WS #257 Emerging Norms for Digital Public Infrastructure

Summary

This panel discussion focused on the concept of Digital Public Infrastructure (DPI) and its implications for global digital development. The panelists explored various definitions and applications of DPI, with a particular emphasis on its role in emerging economies. They discussed how DPI, such as digital identity systems and payment infrastructures, can promote financial inclusion and economic growth, citing examples from India and Brazil.

The conversation highlighted the tension between viewing DPI as a tool for digital sovereignty and concerns about potential fragmentation of the internet. Panelists debated the role of governments versus markets in developing DPI, with some arguing for a minimum viable infrastructure approach to protect free markets and democracy. The discussion also touched on the importance of interoperability, open standards, and multi-stakeholder cooperation in DPI development.

A key point of contention was the definition of DPI itself, with some panelists advocating for a narrower, value-driven definition to avoid confusion and potential misuse of the term. The panel explored the differences in DPI approaches between the Global North and South, with some arguing that DPI discussions are more prevalent in the Global South due to infrastructure gaps and a desire to address digital colonization.

The discussion concluded with reflections on the need for context-specific DPI solutions, the importance of transparency and accountability in DPI development, and the potential for DPI to address digital inequalities. Panelists emphasized the need for careful consideration of trade-offs and long-term impacts when implementing DPI initiatives.

Keypoints

Major discussion points:

– Defining digital public infrastructure (DPI) and its scope

– The role of government vs. private sector in developing DPI

– Concerns about digital sovereignty and fragmentation

– Challenges and opportunities for DPI in developing countries

– Interoperability and open standards for DPI

The overall purpose of the discussion was to explore the concept of digital public infrastructure, examining different perspectives on its definition, implementation, and implications across various national contexts. The panelists aimed to identify key issues and potential action items for the global community regarding DPI development.

The tone of the discussion was largely analytical and academic, with panelists offering nuanced views based on their experiences and research. There were moments of disagreement, particularly around the role of markets versus government in infrastructure development. The tone became more passionate when discussing issues of digital colonialism and inequality between the Global North and South. Overall, the conversation maintained a constructive and collaborative spirit despite differing viewpoints.

Speakers

– Milton Mueller: Director of the Internet Governance Project at Georgia Institute of Technology

– Luca Belli: Professor at FGV Law School, Director of the Center for Technology and Society at FGV

– Henri Verdier: Ambassador from the French Ministry for Europe and Foreign Affairs

– David Magård: Works for the Swedish Public Registration Office, coordinator of EWC (European Wallet Consortium)

– Jyoti Panday: Regional Director for the Internet Governance Project

– Anriette Esterhuysen: Former chair of the Multistakeholder Advisory Group

Additional speakers:

– Kashfi Inua: From Nigeria (audience member)

Digital Public Infrastructure (DPI) Panel Discussion Summary

This panel discussion brought together experts from various fields to explore the concept of Digital Public Infrastructure (DPI) and its implications for global digital development. The conversation covered a wide range of topics, including definitions of DPI, its potential benefits and risks, implementation challenges, and the differing perspectives between the Global North and South.

Defining Digital Public Infrastructure

One of the central themes of the discussion was the lack of consensus on a clear definition of DPI. Several panelists offered their perspectives:

– Henri Verdier described DPI as the “minimum necessary infrastructure to protect free internet, market and democracy”.

– Luca Belli defined it as “digital systems built on open standards that are interoperable and secure to provide services”.

– Jyoti Panday characterized DPI as an “approach for building large-scale networks, platforms and services essential for digital economy”.

– David Magarde noted that DPI is not a commonly used term in the EU, where the focus is instead on interoperability.

The diversity of these definitions highlighted the need for a more precise and value-driven conceptualization of DPI. Anriette Esterhuysen, in particular, emphasized the importance of narrowing the definition to avoid confusion and potential misuse of the term.

Benefits and Risks of DPI

The panelists discussed both the potential advantages and drawbacks of DPI implementation:

Benefits:

– Breaking monopolies and increasing competition, as exemplified by the PIX payments system in Brazil

– Facilitating multi-stakeholder cooperation and intergovernmental collaboration

– Cost-effectiveness, as highlighted by Henri Verdier, who noted that DPI can be significantly cheaper than traditional infrastructure

Risks:

– Centralization and compromise of digital security

– Potential for control rather than public benefit, depending on the context

– Risk of fragmentation if pursued with a sovereignty-based approach

– Potential for creating new forms of monopolies or oligopolies

Luca Belli provided a detailed comparison between Brazil’s PIX system and India’s UPI, highlighting how these DPI initiatives have increased financial inclusion and competition in the payments sector.

Implementation and Governance

The panel explored various approaches to DPI implementation and governance:

– Luca Belli advocated for a bottom-up approach with stakeholder engagement based on local realities.

– David Magarde discussed the EU’s approach to digital identity wallets, emphasizing the need for regulation and openness in DPI development.

– Jyoti Panday highlighted the importance of institutional frameworks and oversight mechanisms.

– Henri Verdier emphasized that DPI should be developed through multi-stakeholder processes.

There was general agreement on the need for proper institutional frameworks, oversight mechanisms, and regulation to ensure openness and protect democratic values. The importance of open standards and interoperability in DPI development was stressed by multiple panelists.

The role of government versus market forces in DPI development was a key point of contention. Some panelists expressed concerns about DPI leading to fragmentation if pursued with a sovereignty-based approach, while others emphasized the need for government intervention to protect public interests.

Global Perspectives on DPI

The discussion revealed significant differences in how DPI is perceived and approached in different parts of the world:

– Anriette Esterhuysen noted that DPI is more frequently discussed in the Global South due to the lack of existing infrastructure.

– Luca Belli observed that the Global North is now studying Global South approaches to digital sovereignty.

– Henri Verdier pointed out the differences in DPI needs between the Global North and South.

– Milton Mueller highlighted institutional barriers to infrastructure development in some countries.

An audience member from Nigeria raised an important point about the need to consider “leapfrog regions” lacking infrastructure in the Global North as well, challenging the traditional North-South divide in digital development discussions.

Challenges and Future Considerations

The panel identified several key challenges and areas for future consideration:

1. Funding: The discussion touched on the potential role of taxation of big tech companies in funding DPI initiatives.

2. Interoperability: The need for open standards and cross-border compatibility was emphasized by several speakers.

3. Balancing sovereignty and global cooperation: The discussion highlighted the tension between national approaches to DPI and the need for international collaboration.

4. Transparency and accountability: Ensuring these principles in DPI development was seen as crucial for building trust and enabling proper oversight.

5. Digital divide: Addressing gaps in connectivity alongside DPI development was identified as a critical concern, with Luca Belli emphasizing the importance of meaningful connectivity for the success of DPI initiatives.

6. Role of AI: The potential impact of artificial intelligence on DPI was briefly mentioned at the beginning of the discussion.

7. Central banks and cryptocurrencies: The panel touched on the role of central banks in relation to cryptocurrencies and DPI.

Conclusion

The discussion concluded with reflections on the need for context-specific DPI solutions, the importance of transparency and accountability in DPI development, and the potential for DPI to address digital inequalities. Panelists emphasized the need for careful consideration of trade-offs and long-term impacts when implementing DPI initiatives.

While there was moderate consensus on the importance of DPI and the need for proper governance, significant differences remained in perspectives on its implementation, funding, and potential impacts. This suggests that further dialogue and research are needed to develop a more unified approach to DPI development and implementation globally.

The panel discussion successfully highlighted the complex, multifaceted nature of DPI, moving beyond technical definitions to consider historical context, geopolitical factors, governance implications, and potential pitfalls. It revealed DPI as a concept with significant implications for digital governance, economic development, and global power dynamics in the digital age.

