinking, philosophy, coming from this idea, valorization of the way, you know, the Silicon Valley has developed, but also, as Luca said, it’s a counter to the way Silicon Valley has developed and, you know, proliferated in all over the Global South, in particular. So DPI is, in that sense, both a counter, but also runs the risk of following some of that same playbook, and I think that’s the first thing we need to be a little worried about. And in my work, I, you know, use this concept of an alt big tech. So in the process of being a pushback against big tech, could we be creating another ecosystem which is pursuing scale in the same reckless manner, which is, you know, really blind to due process, really blind to other things in the pursuit of reaching scale fast? So I think those are things to be conscious about in the DPI story. The second thing I want to focus on is when we talk about the public in DPI, there are many, you know, interpretations, as Luca said, is it public in terms of ownership? Is it? public in terms of the deployment landscape, but I think the really important question is, is it public in terms of the participation? And there again, the question is, when we are thinking about, you know, the Global South context, very often we are in ecosystems where, you know, things like regulatory impact assessment, cost-benefit analysis are not part of the day-to-day culture, and there are exceptions like the TRAI in India, which has this very robust culture of transparency and accountability in its regulation-making process, but that’s unfortunately not something which trickles down to all agencies, and it’s certainly not something we’ve seen in the deployment of the DPI process, so it has been very much a top-down sort of approach, and that’s something to think about. The linked point of that is really, you know, the role of civil society in all of this, and I think the fundamental question of who is civil society, right? Usually you think of civil society by definition as something which is, you know, separate from the state and markets, it’s a voluntary association of people who assemble around certain values, but those values could be very different. The values could be about, you know, that the route for empowerment is through market-based mechanism, and that’s one value, or the value could be that you need to push for more human rights and think about, that’s another value, and I think it’s open for debate which value is better than the other, but it is important to have space for both these values reflected in the platform, and sometimes it does happen that, you know, certain allied values find a space in internal debates, but values which are not aligned either have to resort to the media or have to resort to courts to find a voice, so I think that’s an important conversation to shift that and really bring in some, you know, first principles, accountability, participation. And the final point I want to make from the Indian experience, I think I’ve been a bit dull and negative in what I’ve said, but I want to bring in a point about the value of social construction. Right, so much of what we are thinking about DPI from different perspectives is about techno-determinism. There are some people who think, like, this is the path to, you know, to kind of empowerment, to inclusion. There are some people who think this is the path, a sure path to a… So, there is a counter view that there is a process of social construction that is happening in the DPI deployment. I’ll give the example of Aadhaar in the Indian context. So, when it was first deployed, the idea from the agencies was that this would just be a number, this would not be a card, right? But when society accepted it, they said, you know, we actually need a card, we are used to a card as a form of identity, and the, you know, over time the way it was accepted by society and by the authority was that now it is accepted as a form of card. Similarly, several other features like, you know, initially, whenever you would use your Aadhaar, you would use those digits and, you know, that particular number, but over time a virtual version of it has emerged as a part of the pushback, as a part of the questions by, you know, before the court, the idea that, you know, can you have a masked ID where instead of revealing your main ID everywhere, you just reveal this virtual ID. So, there has been positive social construction which has taken place, but again, coming back to the question of participation, very often, you know, these are brought as new mechanisms, but because they are not originating from the agency itself, they’re coming back as a, you know, as a pushback, as a way to accommodate. Sometimes they don’t become the mainstream solution, and I think that’s where we need to go in terms of DPI in recognizing that there is a lot of, just from experience, from your own country, from each other’s country, so as others do, you know, digital ID, I think it would be valuable to look at the default as virtual ID as opposed to the main ID, to learn from each other’s experience, and I think that would be a positive way to proceed with DPI.

Sabhanaz Rashid Diya: Thanks, Smriti . Lots of little nuggets for me to pick up on later. Maybe I’m going to now turn to Bidisha, as you’re coming online, and I think I want to double-click on the question of ownership. I think we just heard different takes on it, you know, in terms of what Brazil’s experience has been in terms of, you know, these being owned by truly public agencies, and therefore the Right to Information Act and FOIAs kind of play in that space. We are hearing sort of a mixed bag of reactions in terms of, you know, DPIs elsewhere, where maybe the public agency is not fully in charge. in terms are fully present in terms of the ownership of the system. And I think just to build on that, DPS are also highly modular systems, right? It’s not just one big technology. There’s multiple parts to it, multiple stakeholders at play. And to Sridhi’s point, really this conversion from, you know, are we actually creating another monopoly in order to tackle, you know, sort of the existing monopoly, right? So how do we think about monopolization and ownership in this space as well? So I’m curious if you could speak a little bit more to that in terms of, you know, your research, how you’re thinking about, how do we truly make DPI a public interest technology and a public facing technology? And how should policymakers, civil society and other stakeholders be thinking about ownership within the DPI architecture?

Bidisha Chaudhury: Thank you, Diya. And thank you everyone for, I think they’ve all made excellent points. I don’t know how much new things I can add to the points that have already been made. But just to respond to the question of ownership and publicness in DPI, I want to underline, at least in the Indian context, the role, very important role of the state in, you know, in conceptualizing and implementing DPI. And not just building on Aadhaar, but if you look at UPI and then other efforts, even though there has been very, very close partnership between the state and the market, which is also not new in the Indian context and elsewhere state market partnership has always been there. But particularly in case of DPIs, I think there are three specific roles that the Indian state has played. And they have been very clear in owning this DPI, even though there has been considerable amount of effort by, I would say, the software industry partnership in India. And I see these three roles as creating an institutional space, like even the creation of NPCI to, you know, push forward. forward the agenda of financial innovation or digitalization of financial products. The state and the RBI has played a very important role in creating that kind of an institutional space where the DPI-like innovation can take place. The second has been in implementation of DPIs or pushing for implementation of DPI and specifically scaling this up. Came through very, very clear state policy. So Aadhaar was mandated for all welfare, right? So this cannot happen. This kind of scale cannot be achieved without the intervention of the state. The same we see for UPI that all, you know, the fact that all welfare accounts needs to be connected to a mobile phone number and a zero balance bank account. That created that whole scale of public who will now be the market for financial innovation or digital financial inclusion. So here the state played very important role for scaling up DPI or even creating the possibility of scaling up something like an UPI. And the other role that the state plays is regulatory role that how to regulate different, this multi-stakeholder ecosystem that we have, how to regulate each of these stakeholders. But having said that, so I think in the case of India, the state has been very much there in terms of subsidizing the infrastructure on which these are built in regulating these infrastructural spaces and stakeholder and also scaling up this infrastructure to, you know, much of the Indian population, let’s say. But I think the word of caution that I would just put here that while the state does play an important role and I think that should play an important role, especially in a global majority, where the role of the state. in bringing in development, bringing in inclusion is more, should be more explicitly realized. But the caution here, I’ll just take the example of UPI in this case. So one of the things that came out of our research is that when the state is regulating and when you are relying on a very, very global digital infrastructure or global digital platform, there is always, there is already very, very big players. So I understand the DPI is kind of thought as an alternative to big tech, this question of digital sovereignty. But to what extent the big tech can, like you can bring, create an alternative infrastructure. And here I mean the software and the technological infrastructure, which is truly independent of the big tech is something we need to really explore and ask more carefully. For example, one of the things that came up in our study of UPI in India is that even though it is seen as something which is owned by the state, to some extent, subsidized by the state, scaled by the state, the major traffic that is, that controls the UPI ecosystem, or at least the volume of transaction, is GPA. So much so that many of the vendors, small scale vendors that we interview, they don’t even think of UPI as a public infrastructure. They kind of synonymously think of it as a Google infrastructure. So, you know, even without knowing much about it, because that’s what they use nonstop, you know, constantly, like GPay, PhonePay, PhonePay is a Walmart subsidiary. So they call this infrastructure as GPay, or as PhonePay. So they don’t even think of it as an UPI which is owned by the state, or maybe not owned by, maintained and sustained by the state. So I think that is where it is, it becomes very critical to understand, or even explore that how do we actually circumvent. the influence of the big tech when we rely on an infrastructure where the big tech interests are so much more embedded already. I think so that’s a cautionary tale with which I will stop.