Milton Mueller: Well, I’m going to introduce the topic, and then I’m going to introduce the panelists. We had a last-minute cancellation, so we’re lacking perspective from India, but we’re talking today about digital public infrastructure. This has become a fashionable term. Somebody just told me that we should have called it AI digital public infrastructure because that’s even more fashionable, but in fact that is not a very scientific or accurate way of going about this. AI is a digital technology, and the infrastructure supporting digital services and applications, particularly payments, are something where we are bringing together new services in ways that raise policy issues regarding security, trust, competition in the digital economy, and the role of government in the private sector. So we are going to first examine what we mean by DPI. Is this just another buzzword? Is it something real? We’re going to talk about the institutional frameworks for collaboration between states and markets, oversight mechanisms, and the role of multi-stakeholder cooperation in fostering DPI production and governance. So what we’re going to do is spend about 30 minutes in this discussion, including our online audience and the local audience, fully aware of the fact that our online audience will be probably having better sound than we have here. So, let me introduce the panelists now. We have hovering above me on the screen is Ms. Jyoti Pandey. She’s a regional director for the Internet Governance Project. I guess I should introduce myself. I’m Milton Mueller. I’m with the Internet Governance Project, too. I’m the director, and we’re located at the Georgia Institute of Technology. Situated, going from my right to my left, is Luca Belli, and he’s from Brazil. I can’t remember the name of your institute. Yes, the well-known Center for Technology and Society, which goes by the acronym FGV, which nobody relates to technology. FGV is the acronym of the foundation. The foundation, right. From the South to the Vargas. So, it’s the Center for Technology and Society at FGV. And then we have Ambassador Henri Verdier, who is from the French Ministry for Europe and Foreign Affairs. Thanks for joining us. And we have online, we have David Magarde, who’s with the EU, there he is, Digital Identity Wallet Consortium, and the Open Wallet Forum at the ITU. Welcome, David. And last but not least, we have Henriette Esterhausen, a stalwart of civil society participation in these processes. And also, the chair, or the former chair, of the Multistakeholder Advisory Group. Are you still the chair? No, no. OK. So, let’s get started. Let’s begin with definitions and why you think this concept of digital public infrastructure has taken off and become such a buzzword. Let’s begin with Jyoti.

Jyoti Panday: Good morning, everyone. As Professor Muller introduced me, I’m Jyoti Pandey, I work with him at the Internet Governance Project and welcome to Emerging Norms for Digital Public Infrastructure. So the term digital public infrastructure or DPI, as it’s commonly referred to, is an approach or strategy for building large scale networks, platforms and services to mediate key processes or functionality that is essential for operating in the digital economy. So whether it’s digital identity, paying for transactions, it encompasses the underlying design, the institutional frameworks, resources that enable the development and use of these large scale system. DPIs, as we know, are transforming the global economy, they’re impacting business practices, they have altered relations between state, market and citizens. The emergence of DPIs like identity and authentication or interoperable payment systems have blurred the difference between public and private sector, traditional and new economies, tradable and non-tradable products and between goods and services. They’ve also created avenues for development of norms and standards for cybersecurity, privacy, data protection and competition. Approaches such as domestic and cross-border flows of data, intellectual property rights, consumer and data protection, digital security are significantly impacted by the emergence of DPIs and internet governance policies are also being shaped and advanced through DPIs. So as we know, this is a really important topic that everyone wants to weigh in on. But the fundamental issue is that even though we see forums like G20 and bilateral negotiations indicating that the adoption of DPIs are at a tipping point, they are being advanced globally, the lack of definition of consensus around what should be labelled as DPI and what doesn’t fall within that label is missing at the moment. moment. So our aim at this workshop is to have various folks who are part of stakeholder groups who are working and engaging with these processes to kind of shine some light and you know bring to our understanding to increase and expand our understanding of how they are approaching building DPI’s, how they are going about defining the values and infrastructure development and the design of these DPI’s. The advancement of DPI’s is happening even as legitimate concerns about their impact on trust, security and competition on in the digital economy remains unexplored or under addressed. There is a centralization of digital identity and online payments that is happening through under the DPI label and this can lead to policy problems like exclusion, fraud or even compromise of digital security. The rapid development and deployment also has profound implications in terms of disrupting traditional sectors and businesses. Another risk raised by DPI’s which are rooted in digital sovereignty or the claim that states have the stake in running the internet and digital services and should have the maximum say in how they are to be taken forward is that it could lead to fragmentation. So a sovereignty based approach to building and developing DPI’s could lead to fragmentation and we want to kind of explore this tension between fragmentation and cooperation that DPI’s enable. Control of these institutional arrangements and the technical architecture of DPI’s and its usage is creating discrete spaces of data and transactions which can encourage and enable governments to pursue a sovereignty based agenda. Delays in reform and rise in protectionism could hinder the adoption and expansion of DPI’s if embraced elsewhere. Given the stakes and the economic, political, social impact of DPI’s it is important to think through their development. create avenues for oversight. The prevailing discourse is exclusively focusing on legislative and regulatory measures, however interventions such as audits and assessments could also play a vital role. So like Professor Mueller kind of directed that you know we will delve into these other issues as we go by in this discussion, but on definition itself we do have certain global forums that are trying to create the platform and bring together diverse stakeholders to come towards a common definition, but again because of this sovereignty-based approach to developing DPIs the consensus has not been easy to come by and that’s what we are hoping to kind of delve into more over here. I’m happy to like jump in and talk more about it but I want to hear from my panelists. Thank you.

Milton Mueller: Thank you, Jyoti. So why don’t we go next to David. The ITU has played a classical role in basic telecommunications infrastructure coordination. I have never really built infrastructure. Okay, you have to turn it on. Somehow it got turned off. So yes, let’s turn to David and can you give us a brief overview, five minutes or so, of what your perspective is particularly focusing on the definitional issues at this stage?

David Magarde: I can give it a try. So thank you everyone for inviting me to this panel. I sit in a kind of cold Sweden and now I work for the Swedish Public Registration Office under the Ministry of Economic Affairs in Sweden. And also, I am the coordinator of EWC, the wallet consortia. We pilot digital identity wallets for Europe. This is a collaborative effort, 86 organizations from private and public sector, done over two years, 20 million euros, rather big. And I guess in some sense, a digital public infrastructure, although this is not really my area of expertise, so I’m coming with a kind of perspective from what I don’t understand of the public infrastructures and see if that can add to the discussion. When it comes to digital public infrastructures and what we see from, at least what I see, and I’ve been working with the Swedish government, within the Swedish government, for about ten years with digitalization, and of course in the EU as well, part of several expert groups and also the Open Wallet Forum, and some other OECD groups and so on, before. Digital public infrastructures kind of come in, in my view, in the last three years. I wouldn’t say that this is a big discussion in the EU. I haven’t really used it at all, to be honest, in the work that we’ve done. And it’s not something that is really looked at from my perspective when it comes to digital identity ecosystem in the EU. So I think this is interesting for the discussion, because then it seems like we have different kind of angles to it and different understandings. And also, of course, that makes it difficult to come with a kind of cohesive understanding of the public infrastructure and what we should do with it. From my perspective, also speaking about EU and also about Sweden, this is my personal understanding, but digital sovereignty is a core focus of the EU in our digital strategies. I wouldn’t say that it necessarily hinders cooperation with the private sector or other countries, but there is an understanding that we need to some degree have control of some of the key fundamental infrastructures when it comes to digitization, such as digital identity. Although we have a lot of good experience with public infrastructure.

Milton Mueller: We have a lot of ambient noise.

David Magarde: Should I stop?

Milton Mueller: You can go on, but we have a lot of ambient noise here.

David Magarde: Okay. And yes, to finish up, I think what’s mostly interesting, what I can assert is the interoperability between identity systems and interoperability of course on the technological infrastructure layers, but also of course the semantics, legal semantics and definitions and so on. So I’m curious on the interoperability questions in the digital public infrastructure framework, which as far as I’ve seen, it’s not that prominent in the discussion papers and so on that I’ve browsed through the last years.

Milton Mueller: Good. Thank you. I think it’s interesting then that from the perspective of, from David’s perspective, the term is not commonly used and the, whatever concern he has is in fact with interoperability, which as Jyoti flagged, may be an issue if you start taking a sovereigntist approach to DPI. So let’s go on to Ambassador Verdier. Let us, I think you are more involved with this topic.

Henri Verdier: Yeah, I am. So first, let’s say that I’m a veteran of the internet revolution. I started my first company in 1995 and then I became an ambassador. And I say this because that’s always the same thing. The reality is vibrant, evolving, diverse, complex, and we look for words, but the reality doesn’t obey to our words. I say this because DPI movement is one new world for various approaches. And you can connect it to infrastructures, to public service, to sometimes digital commons, to platform strategies. And for example, personally, I discovered probably this kind of ideas. 15 years ago, I was writing a book on platform strategies and I tried to explain, I was trying to explain the success and the strength of the big platforms like Google or Facebook or Apple itself. And suddenly I discovered that some people were speaking about government as a platform. And I was very interested by this approach. And when I became the French head of the French IT department for the government, I tried to develop some building blocks for a government as a platform strategy. And we did develop France Connect or some important APIs. And it was on a certain perspective, a kind of prototype of… what we call now DPI. And I’ve discovered at this time, the Estonian X-Road, for example, that is another kind of ancestor of the DPI approach. And then I discovered what was happening in India. So Pramod is not with us today, but we can here observe a massive impact. Because in less than three years, they did develop a digital ID for more than one billion people and 25% of them didn’t even have a legal existence. And then they had a legal existence plus an ID. Then they did develop a very smart payment interface. That is just, you know, just a set of API. They just decide that the banks has to be able to receive payment orders through this format. And they have a duty to agree to execute payments through this format. That’s all, plus independent body to regulate, to be sure that the banks are correctly implementing the system. And thanks to this approach, a good definition of a set of API, they did manage to let the market conceive 600 payment systems. So of course, Alipay and Google Pay are the biggest, but there is a huge diversity of payment system. You can pay with a QR code or with WhatsApp. It depends on your service. So I mentioned this to say that a lot of people are, a lot of countries or organizations are trying to build this kind of small layer of something, a kind of platformization between the free internet and the market and the society. And that’s very important because as I said, usually in the digital world, platform strategies are very efficient. A platform strategy is a strategy that compacts and distributes some resources for users. Most often they try to take the part of the added value of the user, so they take the platforms, the private platforms. They take data or they try to capture you or they ask for revenue sharing you. But that’s efficient. Very simple and obvious platform strategy is a smartphone itself. If you are a service developer, you are very grateful to Apple or Google to propose to you a connected computer with a camera and a connection and a lot of sensors and a lot of tools plus a good SDK. So you are very grateful and you agree to share 30% of your revenues because they did allow you to innovate and create and try to access to the market. So that’s efficient and that can be dangerous, that was said, because of course the platform is controlling everything. But to conclude this introduction, from my perspective, the most important reason to speak about this here in the IGF is that if we want and we want, if we want to protect a free, open, decentralized, neutral and unified Internet. And if we want to avoid the capture of this Internet by strong companies that are built on the Internet and that try to capture the customer. And if we want to remain free democracies, so we have a fundamental right, for example, we are in Europe, we are very attached to privacy. We have a fundamental right to say we want to protect privacy or we are not democracies anymore. So we have to impose some views to the companies that are. built on the internet. So if you want to manage all of this to protect the free and open and decentralized internet, to avoid the capture by strong actors, to build a democratic feedback, probably the design of this small layer of public services to be the interface between all those principles is a very, very good and efficient approach. But of course, and it will be said, I think, you can also have a public infrastructure that is dangerous. You can have a public infrastructure that is too much owned or controlled by the government itself, that is not transparent enough, that is, or that make mistakes with a security breach, or I don’t know. So we have to design it carefully in a multi-stakeholder way with enough transparency, if enough democratic feedback. But I don’t see any other approach than this layer of public service. If we don’t want the internet to become a kind of new far West. So that’s my view to launch the conversation.

Milton Mueller: Very good. So you have introduced something that I think is the best definition of DPI that I’ve heard, which is a set of APIs that is serving as an intermediary for a platform for many different, particularly as an interface between government services and the broader public. Let’s turn to Luca, who has, I think, some very strong perspectives on DPI based on your experience in Brazil.

Luca Belli: Not very strong. I would not say strong, but so just to provide a little bit of context. So I’m a professor at the FGV Law School, and besides directing the Center for Technology and Society at FGV, as Milton eloquently was announcing, I also direct a project called Cyber Bricks that maps and compares the digital policies of the bricks grouping. So over the past five years, we have. we discussed a lot of issues, and one of them is also amongst data governance and digital transformation AI governance, an overlapping issue is digital sovereignty. So we have a book that should already be available in open access on the Cambridge University Press websites on digital sovereignty in the BRICS. And some of the case studies we have analyzed from India or Brazil are precisely about DPIs and how they can be considered as an example of good digital sovereignty in some cases. Again, I also want to introduce a little bit of caution because DPI as digital sovereignty as pretty much anything we can speak in life, it may be a label. So to understand if it is good or bad, we have to understand what is the content behind the label, right? So DPIs as a definition, the only agreed international definition we have DPIs has been put forward by the G20 last year under the Indian presidency and has been taken off by the UN. For instance, if you look at the reports recent on DPI safeguards done by the DPI and UNDP, you will see that they quote the same definition of the G20 last year under the Indian presidency. And again, the Indians have been very successful because the DPIs for them is a key of their digital transformation strategy that endured more than 10 years under the India stack program. So let’s say that the cherry on top of their strategy was to put this into the G20. So everyone now is speaking about DPIs, which is a very successful strategy and very few countries have been this strategy and this agenda for G20 in my opinion. And so this definition is considering DPIs as digital systems that should be secure and interoperable. And they’re used to provide access to public or private services and deliver therefore not only… public services, although they are mainly used for public services, but if you look at India, there is a very good example of the ONDC, the Open Network for Digital Commerce, which is a protocol for a great bazaar where any private service can be delivered. It’s not only about public services. Usually, we categorize DPIs as digital identities, online payments, and personal data constant managers. These could fit into the public services, especially the first two, but they are not only public services. They can be used for private services. Another good example, also to stress that this does not only happen at the federal level, at the national level. In the city where I live, Rio de Janeiro, there is a very good competitor of Uber called TaxiRio. It’s a DPI made by the local municipality that allows you to take taxis without using Uber and is entirely developed by the local administration. So it’s a very good example of how DPIs can also be local. And also, why we speak about digital sovereignty not only as something that is state-driven, but can be driven by local communities, local municipalities, even local individuals. And why I argue that DPIs like PIX, our national digital payment infrastructure, is a good example of digital sovereignty because it allows, in this case, the individuals and the government to understand how the technology works, develop it, and regulate it effectively, which is how we define digital sovereignty in the book, in its various nuances. It’s all about understanding, developing, and regulating effectively technology. And so before PIX, our national payment infrastructure, was developed, the only exclusive way to process digital payments in Brazil was through Visa and MasterCard. The only way was using two foreign companies that not only charge extortionary fees between 3% and 5% in all the global South, but also they become… over the past 10 years, they became big data companies. Most of their revenues is not the 3% to 5% of the fee, it’s the intelligence and profiling. They elaborate on data collection of every single consumer that uses. People don’t understand it when they use their credit card. They are, in fact, not only paying with money, they are paying with data. And the only ones that were reaping these benefits were Visa and MasterCard before PIX in Brazil or UPI in India. So this is a very good example, in my opinion, of how BPI can be leveraged for good digital sovereignty because it broke the monopoly, actually the duopoly of Visa and MasterCard. The antitrust and competition authorities could have cried and adopted any kind of remedy, they would have never broken this monopoly, never. This has created an alternative that has reverted automatically 3% to 5% of what consumers pay in their pocket. So enormous advancement in terms of competition and consumer benefit. An enormous advancement in terms of informational self-determination or data protection. People now understand that their data are collected and who is collecting them because the online banking intermediaries, they have terms of service and precisely state what they do. When you use a credit card, you don’t even know that your data are collected. So that is an enormous advancement in terms of informational self-determination. And also, to conclude, this has also provided a very good example of how multi-stakeholder cooperation is not only about speaking in nice fora, but it also can be very outcome-oriented and very effective. The Brazilian Central Bank, which is one of the few institutions in Brazil that works very well, has leveraged a lot of public consultation with stakeholders to build PIX, is still working with all financial intermediaries to implement it, and created something called PIX Forum, which is not a nice event, but is a process to collect continuous feedback from those who are implementing the technology to understand how to improve it and what are the pitfalls of the technology. So I think that there are a lot of very good examples here. I would also provide, do I have a last minute? One more minute, yes, please. Meaningful connectivity, this has also explained the value of meaningful connectivity. In India, UPI, the Uniform Payment Infrastructure from which Pixy inspired, was a success only because India in 2016 adopted net neutrality rules that prohibited zero rating. So if you have meaningful connectivity, you can use DPIs. If you, as in all the global south, only access the internet through a meta family of apps, primarily WhatsApp, and in some cases, Facebook and Instagram, you cannot use, have access to DPIs because you pay for them. You pay for access to connectivity. So let me, the reason why India is experiencing a belly pop of innovation is because they prohibited zero rating in 2016, but that’s for the same reason, but it’s actually paradoxically the same reason why Pixy has been a success in Brazil, because when it was going to entry in force in 2020, one month before, WhatsApp was introducing WhatsApp Payment, and the Brazilian central bank know very well that if WhatsApp Payment during the pandemic had been introduced before Pix, no one here would be celebrating Pix as a success stories, but everyone in Brazil would be using WhatsApp Payment. So they suspended the entrance in force of WhatsApp Payment until Pix was also in force for consumer. Let me just explain that to be exact. In Brazil, only because the Brazilian central bank in 2020 understood that if they had allowed WhatsApp Payment to enter in force before Pix, Pix would have been absolutely useless. So before, I think this is a very cautionary tale for all governments willing to do DPIs, think systemically about everything, because if you spend a lot of money to do the best possible DPI, but at the end of the day, everyone only access WhatsApp and Facebook. it’s useless to put your investment in DPIs.