Sabhanaz Rashid Diya: Thanks, Bidisha. I think we’ve heard quite a bit from our non-state stakeholders, if I may, as the term of the multilateral space goes, in terms of the various iterations of DPI, questions around ownership, questions around sovereignty, some positive, some sort of cautionary tale. So it’s been quite a mixed bag, and I think that’s really great because I think it starts demonstrating that not all DPI is the same. It is supposed to be different, and it is in this monolithic technology that just gets replicated around the world, and I think that’s a really important piece of that puzzle. But I will now turn to our government stakeholders, perhaps, in terms of you’ve heard quite a bit from the various panelists here, and maybe I’ll start with you, Mr. Singh, in terms of India has come up quite a bit in this conversation, as it should. I think India has been such a leader in the DPI space in terms of really setting that mark, demonstrating how it’s possible for a global majority country to pull something like that together, and a very deliberate role of the state in terms of tackling issues of inclusion, of whether it’s financial inclusion, whether it is identity inclusion, whether it is socioeconomic inclusion. There have been multiple layers of that, but we’ve also heard these concerns around ownership, these questions around what is really the role of the state? How does the state work with the broader software ecosystem? And perhaps, what are some of the regulatory considerations as the state thinks about scaling this technology? So maybe I’ll turn to you to respond a bit to that, and just one add-on to that question is I know while the DPI infrastructure is a very national, domestic play, India has also been leading in terms of making. need global, right? So I’m curious also as you’re thinking through your response, how does India see sort of its learnings nationally play out in terms of the global scaling of the technology, and particularly some of the principles that it carries forward?

Sheo Bhadra Singh: Thank you, dear. I would just like to start with the problem statement, what actually we are looking at to solve. Because a lot of the times, we go directly to the issues related to privacy and secrecy and all the things. But the problem, and particularly in India, what we were actually looking at is a large section of population which was not having access to the public services. So this was the main driving theme. And when we embarked on this journey of DPI, I think most of us were not aware that this will be named as DPI at all. So because Aadhaar came way back in, I think, 2009, 2010. And that actually laid the foundation of the present-day DPI is what we are having. So the basic requirement to give public services down the pyramid to everybody, access to the public services, was the basic requirement through which this journey started. And once we had the Aadhaar digital identity, and in the present case, we are at almost 1.33 billion Indians are having the Aadhaar digital identity. And that forms the basis of all our services, which are extended to. And once these services are available, it is being used by not only those who were earlier not basically accessing these services. It is being used by the, it has become a citizen-centric. services now. So when we embarked, now we are a lot of in the world it is being discussed actually what DPI is and what should be the foundational principles of these DPI has to be. But as far as we are concerned, we have our own foundational principles when we embarked our journey. And let me say that as we negotiate, as we travel, as we implement, as we learn, we always have some aspects which we change and introduce in our systems. So nothing becomes you can say that it is not like things being cast in stone. Based on our experiences, there has been some judicial pronouncements also. And based on those, we make changes in our systems and we try to be transparent. So the basic principles which we actually have in our systems, I would just like to go through maybe in case of taking the risk of repetition also that we have certain principles like eliminate economic, technical and social barriers. Then the system has to be modular and scalable. Keep privacy enhancing technologies and security features within the core design to ensure individual privacy, data protection and resilience based on standards. And then we have laws, regulations and policies should ensure that these systems are transparently governed and promote competition. So these are the basic principles through which we have been building our DPIs. The government has been the major source of inspiration, major source of finance also in these cases. But I’ll not say that it is basically driven by government. There are a lot of individuals, a lot of organizations who are participating. and helping us build the DPIs. Now once we have this DPIs in place, I just wanted, because we have an experience of already implementing and achieving the scale, which we can showcase to the world at large, and what are the changes it actually makes to the society. I’d just like to give some details of that. But before that, once the Aadhaar was in place, digital identity was in place, we had a challenge of how to provide the financial inclusion to the large section of the society. Most of the Indians, maybe about 10 years back, they were not having bank accounts also. So we went through this digital identity of Aadhaar, and in India, we open an account maybe in less than half a minute. The KYC is being done through Aadhaar, and an account is opened, and that account gets mostly operated through a mobile. So there was in parallel the 4G was being implemented all over India. Mobile phones and the broadband was actually proliferating at a very fast pace, and digital identity, people got, the accounts were being opened, and in the Jan Dhan account, what we call, is the zero balance account which is being opened. The present count is 550 million. So this is the number, so these people who are basically out of any financial inclusion, they got a platform, they got, it is just. a mind-boggling figures, which I am actually quoting, and so many people were out of the banking system. And just because we had these DPIs, we had these platforms, the financial inclusion of large proportion of our population actually happened. Once they were part of the banking system, there were a lot of benefits government was giving, which came to their accounts, money was getting transferred to their accounts, right from building toilets in the villages to building houses in the villages, and pensions were being given. So the experience of India was not very good in giving these benefits to the people because a lot of the proportion of it was being siphoned off so the direct benefit transfer actually came into being. And then UPI happened. UPI happened much at a later date while we had an Aadhaar identity, we had this Jan Dhan account, and what UPI did was that the transaction through the mobile bank happened at a very fast pace. So people adopted, they had an account, so they were able to transact. So these things actually happened, and now, in fact, the last month I am having the details, 18.77 billion transactions were $290 billion. That is only for the month of 2025. The adoption of UPI has been so fast because we had a system of accounts, we had a system of digital identity already in place. And then now India is actually building upon all these advantages. There are three or four examples which I just need to quote to give the context to our audiences today. that we have a GEM, that is Government E-Market Place, where most of you may be aware that the state, federal, and the state governments is one of the biggest, does one of the biggest procurement of services and goods. In our case, there has been a DPI built on that, which we call GEM, and the facilities of procurement of goods and services to all the state governments and the federal government actually happens through that. And if I just give the figures, there are 11,000 products, 336 services, 1,64,000 buyers, and for last year, that is from the April of 2024 to March of 2025, there was a procurement of $625 billion through that. So, and what has happened is, actually before that, there was a small group of suppliers. They had actually controlled the market of procurement, the government was procuring through them. It has basically democratized the suppliers. It has enabled a person sitting in a village, a person sitting there in the remote part of the country to supply, to become a supplier to the government. This is how, actually, the transition is happening. Then we, I don’t know how many of you would be aware of Open Network of Digital Commerce, ONDC, which is being implemented. We have seen, actually, the platforms, which are basically, which provide e-commerce, have not been very democratic in nature. There are issues with small and medium enterprises to actually come on board on them. So, this ONDC also is being made for actually, it doesn’t. So, what we are doing is we are providing a platform for small and medium-sized enterprises to participate in the e-commerce activities. The government actually stops at or it doesn’t say that larger ones cannot come to our platform, but it actually enables small and medium-sized enterprises to actually, it enables them to participate in the e-commerce activity. We have in fact a very popular platform for small and medium-sized enterprises. We have a very large number of small and medium-sized enterprises in the common public, and the number is also, I would like to quote, is 787 million. And there are 409,000 health facilities which have been registered, and 648,000 health professionals. So, what it has done, that you have an animation, and you download to it, regular animation, and the hospitals are registered by the government, health professionals are registered by the government, you go to all these hospitals which are on board, show your card, give your number, and you get the facilities, and the payment is done, the government does the payment for that. So, it’s a very, very, very, very, very, very, very, very difficult thing to do for a human citizen, and earlier, if you just imagine the numbers, and imagine whether it can be done without a DPI in place, the answer is absolutely no. So, in India, we had no option if we have to, the time was very short, services to be provided to the citizens was a very, very, very, very, very, very, very, very, very, very, very difficult. Whether we could have done it without DPI, the answer is no. Whether we could have waited for the definitions to be finalized, the answer is no. We have embarked on our journey, we have achieved a lot, we still learn, and we and based on, there is another aspect, in fact, in our G20 presidency. We had included this DPI as one of the pillars and it was accepted, with certain changes which was made during the negotiations. And I think that was the time when the world actually woke up to the DPIs and started discussing about them. Although a lot of things were happening at other places, but now it is the time for all of us to sit together and to see that there has been a lot of experiences in other places also, how we can learn from them. And as far as India is concerned, we have offered free of cost to any country which wants to adopt DPI platform. Our prime minister has offered this during the G20 presidency, and it can be adopted by any country and we’ll be very happy to actually help that. Thank you.