Milton Mueller: Okay, so now you understand why I said Luca had strong opinions about DPI. Let’s go on to Annelies.

Luca Belli: DPI is on many things. Just before, if I can add one word to make it even more simple. One day I was discussing with Ramod Varma, who was supposed to be there today, and he told me this very simple idea. He told me, you, in Europe, you did build your strength and prosperity, and maybe sovereignty, thanks to public infrastructure. Roads, trains, bridges. And you became very prosperous. And then suddenly you did stop. And today, in the current economy, you need ID, payment, geolocalization. Those are the infrastructures of the 21st century.

Annelies: Thanks, Sam. And Milton, do I also have strong feelings? Actually, I do. I think that DPI is a really important opportunity for us to fill gaps that date back. I mean, I see the ancestors of DPI, and I’m speaking very much from an African perspective. I’m in South Africa, but I work across the continent. Firstly, open government. One of those grand ideas and grand coalitions from about 20 years ago, which still exists, but it’s lost its glamour. It’s lost its attraction. And then the World Summit on the Information Society, where you had action lines on e-government, e-health, e-education, and enabling environment for innovation. All of those elements that were part of, and of course, connectivity infrastructure. And what we have as a result of that 20 years later is a landscape that’s completely characterized by digital inequality, which means that any initiative, be it a health initiative, or of an education initiative, or a government initiative, actually in most cases just increases that inequality. Because if there isn’t connectivity across the board, if people don’t have devices, if people in rural areas cannot afford the cost of mobile broadband, then any investment in any kind of digital public service, or in fact even in other services, commercial services, tends to just increase the gap between those who have and who can benefit and those who don’t. I think I’m not surprised that David hasn’t heard much about digital public infrastructure in Sweden, because you have digital public infrastructure in Sweden, you don’t need to talk about it. But we need to talk about it, and we need to talk about it very seriously. And I don’t think, Milton, it’s a set of APIs. It’s much, much more than that. And I think that the G20 definition, or looking at these four elements of broadband infrastructure, absolutely critical in Africa. I’m not sure how many people are aware that internet penetration in Africa is actually lower now than it was a few years ago. It’s now below 40 percent. It was around 40. It’s now more in the upper 30s. Even access to electricity is reducing in parts of Africa where dependency on hydro is affected by drought. And digital identity systems, again, even identity systems are a challenge in many African countries. And then, of course, the challenge of moving to digital identity when the data governance and the data protection frameworks and the public administration and rights protection frameworks are also insufficient in many cases. Another challenge, finance payments. How do Africans, most Africans, have financial inclusion or any semblance of it? It’s through Safaricom, it’s through mobile operators facilitating quite expensive financial services for the poor. Very few countries in Africa, outside of Nigeria and South Africa, for example, have banking. Banking that actually makes accessible, is accessible to the poor.

Milton Mueller: So let me just be clear here, Anrette, you are broadening the concept of infrastructure to everything, electricity, telecoms, banking, the whole shmear.

Annelies: Well, the G20 definition of digital public infrastructure looks at data, finance, broadband infrastructure and identity. And I think for all of those aspects to be able to be implemented and operate, operated and to become a platform for developing public services, you do need, you need not just the broadband, you also need the electricity infrastructure. Why I think it’s important and why DPI is such a good opportunity is because I think it, I mean, look, I talked a lot about how it can facilitate multi-stakeholder cooperation. That’s absolutely true. But I think even at a national level, it can facilitate intergovernmental collaboration. It can facilitate initiatives such as building, laying out broadband backbone, which is often done by the Ministry of Communications with the Ministry of Finance, Department of Finance is trying to deal with financial inclusion with home affairs. That’s trying to deal with digital identity systems. A country like South Africa, for example, pays social grants. In fact, the only reason people in South Africa don’t starve is because they get grants from the government. Administering that has been a massive challenge because there is no infrastructure to do that easily. It has been done in collaboration of the private sector, but it is extremely complex.

Milton Mueller: And I just think that I think I need to keep us a bit more focused here, because if you’re talking about the development of. telecommunications or broadband infrastructure, right? There is a completely different set of institutions and processes that have facilitated the growth of that. For example, all of these countries that had state-owned telecom monopolies had abysmal levels of penetration before liberalization and competition was introduced in the 90s or 2000s, right? You have to admit that there was public infrastructure and whether it’s power or especially telecoms, these countries that relied entirely on the state to be the developer or on international grants are simply never going to catch up with what was happening in the world where there was commercial development.

Annelies: What I’m saying is that DPI is an opportunity for countries where that gap exists to fill it. But with what capital?

Milton Mueller: With what investment? Where does the money come from?

Annelies: That is why it’s an important discussion and that’s why it’s important that we discuss it in the context of the IGF and WSIS because financing that investment and having private and public partnerships and having the kind of rights protections that need to be built into that is something that we can try and achieve through the WSIS, through the South African chairing of the G20. So I know that Jyoti wants to speak about it. The fact that it’s public infrastructure doesn’t mean that it has to be developed and owned and controlled by the state. It means that it has to be developed and controlled and regulated in the public interest. I’m not suggesting the statist approach.

Milton Mueller: Let me look at Jyoti again and then she’s been jumping up.