Sabhanaz Rashid Diya: I think quite a bit of background from what India’s sort of problem statement was and how it kind of approached that larger ecosystem. And I think one of the things that was also interesting, I think India’s G20 presidency also came right after the pandemic, when the world also felt that if governments could not deliver public services and we were in a situation where physical infrastructures were no longer functional, then this was really the way forward. So I think many things kind of tied together and really helped propel DPI into the spotlight. I’m gonna turn to Armando, who has joined us online from the Dominican Republic. And I know it was very interesting to hear from Mr. Singh in terms of what India’s sort of problem statement was, what it was really trying to solve, and sort of the long journey it went through in terms of various iterations of what we now today know as DPI. I’m curious, Armando, as some, you know, leading a lot of the innovation that’s happening in the Dominican Republic in terms of reimagining government services, reimagining public-to-citizen relationships, and how do we sort of rethink digital transformation. So maybe you could share a bit more about what your experience has been in the DPI space, and also just building on the previous question, how is sort of the national priorities tying up to the more universal principles around DPI, and how is sort of Dominican Republic adopting some of those?

Armando Manzueta: Well, first of all, thank you very much for inviting me to this amazing panel. The conversation has been very rich, and I’m more than eager to share a bit of the work that we’re doing so far in the DR. For the Dominican Republic, the DPI as a concept is something that is fairly new. We basically started back in 2020 when we were thinking about how to solve societal-level problems that are tied to public services, especially when it comes to data, interoperability, and identity, and of course payments. So in that regard, for us, DPI is not just a set of technologies, but a nation-building tool, and one that must be grounded in rights, resilience, and trust. In that regard, basically there are two realities that are driving our adoption, and that’s one which is socially, since we’re responding to citizen demands for accessible and inclusive services that respects their dignity and privacy. Politically, in the sense that DPI enables a more transparent participatory result in the government, and the state of course, and economically, because for us it’s a letter for inclusion that is also unlocking opportunities for not just for the people on their feet, but also for small businesses, young women, and especially people that are in underserved areas. But we realized that we can only fulfill our promise to the people when the foundations are right, and for us that means doing a major groundwork for having a strong legal and regulatory frameworks around data protection. interoperability, and open standards, for us to establish institutions that ensure oversight and safeguard implementation, and of course, having institutions that are involved, especially from the third sector, the civil society, private sector, that the one way or another has some sort of shared responsibility on insight and oversight of the work the government is doing on that front. So, from the very beginning, we started first with the most tackling issue for us, which is data interoperability. So, we decided to build our own infrastructure and data exchange based on the work that the government of Estonia has done before on implementing this broad open source platform called XROAD for data exchange. We did local customization and implementation, and we are basically translating all the documentation into Spanish for institutions, and eventually private sector, to understand how it works and get on board on that. And from that, since we started implementing it in 2021, 2022, we probably have around 71, 72 institutions that are currently using the XROAD ecosystem for data exchange. And the result has been very interesting so far. What we face now, another issue, which is, of course, data governance. And now we’re rethinking on how data governance actually works, not just for public services provision, but also how we can empower people on that front. And in that sense, we’re working on a new set of regulations around that principle. One that has focused not just on the data protection itself, but also on the quality of the data that is actually being exchanged. So, we’re working on defining proper data observability and orchestration method mechanisms, so we can actually make some of the principles around data protection something that should be easy and forcible, because sometimes these concepts can be very abstract and very difficult to grasp by institutions. analyze on a technical level. So in that front, defining this proper set of data rules and having the right platforms alongside the data exchange mechanism, it’s very important, especially for them, to be secure and to respect people’s rights on that front. And of course, having a way for this oversight mechanism that should be implemented from the civil society so they can keep the government accountable on the work that is doing on that front. When it comes for identity, basically what we started doing, since the government in the Democratic Republic has the peculiar characteristic that it doesn’t administrate directly the civil registry, although it’s a public institution, but it’s outside of the government. We have made several workarounds based on the work that we previously have done with data exchange. So what we decided was to build some sort of authentication platform for starting for public services that builds on the same principles of data protection, because it’s mostly focused on data minimization and the rights of the people to authorize what kind of personal data it can actually share to provide the minimum requirement for public services. So what we’re doing right now is basically working closely with several institutions, such as the CDPI, on defining a broader architecture for this data identity infrastructure alongside the civil registry authority for us to start providing different means of not just of data protection around identity, but also how we can actually start building this sort of verifiable credential standard that works for everybody and to start having these sort of documents that are actually verifiable, trustworthy, and of course, administered by the right authorities at the right time. And for payments, I think this is the area that we, somehow we have huge advancements because we have like real-time payments that are. actually enabled and effective and available for the universal population. But the way that it’s done is not necessarily the most efficient way. So the central bank, which is the owner of the ad infrastructure, is actually looking at ways to provide a pixel-like solution for the people that is also based on those principles that we’re trying to build. And of course, trying to have this sort of accountability around the society and also getting the private sector on board. So that solution, when it comes available to the people, actually gets adopted very easily and quickly, especially since the country has serious issues when it comes to financial inclusion. Because roughly half, for example, the population is financially excluded. And regarding the other half that is technically included, probably 30% are actually the ones that are receiving broader financial services, whether that’s digital or not. And we are finding ways to reverse that situation. And of course, having these sort of platforms will allow, especially, to do just that. Not necessarily in the way that we have solved the problem in Brazil, but it’s one of the options that are actually on the table. And we hope that we’re going to find ways to do this in the most inclusive way, but also having this participation for oversight and accountability for all the actors that must be involved in that part.

Sabhanaz Rashid Diya: Thanks, Armando. I think kind of really seeing that sort of spectrum of experiences, right? Sort of, you know, kind of India’s very long journey and now kind of Dominican Republic tackling some of the more emerging challenges in this space. I will come to audience questions right after we bring Rasmus in. So if you have a burning question, now is the time to start thinking about it so that you can ask it in a few minutes. Rasmus, maybe turning to you. Estonia is sort of the other major example in the DPI space. And similar to India, I think it has a very long legacy of really digitizing e-government services. digital, but thinking about e-government services, public to citizen delivery, and solving some of the problems that perhaps Dominican Republic is going through right now, and the whole world is going to know. So think about, how should we think about DPA? What is the right governance model? What are the different ways in which we can really cut down some of the costs of it? So there’s many questions on the table, but I would love to know from you, what has Estonia’s experience been in this space, one. But also just to build on that, Estonia is also leading on the rights respecting DPA principles through the Freedom Online Coalition. And so maybe if you can also tie in how Estonia has actually implemented some of those principles in its past, or how it has taught to those principles, as it is now thinking about either its legacy DPA, or sort of emerging DPA iterations in its work.

Rasmus Lumi: Thank you very much. Well, maybe I should start by saying that when in the beginning, when you introduced me as being on the extreme left, I just wanted to clarify that it doesn’t necessarily reflect my political views. Anyway, I think many great angles and aspects have been raised, and the questions you are posing to me now also probably would allow for hours of discussion. But of course, I will stick to just some of the main things for now, which would probably be the fundamentals of the DPA, because we have already heard a lot about the practical backgrounds of why DPA is being built, and so on. So maybe I will get also to the Freedom Online Coalition’s principles of the DPA, as Estonia is in 2025 the chair of the Freedom Online Coalition. But as a representative of Estonia, I I definitely can give a bit of a background as to how we have built it up over the course of for now I think around 25 years, and how we have managed to tie the fundamentals into the actual digital solutions. So as I said, Estonia I think started with its digital public infrastructure in the beginning of I believe 2000 or so, around 25 years ago, and this has really allowed us to see over the years how this digital public infrastructure and all this digital governance has been able to transform the society, and how it has empowered the citizens and how it has fostered economic growth and so on. At the same time of course we have always been very mindful and vigilant about the risks that this has also brought forward, and we have always tried to mitigate those risks to the best of our knowledge at every certain time or time period when something has been built. Now many of the principles that already have also been mentioned here during previous interventions, I mean these principles seem to be self-evident, but they don’t necessarily self-evidently work within the DPI systems, and there also can be very different understandings of what does it mean for example privacy or security within the digital public infrastructure. I will bring you one example which would be that for example when we talk about privacy then within the DPI, does that mean that the person’s data or the information about the person is private to the outside users? Or is it also private on the inside of the DPI, meaning that is there any control or checks and balances on how, so to speak, the owner of the DPI or the government, how they can access this information about the person? I think these are two different concepts, and that’s why, for example, what we have done in Estonia is that one of the very key principles for our DPI is that every citizen is the owner of their own data and has full control of it. In practice, it means that any user, any Estonian citizen can at any time verify which government official or any other person has accessed their personal data within the DPI systems. So this is our way of trying to make sure that there is preventive accountability for the use of personal data within the DPI. Now, in practice, there are also some other important elements which we have prioritized from the very beginning, and one, of course, would be interoperability, meaning that Estonia’s digital public infrastructure works on a platform which is called Crossroad, and the point is that this is a platform that allows for all different institutions to manage their own systems, but they’re all interconnected on the same platform, allowing for the citizens. and to actually be able to access everything from one place. And as I said, already being able to control their own data. Then what was also mentioned and what we have already done since the very beginning is the e-identity. And this is the key or the central part of the whole system, which allows for the people to actually access all the services. And also, in the case of Estonia, it has already also been taken so far as to be able to participate electronically in the elections and so on. The so-called iVoting, Internet voting, we can do it on the computer as well. At the same time, we have been trying to maximize the possibility for people to access all government services, or I don’t know if they can be called necessarily government services but public services, to close in on having 100% of the services being able to be accessed by people. And I think, not sure, but I think divorce is now, I think, the only thing that we cannot do online yet. But I think as the society develops, then that will be possible as well. But now, not to waste too much time, just I would run very quickly through the Freedom Online Coalition’s work currently, which is for this year in the program of action. Estonia, as the chair, has prioritized rights respecting DPI principles, which obviously would be something that we have already naturally kind of built into our digital public infrastructure. And as the chair of the Freedom Online Coalition, we are hoping that the coalition could… could bring this to the global attention as well. Many of those principles, as I said before, are quite natural. There is one thing that is extremely natural, but you cannot write it into anything. It’s not a standard, and you cannot even put it as a principle, and this is called trust. The public trust is something that is actually the most important thing. You can force people to do anything, but without trust, you will not get the most out of whatever you’re building, because people will not want to use it. But anyway, so trust we have not been able to include into the principles, but what we have included in the principles are the notions of human-centered solutions, meaning that everything has to be focused on humans when we develop technology or digital solutions. Everything must serve this objective. Then something very important, which is inclusivity, and we are feeling very strongly about everybody’s need to have access to the Internet, to all the digital services and skills, in order to be able to benefit from digital transformation. Then we believe that international human rights standards should be followed, because otherwise there are too many risks that we will immediately take in undermining the possibility of trust in these discussions. Transparency and accountability, I think, speak for themselves. Everybody probably realizes that we need this to minimize the negative impact of technologies on human rights. Privacy and security, we have already talked about. Then societal context. I think everybody agrees that their own societies, their own communities need to be respected with their own particular – and so on and this is also very important to make any digital public infrastructure work effectively to take into account a societal context. Then sustainability and resilience and this is because it is important to build sustainable programs and to be able to maximize long-term impact. Data-driven. DPI must be data-driven because we want to ensure that quality information is available for decision-making. Interoperability that I mentioned in our case as well we do want to promote interoperable systems which work together seamlessly across organizations. It will make everything cheaper, it will make everything more effective, it will be easier for both the users and the administrators or managers of the systems. Then of course technology neutrality because we we believe that the digital public infrastructure should be built in a technology neutral manner. Openness of course open approach to digital development. It means open standards, open source, open data. This is something that will nicely tie in with privacy and security. And finally of course collaboration meaning both collaboration domestically and and also on the international level. So these are the principles that Freedom Online Coalition is currently working on. We hope to be able to adopt them during this year and then they may be able to serve as some like a guiding light to everybody who wishes to further develop their own digital public infrastructure. Thank you.

Sabhanaz Rashid Diya: Thanks Rasmus. A very comprehensive set of principles I think for all of our stakeholders to think through. We’ve reached we don’t have a lot of time for our panel so I’m curious if anybody in the audience here has any questions they would like to ask. the panel. There are mics there, but we’re also a very intimate room, so also I think we can hear if people are speaking. Or maybe if online also. Yeah, we actually have an online question, but I wanted to make sure. Okay, great. I think we have an online question. I think Rasmus touched on it a little bit more, so maybe, oh, okay, yes. Thank you. You should take the mic. Yeah, maybe the mic will be useful.

Audience: Thank you very much for this insightful question. My name is Elizabeth. My first, I have two questions, if you don’t mind. My first question is to Professor Belli. Thank you very much for the insights around, like, collaboration between government and other stakeholders. I had worked with the government of Nigeria on digital policy reforms, one of which was digitizing business registration processes to reduce the time, the cost, and the procedure of doing business. And my experience with public service is there’s still that gap in terms of getting that sort of entrepreneurial mindset, in seeing how this could be of great benefit, and then fully owning this service, service delivery to the public. So I now currently work with a mobile industry association, and the big question is, for telcos, mobile network operators, what are, like, the opportunities in terms of collaborating with government? Because there’s this buzz around mobile data. Increasingly, mobile devices are becoming very instrumental in increasing access to public services, to digital services across the world, especially in emerging markets. So I don’t know if you could share some examples and some insights on that, that would be great. The second question I have is to Professor Breckenridge. out the session in South Africa and also to Simriti. I’m so sorry. I referred to it properly. It’s on digital financial inclusion. So the comments around really thinking through on how we can guard against digital financial inclusion not being a risk more or less an innovation addressing the problem as opposed to being a risk for misuse, abuse, data harvesting and the likes. And it got me thinking around AI ethics and transparency. And my question is to Professor Breckenridge in South Africa. What is, what would you say are like top priorities in terms of building this very strong AI ethical standards around that? And for Simriti, I’m sorry once again, is in the context of digital financial inclusion. Thank you very much.

Sabhanaz Rashid Diya: Thank you. Before I go to the panelists to respond to this question, we also have some online questions along the same lines. Maybe I can tie the two questions together just in the interest of time. I think we also have an online question that is also asking similar sort of approach as to how do we actually weave in some of the human rights principles within the DPI architecture. And then particularly just speaking on the AI piece, you know, we are now also seeing a moment where using the existing DPI, a lot of AI developments are happening, a lot of API access is being provided. And so that architecture is also scaling some of the AI expansion particularly in the global majority. So as we think about these expansions, how do we actually weave in some of these principles that Rasmus mentioned but also some of the learnings that we have had in the past? And where do the different stakeholders that we have, whether civil society, technical communities, what could be their different roles be? So that’s a question from the online audience and maybe it’ll be a good way to kind of weave them in. So maybe Luca, since we can start with you and then I can maybe turn to some of the other speakers.