Jyoti Panday: These are really interesting perspectives, but I just want to, as the token representative of India and living through this great digital public infrastructure revolution, I feel like I have to chime in. A little bit going back into the history of how did the label DPI come about, and Riyad is very correct that the roots of this lie in the e-governance, the digitization initiatives that have kind of evolved with time and investment and stakeholder priorities. So it is running along the same trajectory, but for example, digital identity is rooted in the notion of national security, and therefore the state has way more say in running digital identity within the India stack than it has in matters of UPI, which has been developed by banks coming together to compete against MasterCard and Visa. So even within this whole label of India stack, which later got reformulated as DPI as they wanted to export it to other countries, the idea of digital identity is coming from a different perspective. Digital payments and wallets are coming from a different priority and perspective, and digital sovereignty seems to be the thread that ties these two different efforts and neatly brings them together under a common label, and therefore facilitates adoption, etc. I want to throw in a couple of questions here for everyone’s consideration. I like what Ambassador talked about, where he’s like, you know, your intention versus your experience, there can be a huge gap between that. So for example, when we talk about DPI and digital identity in India, one of the most common refrains I hear from everyone is that look at the scale, and look at how many people have digital identity, and these were people, large parts of the population were lacking any form of legal identity. Firstly, that claim is very, very contested, or there were various forms of identity available to the population. But of course, the state apparatus was backing this digital identity project. And therefore, one of the rationales to push it forward was to make the claim that people lack identity and therefore need a digital version of that identity to upgrade the population and bring them into the modern world into this revolution, right? But also I want to ask that after the digital identity got built, it got integrated into welfare not because this was strategic from day one, because the government and the players who were developing digital identity realized that the swiftest way of ensuring adoption and achieving that scale would be to integrate it with welfare schemes because the state has greater control over it and therefore would be able to push and apply its might behind the adoption of these services. So again, the intention, how has it reached that scale and is the scale the only parameter of success? And I think this is what Andrea is referring to when she talks about that we can’t only look at the scale and then reach and let’s talk about roads, like right, people, all countries have roads, but in some countries, the roads are excellent and talked about globally, but in other countries, they’re not that well developed. Some villages and towns in India still have dirt tracks that are referred to as roads, right? Who is responsible for building roads? It can happen both through public funding, it can happen through public-private cooperation, or private sector can be contracted to build roads. But it’s not just building roads once, roads require maintenance, roads will have to be upgraded as traffic piles up, different aspects need to be considered as the use of the road evolves. And I feel in a lot of these discussions about DPI, we get stuck at pointing to the success in India and not delving deep into that fine, no matter what the justification as to why this label has become really popular, but what are the impacts of DPI on the ground? What is the reality of the people who are actually engaging and interacting with these public infrastructure and how public are they? So, for example, the UPI API is owned by the consortium of Indian banks. And they did defer, they developed the API before, and they actually kind of didn’t allow WhatsApp to launch its payment or delayed, you know, WhatsApp adopting the protocol that was developed for interoperable payments in India. So if the idea of digital public infrastructure is that it is an alternative to big tech, monopolistic structures that are very prevalent in the digital ecosystem currently, are we replacing one monopoly with another kind of monopoly? And I would really want to hear more about these themes from our panelists.

Milton Mueller: Okay.

Annelies: I want to react to your tea, please, Milton. I agree 100% with you. And this workshop is about norms. And I think I’ve looked at quite a lot of DPI norms, the UNDP, safety guidelines, whatever they’re called. And I think none of them mention what you’ve just talked about is, how do you maintain these roads? How do you assess whether they’re getting to the people that need them most? And whether those people are actually able to find buses that they can ride on on those roads to get where they need to. I haven’t seen that level of practical looking at the sustainability and the impact of DPI initiatives. And in fact, I think if we don’t reframe how we talk about DPI, it’s just going to become another opportunity for vendor-driven pseudo-public investment, taken advantage of by corporations that are set up to do that, as has been the case with e-education, the WSIS Action Line.

AUDIENCE: I want to connect some dots. First, your former question. I think that’s of the utmost importance to understand what must be mutualized. And the difference between the, you said sometimes the public telecommunication did not reach out so much. The growing consensus within the community of DPI builders is that. we must find what has to be done by with this kind of public service approach and what must be led to the markets. So most of the DPI’s architects, they are looking for the minimum necessary infrastructure, if I may. And for example, in India, they don’t make, UPI doesn’t make the payment systems. They have more than 600 private payment systems. But built on a common infrastructure. So that’s much more efficient. Same for roads. So the importance is to have an interoperable road systems. I cannot change my car because I’m entering on another private road. So I need to be able to drive everywhere. If some of those roads are highway and I have to pay because this is a very expensive highway and others are communal small streets, it doesn’t matter. The importance is to have one unified system. And that’s the beauty of the DPI movement. We are looking for, yes, what must be done as a common and clear infrastructure for the market. Right. I think this is a well-known, although not frequently acknowledged, we tend to have this polarity, more government, more market. The fact of the matter is that markets function well when they have the right commons in which all the market players can interact. And maybe just one last thing. For most Europeans, when you say public services, you call lawyers too, because there is a long tradition of juridical definition of what must be a public service. A big service has to be neutral, equal access. You have the duty to maintain it. You cannot change your mind and say, oh, sorry, I stopped this project. There are very clear and strong definitions of public service. It can be done through private sector or not, it can be free or not, but there are some specific duties. And public service, at least in Europe, it might be public, it can be state-owned, but public servants are not governments and they have for example the right to contest the authorities because they obey to the definition of their public service.

Milton Mueller: In some countries, that is true. Luca, are you trying to get in here?

Luca Belli: Yeah, I wanted to first to attract everyone’s attention to a rare moment of agreement between me and Milton because I concur with Milton in the fact that I think it is better to keep separate the distinction between digital public infrastructure and classic infrastructure because when we speak about DPIs, it’s about digital systems built on open standards that are interoperable and secure to provide service. We might include into this telecom infrastructure, but I think it’s an overstretch. I would keep it to software rather than getting into hardware. Having said that, as I was mentioning at the very beginning of my remarks, I think one really has not only to think that DPI, although there is this nice internationally agreed definition, there are very different implementations of the concept and it’s essential to study the details. Of course, in this panel we can only have a superficial discussion about it, but let me provide some key examples between India and Brazil on the very same type of digital public infrastructure for payments to illustrate how this can be radically different. I think that the Brazilian experiment with PIX has been an evolution of what has been copied de facto from the UPI in India and actually maybe the Indians will not disclose that they have copied this from Russia because Russia introduced his system MIR after the invasion of Crimea in in 2014, when it was sanctioned immediately the day after by the US with the prohibition of Visa and MasterCard. So from night to day, Russians could not pay anything with their cards. So they had to come up with a new system, and a domestic one, MIR, like the space station, MIR. But it was a very old system based on cards, so a physical system. Now they have copied Brazil and India, and they also have a digital public infrastructure, a software one. But the origin were the Russians that developed this for an existential reasons. From night to day, you don’t have an online payment, so you have to come up with a solution. And the Indians were thinking about this, and were like, you know what? At the same time, they have disrupted Visa and MasterCard duopoly, so let’s do it. But the way in which the Indian did it, as Jyoti was mentioning, was to create a not-for-profit corporation, probably because they were skeptical about bureaucrats being able to do this. So they created the Reserve Bank of India, together with the main financial intermediary, created this National Payment Corporation of India. It’s like an icon for payments. And they did the digital public infrastructure. They created the UPI. But the reason why I argue that, actually, the Brazilian experiment went farther, because it, in my opinion, is much more transparent and accountable, is that the Brazilian experiment was directed by the Brazilian Central Bank. It’s a public institution. So if I file a request of access to information to the Brazilian Central Bank, and I ask them, tell me which data you have about me, tell me with whom you share them, tell me which are the cybersecurity standards that you adopted, they are obliged to reply to me. And if they don’t reply to me, I sue them. If I file this same kind of request for access to information to the National Payment Corporation of India, they reply to me, I’m sorry. We are not a public institution. We have no duty to reply to you. And we can keep everything totally opaque as much as we want, because this is not a public institution. So of course this could be changed in India if the National Payment Corporation of India became a public institution and so they would be obliged to be more accountable. Not in data protection terms because those who know the Digital Data Protection, Personal Data Protection Act of India know very well that all public institutions are exempted from respecting the data protection law of India. So even if it was a public institution would not be bound by respecting the protection law because data protection law in India has carved out an enormous exception for all public institutions. But that is the reason why I argue that we really should look at the details of these things because calling something digital public infrastructure doesn’t make it good or bad. We have now Bill Gates blogging about digital public infrastructure because of course this is a fashionable concept that’s been co-opted by Microsoft as usually happens with all nice concepts. Last year at IGF we were speaking about AI sovereignty. We released a book last year about this and two months after NVIDIA and Oracle started to blog about how their product were excellent for AI sovereignty which I think it’s a little bit of overstretch of the concept because of course they argue that you can be AI sovereign if you buy their tools which is again a little bit of overstretch of the concept. But again, we really have not only to look at the label but what is behind the label to understand if we can form our opinion and to understand if it’s good or bad.