Luca Belli: Yeah, I think those questions are excellent and I So, I’m going to talk a little bit about what we are doing in the digital world and what we are doing in the digital world. And I think this actually allows me to delve a little bit more into the details of what could be the success or the failure of this kind of efforts. I think as anything in digital, one has to consider that we are dealing with systems, right? There are a lot of elements and a lot of layers. I like the India stock vision because it revisualizes that the system has layers. So, there are a lot of layers. And the reason why the central bank is successful in implementing PIX, the other reason is because they are a very independent, well-resourced institution, which is key for the success of the DPIs, knowing that the institution that implement them understand what they are doing. They have the resources to do this, both intellectual and financial resources to stimulate this. And they are able to implement it with other stakeholders. So, I think this is a very good example of how the PIX was introduced. So, in the context of PIX and the Brazilian central bank, they really understood that it was not only the PIX, but it was how the PIX was introduced, and they understood very well the system. Let me give you a very concrete example of why. So, Brazil, as most global south countries, although Brazil has a very good level of connectivity, most of the country is not meaningfully connected. As most of the global south, we are connected to the internet, we are connected to the internet, we are connected to the internet, we are connected to social media, and all the rest is very expensive, so, if you want to introduce a new digital service, a DPI, people will not use it, because they have to spend money, and that is why, when the Brazilian government was going to introduce PIX, and META proposed to introduce what’s a payment, three months before the PIX, the Brazilian central bank had the power to minimize, in fact, the risk of falling at the beginning of the first section of the world, and watch the power of the PIX entering force due to competition, data protection, and consumer protection. We will speed this up a little bit, but before it end, a little cautious shift for today. So, they understood very well that if WhatsApp had introduced WhatsApp payment before the PIX, today we would not be celebrating the PIX as a great success. Everyone in Brazil would be using WhatsApp payment. And that leads me also to celebrate another great success of India, and I think that colleagues from the Telecom Authority of India should be really praised for their vision in 2016 when they adopted very strong net neutrality rules that prohibited zero rating in India. And I think that the reason why DPIs have been so successful in India, and I think this is the greatest untold story of good digital sovereignty, India prohibited zero rating in 2016. It’s the only global South country that did this, and the result is not only that connectivity prices have dropped by 90% and connectivity has boomed more than 300% of adoption increase in the past eight years, but also that India is experiencing a belle époque of innovation. Why? Because you can access everything. You don’t have only social media for free and all the rest extremely expensive. Everything is cheap. And that is something that you need to consider. So, you need really to have a systemic vision before thinking about the DPI, and also you need to have an institution that is able to implement that vision. We had this chance in Brazil with the Brazilian Central Bank. I’m not sure all other institutions would be able to do this same success.

Sabhanaz Rashid Diya: Thanks, Luca. Smriti , I’ll turn to the question specifically to you on the role of different stakeholders in that process.

Smriti Parsheera: Yeah, thanks for that question. You know, on the financial inclusion question and stakeholders in that process, I think it’s an important question to think about what were the causes of exclusion, and I think to the extent ID was a cause, you know, we have seen DPIs addressing that, but there is, I think, an underlying question that, you know, what sort of ID did you need? Do you necessarily need a biometric-based ID to solve for financial inclusion? I think those are the debates to be had. So, I think, you know, I think it’s important for us to think about, you know, how do we define financial inclusion? How do we define it in terms of, you know, how do we, you know, certainly, you know, have recognition and being able to verify and identify is a building block. The other question, I think, which is important to the DPI conversation is about, you know, how do you define financial inclusion? Is it only about payments and remittances? A lot of the DPI focuses around that, or is it about a broader bouquet of financial services? And are we thinking about, you know, insurance? Are we thinking about DPI on those fronts? So far, we don’t very actively seem to be solving those through DPI, and I think that’s where, you know, the next layer of thinking should be.

Sabhanaz Rashid Diya: Thanks, Smriti . Keith, there’s a question from, specifically, for South Africa, as well, from your experience. And then maybe just for you. Sorry? I don’t think it’s a South African question.

Keith Breckenridge: I think it’s a global question. For me, the first question is the liability question. You know, who should take responsibility? Someone finds out your information on Facebook, uses a network provider to send you an enticing SMS, you click on the link, and then your pension fund is gone, you know? And this is happening, let’s be honest, it’s happening to, you know, a significant proportion of the populations that are meant to be basically getting resources and capabilities from this. And I think it may be on the age. So, I’m 60, which is, you know, I spent quite a lot of time working with people who are 85, 90, old people, the normal old people who are the actual targets. They’re not 25-year-old Estonians, they’re 80-year-old Africans, and they really don’t have the skill set that they need, that the designers of these systems are assuming are actually available to them. They can’t, they don’t see the buttons, they can’t read the text, they can’t, you know, they can’t, you know, they can’t, you know, they can’t read the text, it’s so easy for them to be robbed. I mean, it’s truly a joke. And no one is really looking at this. No one wants to take it seriously. seriously. The only people who are also the English the British are doing this and you know this they’ve imposed an eighty five thousand pound liability on the banks. Regardless of what happens someone empties your bank account. The banks are responsible. And that’s as far as I can see the only country in the world that’s doing that. And when we start talking about sharing liability the network providers will stop allowing criminals SMS to follow to flow on their systems. You know then then we can talk. So if we’re doing this let’s be aware of what we’re doing. We’re building an infrastructure that’s going to make a huge population newly vulnerable to really serious crimes. And there’s no discussion of what we should be doing about that. So I would start with that. I think that the second. The second issue for me is really the question about asset creation. I’m much more worried about asset creation than I am about what we call usurious interest rates. It’s people need to be getting something out of these mobile systems. It’s not enough to make it a kind of subsistence question. We have to look for how people can create buy things material things that they can keep and not turn it towards consumption. And that for me is fundamentally a matter of fiduciaries and also these kind of collective forms of savings association that are very common everywhere. You find them of one kind or another. They’re very normal. But DPI privileges what we call individual biographical credit surveillance. It gives you an ID number that makes the ID number the core of the credit scoring system. It’s also one of the reasons why everywhere even in the U.S. firms are much less visible. That’s the real problem we’ve got. And again I’d like to hear the engineers who are building this huge infrastructure that’s putting us all at risk speak about the implications of that.

Sabhanaz Rashid Diya: Thanks Keith. In the interest of time maybe I’m going to you know probably not come back to Sreti Luka and Keith but actually give Bidisha Armando. So, Mr. Singh here and Rasmus, a chance to give their closing remarks with like one quick question, which is also from the participants, is really how do we think about weaving in these principles into DPI governance? Some are very legacy systems, some are newer systems, so how do we actually do it? So maybe a quick 30-second, I’ll start with Rasmus.

Rasmus Lumi: Well, I will use my 30 seconds to maybe be quiet. I don’t know how to actually do it. I can only maybe refer to or reuse something I said 10 years ago and with reference with what Keith was mentioning about old people, young people, that I am firmly confident that old people are not as stupid as we think they are and young people are not as smart as we think they are, even in digital.

Sabhanaz Rashid Diya: Fair point. Bidisha?

Bidisha Chaudhury: Yeah, I also don’t have a profound thought last, you know, closing statement, but I think one thing that in the discussion we see that India being this success story in terms or other countries being success story in terms of how many people have bank accounts or how many people, so my broader question would be like, is digitalization equal to inclusion? Is the question that I think I would end with.

Sabhanaz Rashid Diya: Thanks, Bidisha. Armando?

Armando Manzueta: Well, building on the same statements that preceded me, basically, I just want to remark that to earn people’s trust, DPIs must be more than efficient. They must be ethical, inclusive, and accountable. And that means that safeguards, so an AI on the broad DPI effort should be qualified into law, should be embedded into law, ensuring that civil society has a seat on the table, and making systems transparent and elitable. We in the Dominican Republic, we are committed to that. we’re actually trying to do that. We’re committed to building a DPI, not just as a government infrastructure, but rather a civic infrastructure that are governed not just by code, but by values that protects people and empowers them to achieve their very best. That’s all I can say.

Sabhanaz Rashid Diya: Thanks, Armando. Over to you, Mr. Singh, last word.

Sheo Bhadra Singh: I would just like to say that any DPI or any system which is being built has to be user-centric, citizen-centric, and the goal and the principles have to be to how the society is to be served. That is the basic principles, and everything else can be actually built. There are no issues.

Sabhanaz Rashid Diya: Wonderful, on that closing thought of building people-centered DPI, thank you, everyone, for joining in on our panel. I’m sure lunch is being served right now, so I hope to continue the conversation. Thank you for all our participants for joining and staying on, and I look forward to the next IGF. Thank you. Thank you.

Agreement points

DPI requires strong institutional capacity and multi-stakeholder governance

Speakers

– Luca Belli
– Sheo Bhadra Singh
– Armando Manzueta
– Rasmus Lumi

Arguments

Successful DPI implementation requires independent, well-resourced institutions that understand the technology and can work with stakeholders, plus systemic vision considering connectivity and affordability – Institutional Capacity Requirements

Indian state played three key roles: creating institutional space (NPCI), scaling through policy mandates, and regulatory oversight of multi-stakeholder ecosystem – State Leadership Model

Dominican Republic emphasizes shared responsibility and oversight from civil society and private sector in DPI governance and implementation – Multi-stakeholder Oversight

Freedom Online Coalition’s rights-respecting DPI principles include human-centered solutions, inclusivity, transparency, privacy, security, sustainability, data-driven approaches, and technology neutrality – Rights-Based Principles Framework

Summary

All speakers agree that successful DPI implementation requires capable institutions with adequate resources, technical expertise, and the ability to coordinate multiple stakeholders including government, private sector, and civil society.

Topics

Infrastructure | Development | Legal and regulatory

User-centric and citizen-centric design is fundamental to DPI success

Speakers

– Sheo Bhadra Singh
– Armando Manzueta
– Rasmus Lumi

Arguments

DPI must be user-centric and citizen-centric with the goal of serving society as the basic principle from which everything else can be built – User-Centric Design Principle

DPI is not just a set of technologies but a nation-building tool that must be grounded in rights, resilience, and trust – Holistic Definition

Freedom Online Coalition’s rights-respecting DPI principles include human-centered solutions, inclusivity, transparency, privacy, security, sustainability, data-driven approaches, and technology neutrality – Rights-Based Principles Framework

Summary

Speakers consistently emphasize that DPI systems must prioritize citizen needs and human-centered design principles rather than technological or institutional priorities.

Topics

Development | Human rights principles | Infrastructure

Transparency and accountability are essential for DPI governance

Speakers

– Luca Belli
– Keith Breckenridge
– Smriti Parsheera
– Rasmus Lumi

Arguments

Brazil’s PIX is truly public because it’s delivered by government (Central Bank), ensuring freedom of information access, unlike India’s foundation-based model that can avoid transparency requirements – Public vs. Quasi-Public Debate

Metrics of exclusion should be made visible, including publishing numbers of disabled identity accounts and response times for queries, to ensure accountability – Transparency in Exclusion

Civil society participation is limited in DPI deployment, which follows top-down approaches without robust transparency and accountability culture like regulatory impact assessments – Limited Participation

Freedom Online Coalition’s rights-respecting DPI principles include human-centered solutions, inclusivity, transparency, privacy, security, sustainability, data-driven approaches, and technology neutrality – Rights-Based Principles Framework

Summary

Speakers agree that DPI systems must incorporate strong transparency mechanisms and accountability measures, including public access to information about system performance and failures.

Topics

Human rights principles | Legal and regulatory | Development

Similar viewpoints

These speakers share critical perspectives on DPI implementation, highlighting risks of creating new forms of exclusion, exploitation, and corporate dominance despite public ownership claims.

Speakers

– Keith Breckenridge
– Smriti Parsheera
– Bidisha Chaudhury

Arguments

Financial inclusion through DPI risks encouraging usurious lending at 1% daily interest and pushing people into blacklisting systems, with South Africa seeing one trillion Rand spent on online gambling (one-third of GDP) – Financial Inclusion Risks

Strong state-market alliance exists in DPI implementation, but risks creating ‘alt big tech’ that pursues scale recklessly while being blind to due process – State-Market Alliance Concerns

Even with state-owned DPI like UPI, big tech influence remains embedded as major traffic controllers, with vendors often identifying the system with Google Pay rather than public infrastructure – Big Tech Persistence Challenge

Topics

Development | Digital business models | Consumer protection

Both speakers present DPI as successful tools for financial inclusion and breaking private monopolies, emphasizing the scale and positive impact of their respective national systems.

Speakers

– Luca Belli
– Sheo Bhadra Singh

Arguments

Brazil’s PIX payment system broke Visa/MasterCard duopoly, reduced transaction costs from 3-5% to zero, and distributed data collection while maintaining public accountability through freedom of information laws – Brazilian Success Model

India’s UPI processed 18.77 billion transactions worth $290 billion in one month of 2025, demonstrating massive adoption enabled by existing digital identity and banking infrastructure – Transaction Volume Success

Topics

Development | Digital business models | Infrastructure

Both speakers emphasize the importance of genuine public ownership and citizen control over DPI systems, with strong accountability mechanisms and transparency requirements.

Speakers

– Rasmus Lumi
– Luca Belli

Arguments

Estonia’s model ensures every citizen owns and controls their data, with ability to verify which officials accessed their information, providing preventive accountability – Citizen Data Ownership

Brazil’s PIX is truly public because it’s delivered by government (Central Bank), ensuring freedom of information access, unlike India’s foundation-based model that can avoid transparency requirements – Public vs. Quasi-Public Debate

Topics

Privacy and data protection | Human rights principles | Legal and regulatory

Unexpected consensus

Trust as fundamental but uncodifiable element

Speakers

– Rasmus Lumi
– Armando Manzueta

Arguments

Trust is the most important but uncodifiable element for DPI success – people can be forced to use systems but won’t maximize benefits without trust – Trust as Foundation

DPI is not just a set of technologies but a nation-building tool that must be grounded in rights, resilience, and trust – Holistic Definition

Explanation

It’s unexpected that both a European developed country representative and a Latin American developing country representative would emphasize trust as a fundamental but intangible requirement for DPI success, showing convergence across different development contexts.

Topics

Development | Human rights principles | Infrastructure

Need for liability frameworks and consumer protection

Speakers

– Keith Breckenridge
– Audience

Arguments

Need for liability frameworks where financial institutions bear responsibility for fraud losses, as implemented in Britain with £85,000 bank liability regardless of circumstances – Liability Gap Problem

Digital financial inclusion risks creating opportunities for misuse, abuse, and data harvesting rather than solving problems, requiring strong AI ethical standards and transparency measures – Financial Inclusion Risk Management

Explanation

The convergence between an academic researcher’s detailed analysis and audience concerns about consumer protection shows unexpected alignment on the need for stronger liability and protection frameworks in DPI systems.

Topics

Consumer protection | Legal and regulatory | Inclusive finance

Overall assessment

Summary

The panel shows strong consensus on foundational principles (user-centric design, transparency, institutional capacity) while revealing significant disagreement on DPI outcomes and risks. Government representatives emphasize success stories and scale achievements, while academics and civil society highlight implementation challenges and unintended consequences.

Consensus level

Medium consensus on principles but low consensus on outcomes. This suggests that while stakeholders agree on what DPI should achieve in theory, there are fundamental disagreements about whether current implementations are delivering on these promises. The implications are that DPI development needs more robust evaluation frameworks and stronger accountability mechanisms to bridge the gap between aspirational principles and practical outcomes.

Different viewpoints

Ownership structure and accountability of DPI systems

Speakers

– Luca Belli
– Smriti Parsheera

Arguments

Brazil’s PIX is truly public because it’s delivered by government (Central Bank), ensuring freedom of information access, unlike India’s foundation-based model that can avoid transparency requirements – Public vs. Quasi-Public Debate

Strong state-market alliance exists in DPI implementation, but risks creating ‘alt big tech’ that pursues scale recklessly while being blind to due process – State-Market Alliance Concerns

Summary

Luca Belli advocates for direct government ownership of DPI (as in Brazil’s PIX) to ensure transparency and accountability through freedom of information laws, while Smriti Parsheera warns about the risks of state-market alliances that may replicate big tech problems and pursue scale without proper due process.

Topics

Legal and regulatory | Human rights principles | Infrastructure

Assessment of financial inclusion benefits through DPI

Speakers

– Sheo Bhadra Singh
– Keith Breckenridge

Arguments

India’s DPI journey started with Aadhaar in 2009-2010 to provide public services access to large excluded populations, leading to 1.33 billion digital identities and 550 million zero-balance bank accounts – Indian Scale Achievement

Financial inclusion through DPI risks encouraging usurious lending at 1% daily interest and pushing people into blacklisting systems, with South Africa seeing one trillion Rand spent on online gambling (one-third of GDP) – Financial Inclusion Risks

Summary

Singh presents India’s massive scale of financial inclusion as a success story with hundreds of millions gaining bank accounts, while Breckenridge warns that financial inclusion through DPI can lead to exploitation through predatory lending and gambling addiction, citing South Africa’s experience.

Topics

Development | Inclusive finance | Consumer protection

Role of big tech in DPI ecosystems

Speakers

– Luca Belli
– Bidisha Chaudhury

Arguments

DPI represents ‘good digital sovereignty’ that fosters competition, domestic innovation, and people empowerment, contrasting with authoritarian or protectionist approaches – Positive Sovereignty Framework

Even with state-owned DPI like UPI, big tech influence remains embedded as major traffic controllers, with vendors often identifying the system with Google Pay rather than public infrastructure – Big Tech Persistence Challenge

Summary

Belli sees DPI as an effective counter to big tech dominance that can foster competition and innovation, while Chaudhury argues that big tech influence persists even in state-owned DPI systems, with users often not recognizing the public nature of the infrastructure.