Milton Mueller: Good, I need to get in here. So I really appreciated your story about the competition issue related to Visa and MasterCard in Brazil. I would tell you that within the US this is recognized as a competition issue but you need to understand how deeply embedded the dominance of those two payment networks is in our system of banking regulation. In other words, in some ways the regulatory system has created conditions which lead to an oligopoly. One of the concerns I have with this notion of sovereignty when it comes to DPI is that you could very easily recreate that situation in which local actors within a state essentially, if the DPI is not open and not standardized in a way that is facilitating competition, it can very easily lead to recreation of national oligopolies or national monopolies. One of the big competitors to a new form, emerging form of DPI, i.e. cryptocurrency, are of course the central banks. This is a very clear contradiction or clash between a sovereignty-based payment system and monetary system and a globalized one facilitated by the Internet. The other point I’d like to make here is that when we talk about interoperability and commons versus market, it seems that we’re not fully recognizing the revolutionary impact of the Internet itself. What is the Internet? It’s a set of non-proprietary protocols, right? Non-proprietary protocols that made all of this connectivity possible and it essentially facilitated a digitally networked economy, right? It wasn’t like the government said, we’re going to create a globally integrated economy and that’s why we’re funding this protocol. It was simply, the protocol was created through computer scientists and researchers and then it was, because it was non-proprietary, which is a good part about the fact that the government funded it, but it was non-proprietary so neither the US government nor any private actor controlled the use and development of the infrastructure around the Internet and one of the things we can even talk about digital payments now coming down to our phones is because of this. commons, this digital commons created by the internet protocols?

Henri Verdier: It’s real. So you’re right. Internet is based on open standards and we have to protect it. But we have also to recognize that the internet is captured because of through ID, payment and some things like this. So DPI could be an answer, but you’re right too. Maybe we should start thinking about a kind of ITU for DPI’s because interoperability, free exchange. I should be able to use my European digital wallet in India. That should be the goal. I should be able to travel, to exchange, coming from my DPI and negotiating with others.

Luca Belli: Just to complement your point because there is already an ITU for DPI which is the ITU because it has already released two years ago the GOVSTAC which is a set of open protocols for DPI’s. But what is the problem with the ITU? The ITU is an intergovernmental organization that sets norms. The reasoning behind the DPI is building the infrastructure. I honestly use this in class with students as an example of what Lessig called 25 years ago regulation by architecture because you are not regulating the market by imposing a sanction or by defining a norm. You are regulating the market by creating an infrastructure, an architecture that competes with what exists. And you’re perfectly right. I mean both of you are right. You are right because the internet is based on, it’s probably the only part of the internet that is public, is the protocols. Public, exactly. But you are right, but on that public infrastructure leading U.S. big tech has concentrated dominance and so the need for DPI is now is to revert. this concentration, it is evident, it’s quite blatant. And that is why I think that DPIs are, they could be seen, again, once again, word of caution, DPI is not always good or not always bad, but some of the examples could be seen as very good examples to reclaim digital sovereignty. So to reclaim the capability to understand how the technology functions, develop it and regulate it effectively, which is something that we have, pretty much everyone has lost over the past 20 decades because how the internet is now regulated is not through our nice laws. I say this with enormous frustration as a lawyer. The way in which the internet is regulated is not the law that I teach to school, at school, to students. I tell them, this is one vector of regulation. Reed Lessig, he was mentioning this 25 years ago, is one of the vector, is much more effective through you, through regulate, through infrastructure, what Lessig called architecture, but it de facto is what Susan Strange called structural power 30 years ago, or through marketing incentives, through subsidies and taxation. Why do we, everyone uses WhatsApp for all communications in the global South? Because it’s considered as free. It’s a marketing incentive. If WhatsApp had to be paid $100 per month, no one would use it in the global South. But the way we communicate is regulated by a subsidy that…

Milton Mueller: Like most of the infrastructure is actually created. I mean, the telecom infrastructure, the power infrastructure, that’s all. Let me get David in here, see if he has anything to say at this point. We’ve kind of left him in the dust.

David Magarde: No, no, thank you. Thank you for inviting me. So I have two things that I would like to just put on the table. So one may be interesting fact is from Sweden. So we have… We’re using BankID, which is an identity made by the corporation of the big banks in Sweden. It’s used by 99.7% of the Swedish adult population, so that is 18 to 67. So it’s of course very successful when it comes to usage and adoption. But now, the last three years, we have started work for a government-issued identity, because we’ve kind of seen that giving away all of the control of the digital identity also creates some issues. Those could be kind of on the kind of security, but also inclusion, and also the development actually of the infrastructures, because then we are from the public side and in the hands of BankID, which is sometimes good, because we have a good corporation, but sometimes it also stifles us a bit. So in my mind, that says that every situation is different, so it’s not possible to kind of have a framework that you can adopt to one country. It will differ. That’s very interesting. So I would like to hear more about this BankID offline. I also want to say the second part is that what we’re doing in EU now with the wallets is that we actually regulate open source, open standards for digital identity. And I think that the digital infrastructures, the public infrastructure, if you want to call them that, the trigger is really when you kind of make it cross-border, because then you need the interoperability, you need the kind of guardrails and the railroads and all of these things kind of working. And it’s really, really hard to do that even with standardization. We can see that with different standards and so on. But what we’re seeing now in the EU is the regulation of it. And it comes from an open source kind of idea. So I think that is very promising for making these things that you’re speaking about, being the kind of digital commons and the possibility for everyone to kind of have the same set of infrastructure and standards and so on. Yeah.

Milton Mueller: I think we’re supposed to end at 1045. So I’m going to ask now for all of you, starting with Jyoti, to provide some wrap up comments. And I want you to focus in particular on what action items you think the global community should take with respect to this issue. What can we do globally? And Amrita, I will not allow you to say that we should suddenly magically appear with $7 trillion to build broadband infrastructure everywhere. I’m going to have to ask you where that money comes from. But in terms of real feasible action items, what should we do next? Let’s start with Jyoti.

Jyoti Panday: So before I come into the action items, just briefly, one big takeaway for me from this conversation, and it’s also a question, is that which digital services qualify for the label of digital public infrastructure? Is social media DPI? Is digital identity DPI? What are the values that actually translate into something being labeled as DPI? Does it have to be open standard? Does it have to be public funded for the public part of the public infrastructure to hold? Also, I think another big point of confusion here is that the term infrastructure or architecture, as Luka helpfully referred to, has a certain notion that there is a common ground around. these ideas, right? What is happening in the case of India’s tax specifically, and maybe DPI more broadly, is that software and applications are being relabeled as infrastructure. And this is a very problematic pathway in my view, because, you know, as your digitization and your growth of your digital economy goals keep shifting, you can’t keep relabeling things that are agreed upon to suit your strategy and your convenience and to propel you into the global conversation on regulating the internet and digital services. It is a very effective strategy, but it causes confusion and it will lead to exactly this kind of fragmentation where everybody in various jurisdictions, stakeholders that have a stake or are behind certain projects, see the benefit of labeling their projects as DPI to attract more funding, to attract more mileage, to attract more attention. And we need to be wary of that. In terms of action points that we really need to focus on, I think, you know, unpacking these terms, are all digital services DPI? What constitutes, what values constitute as public of the DPI label? And what can applications and software become infrastructure? What are the infrastructural dimensions? So can you make a digital identity, the Indian digital identity, Aadhaar, mandatory for me to engage or access services on the internet? Right now in India, Aadhaar is used by the Supreme Court intervention has been restricted to services that have been paid for by the public fund of India, so our taxpayers money. But even though that decision is being completely flouted by stakeholders in both private sector and public sector in India, because they see the benefit of pushing digital identity adoption to achieve that scale. So I think drawing the guardrails and working closer to a more narrow. value-driven, precise definition is going to take this in a much more productive direction than the confusion that is currently in, you know, informing our discussions everywhere. Thank you.

Milton Mueller: Thank you, Jyoti. Let’s start with Anirudh and then just go around the circle, ending with David.