Topics

Digital business models | Infrastructure | Development

Effectiveness of digitalization for inclusion

Speakers

– Sheo Bhadra Singh
– Bidisha Chaudhury

Arguments

DPI must be user-centric and citizen-centric with the goal of serving society as the basic principle from which everything else can be built – User-Centric Design Principle

Question whether digitalization equals inclusion, challenging the assumption that digital access automatically leads to meaningful social and economic inclusion – Digitalization vs. Inclusion Debate

Summary

Singh advocates for user-centric DPI design as the foundation for serving society, while Chaudhury fundamentally questions whether digital access automatically translates to meaningful inclusion, challenging the basic assumption underlying DPI initiatives.

Topics

Development | Digital access | Human rights principles

Unexpected differences

Trust as a foundational element versus systemic risks

Speakers

– Rasmus Lumi
– Keith Breckenridge

Arguments

Trust is the most important but uncodifiable element for DPI success – people can be forced to use systems but won’t maximize benefits without trust – Trust as Foundation

DPI systems face backend hacking risks where criminals steal encryption codes to access online accounts, plus pig butchering scams, particularly targeting vulnerable elderly populations – Security Vulnerabilities

Explanation

This disagreement is unexpected because both speakers represent countries with extensive DPI experience, yet they have fundamentally different perspectives on the primary challenge. Lumi emphasizes building trust as the key success factor, while Breckenridge focuses on protecting users from systemic vulnerabilities and exploitation.

Topics

Development | Human rights principles | Cybersecurity

Scale and speed of implementation priorities

Speakers

– Sheo Bhadra Singh
– Smriti Parsheera

Arguments

India’s UPI processed 18.77 billion transactions worth $290 billion in one month of 2025, demonstrating massive adoption enabled by existing digital identity and banking infrastructure – Transaction Volume Success

Strong state-market alliance exists in DPI implementation, but risks creating ‘alt big tech’ that pursues scale recklessly while being blind to due process – State-Market Alliance Concerns

Explanation

This disagreement is unexpected as both speakers are from India and discussing the same DPI systems, yet Singh celebrates the massive scale achievement while Parsheera warns about the risks of pursuing scale recklessly without proper due process.

Topics

Development | Digital business models | Human rights principles

Overall assessment

Summary

The discussion reveals significant disagreements on fundamental aspects of DPI implementation: ownership models (public vs. quasi-public), effectiveness of financial inclusion, persistence of big tech influence, and whether digitalization equals inclusion. There are also tensions between celebrating scale achievements versus warning about systemic risks.

Disagreement level

Moderate to high level of disagreement with significant implications for DPI policy. The disagreements suggest that there is no consensus on best practices for DPI implementation, governance models, or even basic assumptions about benefits versus risks. This lack of consensus could hinder the development of universal DPI principles and standards, as different stakeholders have fundamentally different perspectives on what constitutes success and what risks should be prioritized.

Partial agreements

Similar viewpoints

These speakers share critical perspectives on DPI implementation, highlighting risks of creating new forms of exclusion, exploitation, and corporate dominance despite public ownership claims.

Speakers

– Keith Breckenridge
– Smriti Parsheera
– Bidisha Chaudhury

Arguments

Financial inclusion through DPI risks encouraging usurious lending at 1% daily interest and pushing people into blacklisting systems, with South Africa seeing one trillion Rand spent on online gambling (one-third of GDP) – Financial Inclusion Risks

Strong state-market alliance exists in DPI implementation, but risks creating ‘alt big tech’ that pursues scale recklessly while being blind to due process – State-Market Alliance Concerns

Even with state-owned DPI like UPI, big tech influence remains embedded as major traffic controllers, with vendors often identifying the system with Google Pay rather than public infrastructure – Big Tech Persistence Challenge

Topics

Development | Digital business models | Consumer protection

Both speakers present DPI as successful tools for financial inclusion and breaking private monopolies, emphasizing the scale and positive impact of their respective national systems.

Speakers

– Luca Belli
– Sheo Bhadra Singh

Arguments

Brazil’s PIX payment system broke Visa/MasterCard duopoly, reduced transaction costs from 3-5% to zero, and distributed data collection while maintaining public accountability through freedom of information laws – Brazilian Success Model

India’s UPI processed 18.77 billion transactions worth $290 billion in one month of 2025, demonstrating massive adoption enabled by existing digital identity and banking infrastructure – Transaction Volume Success

Topics

Development | Digital business models | Infrastructure

Both speakers emphasize the importance of genuine public ownership and citizen control over DPI systems, with strong accountability mechanisms and transparency requirements.

Speakers

– Rasmus Lumi
– Luca Belli

Arguments

Estonia’s model ensures every citizen owns and controls their data, with ability to verify which officials accessed their information, providing preventive accountability – Citizen Data Ownership

Brazil’s PIX is truly public because it’s delivered by government (Central Bank), ensuring freedom of information access, unlike India’s foundation-based model that can avoid transparency requirements – Public vs. Quasi-Public Debate

Topics

Privacy and data protection | Human rights principles | Legal and regulatory

Key takeaways

DPI is not a monolithic technology but varies significantly across countries based on national contexts, with successful implementations requiring strong institutional capacity, systemic vision, and public trust

There are fundamental tensions between DPI as national infrastructure and the need for global interoperability and common principles

Ownership models matter critically – truly public systems (like Brazil’s PIX) offer better accountability through transparency laws compared to quasi-public foundations that can avoid information disclosure requirements

DPI can successfully break monopolies and reduce costs (Brazil’s PIX eliminated 3-5% transaction fees from Visa/MasterCard duopoly) while distributing data collection and fostering local innovation

Financial inclusion through DPI carries significant risks including usurious lending, vulnerability to fraud, and potential harm to elderly populations who lack digital skills

Big tech influence persists even in state-owned DPI systems, with users often identifying public infrastructure with private platforms (e.g., UPI being seen as Google Pay)

Civil society participation in DPI development has been limited, following top-down approaches without robust transparency and accountability mechanisms

Technical success requires addressing systemic factors like connectivity costs and net neutrality – India’s zero rating prohibition was crucial for DPI adoption

Trust is the most important but uncodifiable element for DPI success, and digitalization does not automatically equal meaningful inclusion

Resolutions and action items

Freedom Online Coalition under Estonia’s chairship is developing rights-respecting DPI principles including human-centered solutions, inclusivity, transparency, privacy, security, and interoperability

India has offered to provide its DPI platforms free of cost to any country wanting to adopt them, as announced during G20 presidency

Dominican Republic is implementing specific technical governance standards focusing on data quality, observability, and orchestration mechanisms to make data protection principles enforceable

Need to establish liability frameworks where financial institutions bear responsibility for fraud losses, following Britain’s model of £85,000 bank liability

Unresolved issues

How to effectively weave human rights principles into existing legacy DPI systems versus newer implementations

Whether digitalization truly equals inclusion or if it creates new forms of exclusion

How to balance the state-market alliance in DPI without creating ‘alt big tech’ that pursues reckless scaling

How to address the vulnerability of elderly and digitally unskilled populations to fraud and exploitation in DPI systems

How to ensure meaningful civil society participation beyond token consultation in DPI governance

How to define and measure financial inclusion beyond basic payments and remittances to include broader financial services

How to create effective trusted fiduciary systems that can help citizens navigate DPI complexities and resolve issues

How to make metrics of exclusion visible and ensure accountability for system failures and disabled accounts

Suggested compromises

Implementing virtual ID as default instead of main ID to balance privacy with functionality, learning from India’s Aadhaar experience

Creating hybrid governance models that combine public ownership with multi-stakeholder oversight including civil society and private sector participation

Developing modular DPI architectures that allow different institutions to manage their own systems while maintaining interoperability through common platforms

Establishing shared liability models between network providers, financial institutions, and government agencies for fraud prevention and victim compensation

Building DPI with both external privacy (from outside users) and internal privacy (with checks on government access) to address different stakeholder concerns

Allowing space for different civil society values – both market-based empowerment and human rights approaches – in DPI platform governance

So what is interesting here is not only that the customer welfare is increased, but also that competition is automatically enhanced, innovation is produced, and data governance is distributed. And the local entrepreneurs that innovate thanks to this, then they have to pay taxes in Brazil. So we have a very different dynamic here of empowerment through DPIs.

Speaker

Luca Belli

Reason

This comment reframes DPI from a purely technical solution to a comprehensive economic sovereignty tool. Belli demonstrates how Brazil’s PIX system simultaneously broke foreign duopolies (Visa/MasterCard), reduced costs for consumers, distributed data collection power, and kept innovation benefits within the national economy through taxation.

Impact

This shifted the discussion from abstract principles to concrete economic impacts, establishing a framework for evaluating DPI success beyond just service delivery. It influenced subsequent speakers to consider sovereignty and economic empowerment aspects of their own DPI implementations.

One trillion Rand was spent on online gambling systems in 2023. That’s a third of GDP one third of GDP spent on gambling… We’re building an infrastructure that’s going to make a huge population newly vulnerable to really serious crimes. And there’s no discussion of what we should be doing about that.

Speaker

Keith Breckenridge

Reason

This comment provides a stark counternarrative to DPI success stories by revealing unintended consequences. Breckenridge exposes how financial inclusion through DPI can enable harmful behaviors like gambling addiction and financial crimes, challenging the assumption that digital access automatically equals empowerment.

Impact

This fundamentally shifted the tone from celebratory to cautionary, forcing other panelists to acknowledge risks. It introduced the critical concept of ‘metrics of exclusion’ and liability questions that weren’t being addressed in DPI design, adding necessary complexity to the discussion.

So in the process of being a pushback against big tech, could we be creating another ecosystem which is pursuing scale in the same reckless manner, which is, you know, really blind to due process, really blind to other things in the pursuit of reaching scale fast? So I think those are things to be conscious about in the DPI story.

Speaker

Smriti Parsheera

Reason

This insight introduces the concept of ‘alt big tech’ – the idea that DPI, while intended as an alternative to Silicon Valley dominance, might replicate the same problematic approaches to scaling technology without adequate safeguards. It’s a meta-critique of the DPI movement itself.

Impact

This comment prompted deeper reflection on DPI governance models and participation. It challenged the panel to consider whether state-led DPI initiatives were truly different from private tech monopolies, leading to more nuanced discussions about ownership, accountability, and the role of civil society.

So they don’t even think of it as an UPI which is owned by the state, or maybe not owned by, maintained and sustained by the state. So I think that is where it is, it becomes very critical to understand, or even explore that how do we actually circumvent the influence of the big tech when we rely on an infrastructure where the big tech interests are so much more embedded already.

Speaker

Bidisha Chaudhury

Reason

This observation reveals a critical gap between DPI policy intentions and ground reality. Despite UPI being state-owned infrastructure, users primarily interact with it through Google Pay and PhonePay, effectively making big tech the face of public infrastructure.

Impact

This comment exposed the complexity of true digital sovereignty, showing how public infrastructure can still be captured by private interests at the user interface level. It added nuance to discussions about ownership and control, influencing later conversations about the need for stronger governance frameworks.

The fundamental issue really has to be making what we call metrics of exclusion much more visible… This is the number of people whose identity numbers are at the moment disabled. And this is how long it’s taken for us to basically answer queries around those numbers.

Speaker

Keith Breckenridge

Reason

This comment introduces a crucial accountability mechanism that’s missing from most DPI discussions. Instead of just measuring inclusion success, Breckenridge argues for transparency about who gets excluded and why, using South Africa’s example of 2 million disabled identity numbers.

Impact

This shifted the conversation toward concrete governance mechanisms and transparency requirements. It influenced the discussion about the need for oversight, fiduciaries, and accountability measures, moving beyond theoretical principles to practical implementation requirements.

is digitalization equal to inclusion? Is the question that I think I would end with.

Speaker

Bidisha Chaudhury

Reason

This closing question encapsulates a fundamental tension running throughout the discussion. It challenges the core assumption underlying much DPI advocacy – that providing digital access automatically translates to meaningful inclusion and empowerment.

Impact

Though brief, this question crystallized the skeptical undercurrent that had been building throughout the discussion, serving as a powerful summary of the concerns raised about conflating technical solutions with social outcomes.

Overall assessment

These key comments transformed what could have been a promotional discussion about DPI success stories into a nuanced examination of digital governance challenges. The interplay between optimistic government perspectives (Singh, Manzueta) and critical academic analysis (Breckenridge, Parsheera, Chaudhury) created a productive tension that elevated the conversation. Luca Belli’s economic sovereignty framework provided a middle ground, showing how DPI can deliver benefits while acknowledging implementation complexities. The discussion evolved from celebrating technical achievements to grappling with fundamental questions about power, accountability, and the relationship between digitalization and genuine inclusion. The critical voices didn’t dismiss DPI but demanded more sophisticated approaches to governance, transparency, and protection of vulnerable populations – ultimately strengthening the case for more thoughtful DPI development.

How do we truly make DPI a public interest technology and a public facing technology, and how should policymakers, civil society and other stakeholders be thinking about ownership within the DPI architecture?

Speaker

Sabhanaz Rashid Diya

Explanation

This question addresses the fundamental challenge of ensuring DPI serves public interests rather than creating new monopolies, and clarifying ownership structures across the modular DPI ecosystem.

To what extent can we create an alternative infrastructure that is truly independent of big tech when big tech interests are already so embedded in the digital ecosystem?

Speaker

Bidisha Chaudhury

Explanation

This explores whether DPI can actually achieve digital sovereignty or if it remains dependent on existing big tech infrastructure, as evidenced by Google Pay’s dominance in India’s UPI system.

Is digitalization equal to inclusion?

Speaker

Bidisha Chaudhury

Explanation

This fundamental question challenges the assumption that digital access automatically leads to meaningful inclusion and empowerment.

How do we define financial inclusion – is it only about payments and remittances, or should it encompass a broader bouquet of financial services including insurance?

Speaker

Smriti Parsheera

Explanation

This questions whether current DPI approaches to financial inclusion are too narrow and whether they should address broader financial service needs.

What sort of ID do you actually need for financial inclusion – do you necessarily need a biometric-based ID?

Speaker

Smriti Parsheera

Explanation

This challenges assumptions about the necessity of biometric identification systems for achieving financial inclusion goals.

Who should take liability when people are defrauded through DPI systems – network providers, banks, or government?

Speaker

Keith Breckenridge

Explanation

This addresses the critical gap in accountability when vulnerable populations are targeted by criminals exploiting DPI infrastructure.

How can we make metrics of exclusion more visible in DPI systems, including publishing data on disabled identity numbers and response times for queries?

Speaker

Keith Breckenridge

Explanation

This calls for transparency about who is being excluded from DPI systems and how quickly exclusion issues are resolved.

How do we create space for trusted fiduciaries who can help people navigate DPI systems when errors occur?

Speaker

Keith Breckenridge

Explanation

This addresses the need for intermediaries to help vulnerable populations when they encounter problems with DPI systems.

How do we weave human rights principles and AI ethical standards into existing DPI architectures, both legacy and newer systems?

Speaker

Online participant and Sabhanaz Rashid Diya

Explanation

This explores the practical challenge of retrofitting ethical frameworks into operational DPI systems, especially as they expand into AI applications.

What are the opportunities for mobile network operators to collaborate with government in DPI implementation?

Speaker

Elizabeth (audience member)

Explanation

This seeks to understand the role of telecommunications infrastructure providers in the DPI ecosystem and potential partnership models.

How do we guard against digital financial inclusion becoming a risk rather than a solution, particularly regarding data harvesting and abuse?

Speaker

Elizabeth (audience member)

Explanation

This addresses concerns about unintended negative consequences of DPI implementation, particularly for vulnerable populations.

How do we address the problem of online gambling consuming significant portions of GDP through DPI payment systems?

Speaker

Keith Breckenridge

Explanation

This highlights an unexpected consequence of financial inclusion through DPI that requires policy attention and research.