AUDIENCE: Thanks, Milton. I agree very much with Anirudh. I think digital infrastructure, my understanding, what I think we can do with digital public infrastructure is to look at it as an infrastructure that can be used to enable digitally powered public services and benefit. And I do think a value-based and narrower definition is a good way of looking at it. And, Milton, maybe those millions and millions of dollars can, and trillions can come from all that enhanced economic growth you’re going to get from your protectionist economic policies in the U.S. for years. But I absolutely am not going to take physical infrastructure off the table, because unless there is internet infrastructure, internet-enabled digital public infrastructure cannot exist. And I just, the other point I want to make is that DPI is not neutral. DPI and how it’s rolled out, how people experience it, will be shaped by the context. An authoritarian regime is going to use DPI to enhance control. And a revenue-focused, revenue-connection-focused regime is going to use it to enhance collection. And a public services, public benefit-oriented regime is going to use it to create more inclusion. And I think we cannot take that out of it. In other words, DPI is not neutral. It’s going to rise and fall in how much inclusion there is, how much oversight there is, and how much public engagement there is. But I do think it’s an opportunity to come back to looking at what we mean by internet and digital-enabled public services. To go back to all those very fragmented initiatives that emerged from open government, from WSIS, which are not connecting at all. And which are in most cases not actually creating more digital equality or digital inclusion. And I think we can use DPI to bring that conversation back to how do we collaborate? How do we integrate to achieve actual benefit for people?

Milton Mueller: Thank you, Andrea. A few ideas. First, regarding the cost, you did mention the cost. I think that Luca was right to say there is a difference between physical infrastructure and digital public infrastructure. The difference is that basically, first approximation, digital infrastructure are non-rival goods. You can have a lot of uses without scaling the cost. Of course, you have to pay the servers. I was impressed in India because basically the cost of Aadhaar is $1 per person. The cost of the French and most European digital IDs is $100 per person. So, that’s not so expensive. And the cost of UPI. So, UPI did allow more electronic transactions on US plus China together. And that’s basically a set of APIs plus a governance body. That’s such inexpensive. So, the question you did ask, what should be done now? First, like Luca too, I’m not sure we have a very precise, unified definition for everything. We have various national approaches. But probably what we should look for should be the minimum viable infrastructure that protects the free Internet, the free market. democracy. And just thinking like this, you can find some answers. And for example, you did ask, should social network be a DPI? I don’t think so. But I think that curation algorithm should, we should impose a market as a diversity of curation algorithms. We should not let 3 billion people access to information through one algorithm designed by Mark Zuckerberg, because that’s not democratic. So we should use API movements to impose a variety, to de-group, to impose a variety of algorithm. Thanks to one app store for algorithm, that social network should have to agree to accept. And we should start learning to think like this, what should be the minimum infrastructure to protect the free internet, to unleash a free market, and to protect democratic feedback. And when you start thinking like this, you find sometimes some, maybe not answers, but propositions of answers. Thank you, Ambassador. We’ll go to Loka next. And again, action items, if you can.

Luca Belli: So again, I think that there are drawing conclusions from what we have discussed. And from what I was trying to convey, there are multiple layers that we should consider in this quite complex situation. First, we should not think that DPI is bad or good, but analyze how it is proposed, and how I’m speaking about the how, how it emerges as a discussion, because I’m very concerned, for instance, that now, IMF is imposing digital identity as a counterpart for having loans, because this is an imposition, is not something that is emerging as a bottom up process from the country that wants to develop it and knows how to develop it, but is being imposed and likely adopting another type of technology that does not necessarily fit the specific reality. So you were speaking about Visa and MasterCard being part of the U.S. let’s say ecosystem that may work very well in the U.S., but we have also to acknowledge that, I don’t know, I’m not an expert in the U.S. I am needed and it is not a country I studied too much, but every country is very specific. So there is no, first, there is no silver bullet. So if we cannot impose, that is probably the reason why the ITU GovStack doesn’t work very well, because we cannot impose or hope that everyone will magically adopt a fantastic solution. It’s much better to bet on the fact that those who are ready and want to do this might be able to develop it. And here, my second point, institutions that drive this process are very important. The Brazilian example, I want to stress this, is very special example driven by an institution that, although it is in an emerging developing economy, it’s very well resourced, has very good command and control capabilities, and understands systemically how to do this. And they show that the PIX system is a success because they’ve understood all the layers of complexity, starting from the access, meaningful connectivity, or the lack of meaningful connectivity in Brazil. Because the reason why they postponed WhatsApp payment is because they are well aware that Brazil is not meaningfully connected. 22% of the population has meaningful connectivity, 78% of the population is connected only through WhatsApp and Facebook. So that is a very important point, because if we are investing a lot of money in a digital public infrastructure and everyone will only use Facebook and WhatsApp, we are wasting our money. So we have to think systemically. And last but not least, I think that this also gives a lot of hope to those who have been over, like some of us, for 15 or 20 years about speaking about multi-stakeholderism. It’s a very good example of how multi-stakeholderism… and can go beyond nice chats in exotic places and being very outcome-oriented and effective when it is well orchestrated.

Milton Mueller: So more bottom-up where it’s ready, no top-down.

Luca Belli: Understanding the local realities and engaging the stakeholder in the most effective way.

Milton Mueller: And good, but recognizing that when things evolve in that way, they tend to be, could be fragmented, right?

Luca Belli: Yes, but in this, they could be fragmented, but if we cannot think that to fight against fragmentation, we impose concentration and accumulation of wealth by only three or four enormous tech companies that actually are not even taxed as they would be. So if you want to find the money to buy all this, start to tax Google, Facebook, Amazon, that have reaped enormous benefits since the pandemic, but have never been taxed properly. I mean, there are enormous reports about only Brazil, India, and Indonesia lose $4 billion per year in tax evasion. So if you want to find the money, let me tell you, my friend, you can find it.

Milton Mueller: Four billion might finance the broadband in what, two cities, but anyway. But David-

Luca Belli: I’m not speaking about broadband, I’m speaking about software, much cheaper, as Henry was saying.

David Magarde: Yeah, sure. I think a bottom-up approach will give the, I mean, there are already key players here that kind of control most of the environment, so they will effectively control more of the environment if it’s only bottom-up. So I think regulation needs to be there, some institutions, but also with, I know that ITU libraries and so on may not be fully used as was hoped by some, but I really do like the idea that we have libraries with open standards. source code and so on that can be used because that’s also drives innovation and ensures that the civil society but also the private companies can look into what’s happening at the public sector side so that the government can don’t take control of all of the relevant infrastructures because I think to develop an inclusive digital infrastructure needs to be cooperative between the public private sector but also needs to be transparent and ensure that everyone can kind of everyone with skills can go in there and look what’s actually happening and also I was surprised that India have excluded their public sector from the data protection side because that is one thing that I think is working quite well in EU when it comes to ensuring kind of transparency and openness on what of the data which is of course what’s the most core to the people their data or data about them. So I would say regulation and kind of openness and providing like tech stacks for everyone to use and then we will have some players or some areas or nations that will go before the other ones and hopefully they will do a good job so that the other ones can kind of jump into that train on that train and then we can see kind of interoperable basic infrastructures that people can build up using their kind of natural context to that and that in itself I hope would enable people to travel more freely both digital and physical using their kind of uncontrolled digital identity which is the key enabler for most of these things.

Milton Mueller: Okay so I just realized that I’ve totally excluded the audience from the discussion it’s because so much vigorous stuff was going on here but it was hard to to get in, and I’m sure we’d be talking for another hour if I didn’t rein them all in. So, I would encourage you that if you’re here to approach the speakers, take off your headphones and talk to people about some of their ideas. I think we have about five minutes. Two minutes. We have eight minutes? Seven. Seven minutes. Let’s unleash the audience. All right. Here, we’ve got somebody. You’ll have to give him a microphone.

Kashfi Inua: Thank you very much. My name is Kashfi Inua from Nigeria. Quite insightful and interesting conversation. But my question is, why is it it’s only in the global south we talk about DPI? In the global north, nobody talks about DPI. And looking at how they evolve in terms of building the infrastructure, I am slightly lost to what is happening in the enterprises. By having a backbone. And through that backbone, they are connecting everyone. So, sometimes I just get confused. Let’s talk about DPI, global south. Thank you.

AUDIENCE: So, first, I don’t think there is such thing as a global south. There is a huge diversity in the south and in the north. But I work on this question for a long time. Your question, because the truth is, as it was said, we have most ancient and complex systems that is working a bit like a DPI. For example, in France, you can buy a baguette with your credit card without paying any fee to anyone. So, you don’t really need UPI because you have this system. But our system is very ancient. with a strong legacy, very expansive, but the need to change is not the same. But as I said, you have X-Road in Estonia, you have France Connect in France, you have pieces of DPI, but they were designed before the DPI movement.

Milton Mueller: Luca?

Luca Belli: Just a quick comment on this, because I think that the Global South has realized that there is a problem. We have been digitally colonized, and I think that the only thing that is in common in the Global South, or the global majority, they are extraordinarily diverse countries, but the only thing they have in common, they have been colonies. So they understand that they are colonized again. The Global North is used to be the colonizer, and they are not being used to be colonized, now they are realizing it now. Let me tell you, I was invited, I was at the European Parliament in September to present the research on digital sovereignty in the BRICS, because now the Global North is studying the Global South, because there are a lot of things to be learned from India, Brazil, China, South Africa, in terms of how to react to this kind of digital colonialism that is, of course, ongoing, and from which very few countries, I would say only one, is benefiting, and the others are not benefiting at all. So the others have a problem and need to find a solution.

Milton Mueller: I would want to challenge-

Annelies: Simply, you are right. We speak about it in the Global South because we don’t have infrastructure. You don’t speak about it as much in the Global North because you have infrastructure. You know what it feels like to go to the Netherlands and they complain that their trains are late, when in South Africa, getting a bus is impossible. It’s a vast difference, and I think that the sad thing is that these conversations replicate. During the structural adjustment era, when we were trying to invest in public education and public health, we were told by the international financial institutions and the Global North, don’t, no, don’t, let the market do it. When we talked about building broadband capacity, we’re told by Milton and other people, let the market do it. The market fails to do that. I’m not saying the market is to blame for that. I think there are a whole set of complex factors. But so that’s why we keep having these conversations. And I think the challenge is, are our governments having them in the right way, in the most effective way? Are financial institutions responding in the most effective way? But I think that’s very clear why we have this conversations. I think we just need to use these conversations better, be more action oriented and more context oriented in how we actually try and redress these gaps in infrastructure that we are talking about.

Milton Mueller: I don’t think any of you actually answered his question, which was like, I mean, partly you did. You’re saying, yeah, we don’t have infrastructure. None of you have dealt with the question of why they don’t. Now, part of it is the legacy of colonialism. And by that, I mean real colonialism, not this fake thing called digital colonialism in which you’re simply talking about the advancement of technological capabilities and their extension from the leading countries. But a lot of it is also the fact that your institutions have not allowed for the market development that is possible in other countries, right? You have telecom monopolies still in many African countries still stayed on telecom monopolies. You have, you know, it’s illegal to have voice over IP in many of these countries for a very long time. All the forms of competition and innovation that were enabled by digital technology have frequently been stifled by extremely authoritarian or protectionist governments that will not allow the infrastructure to develop.

AUDIENCE: I just like to jump in here. I think Yoti has something to say, but I was also gonna direct my next question to you. And it builds off of this, if you’d like to just take that. I think what he is trying to say and what you were trying to say about the global South differences, I think what we actually have to acknowledge and what we publish in our own research is that these are places called leapfrog regions and that they exist in the global North. You can look at Flint, Michigan. You can look at different parts of Atlanta, New York. there are regions that when COVID hit, we saw that there was not digital public infrastructure in the North as well. And so we need to acknowledge that digital public infrastructure does go across both sides. And so my question would be, I think, when you talk about authoritarian governments and monopolistic telecoms companies, I think what we feel in the Global South is that those companies and that structure is not structured by our governments, it’s structured by outside forces. And so what can we do as parties in the Global South, like I come from Northern Ethiopia, he comes from Nigeria, I’m not sure where you come from, but I think what I’m curious about to hear, and Jyoti, I’d love to hear you start off, is how do the youth from these places, how do we decide to, like, we’d like to take control of our own infrastructure, and how do we leapfrog past what we see in front of us, because it’s highly, highly unfair.

Jyoti Panday: Yeah, thank you for these amazing questions from the audience. And I have a very skeptical view of this. To the first gentleman’s question as to why we hear this label more from Global South and not so much in Global North, is because the label has actually been developed in the Global South. And, you know, these advanced nations with their advanced technological capabilities are, you know, very of adopting terms that they have not been involved in directly, right from the outset, been dictating terms of their development. So it’s like, it’s a little bit coming from geopolitical competition around tech policy, but also, you know, people are wanting to see how this label evolves, what is the impact, and then when they see it is a great success. So to Luca’s point, why Gates Foundation or the IMF are now suddenly using these labels, once they become, once they pick up, once they become popular, and they see the benefits of these labels, they will start using it more. In terms of how do we leapfrog, I want to give an example from India on broadband, actually.

Milton Mueller: Jyoti, we’re about to get kicked out of the room. It’s 11 o’clock.

Jyoti Panday: So Reliance brought the data rates completely low in India, and everybody clapped and said this is wonderful. But now, 10 years down the lane, we see that the investment in actual internet infrastructure has dwindled down because we have, there’s a monopoly or the, you know, there are only two telecom operators in India. So we have to take these decisions carefully. There are always trade-offs involved. And it’s, I can, I’m happy to discuss this offline. I don’t want to hold anyone in the room. But thank you again for the questions and for all your inputs and for making the time to all the panellists and the speakers. Thank you. Thank you again very much for being a part of this.

Milton Mueller: Thank you.

Agreement Points

Need for a clearer definition of DPI

Jyoti Panday

Luca Belli

Annelies

DPI as approach for building large-scale networks, platforms and services essential for digital economy

DPI as digital systems built on open standards that are interoperable and secure to provide services

Need for narrower, value-driven definition of DPI

The speakers agree that there is a need for a more precise and value-driven definition of DPI, focusing on its role in building essential digital infrastructure and services.

Importance of institutional frameworks and governance

Jyoti Panday

David Magarde

Henri Verdier

Importance of institutional frameworks and oversight mechanisms

Need for regulation and openness in DPI development

DPI as minimum necessary infrastructure to protect free internet, market and democracy

The speakers emphasize the need for proper institutional frameworks, oversight mechanisms, and regulation in the development and implementation of DPI to ensure openness and protect democratic values.

Similar Viewpoints

Both speakers view DPI as a tool to promote competition and protect democratic values in the digital economy.

Luca Belli

Henri Verdier

DPI can break monopolies and increase competition, as with payments in Brazil

DPI as minimum necessary infrastructure to protect free internet, market and democracy

Both speakers highlight the potential risks of DPI, emphasizing that its implementation can lead to centralization or be used for control depending on the context.

Jyoti Panday

Annelies

DPI can lead to centralization and compromise digital security

DPI is not neutral and can be used for control or public benefit depending on context

Unexpected Consensus

Global South leading in DPI development and discourse

Luca Belli

Jyoti Panday

Annelies

Global North studying Global South approaches to digital sovereignty

DPI more discussed in Global South due to lack of existing infrastructure

There is an unexpected consensus that the Global South is leading in DPI development and discourse, with the Global North now studying and learning from these approaches. This represents a shift in the traditional flow of knowledge and policy ideas.

Overall Assessment

Summary

The main areas of agreement include the need for a clearer definition of DPI, the importance of institutional frameworks and governance, and recognition of both the potential benefits and risks of DPI implementation. There is also consensus on the growing importance of Global South approaches to DPI.

Consensus level

Moderate consensus with some diverging views. While there is agreement on the importance of DPI and the need for proper governance, there are differing perspectives on its implementation, funding, and potential impacts. This suggests that further dialogue and research are needed to develop a more unified approach to DPI development and implementation globally.

Different Viewpoints

Definition and scope of Digital Public Infrastructure (DPI)

Henri Verdier

Luca Belli

Jyoti Panday

David Magarde

DPI as minimum necessary infrastructure to protect free internet, market and democracy

DPI as digital systems built on open standards that are interoperable and secure to provide services

DPI as approach for building large-scale networks, platforms and services essential for digital economy

DPI not commonly used term in EU, focus instead on interoperability

Speakers had different perspectives on what constitutes DPI, ranging from a minimal infrastructure approach to a broader definition encompassing various digital systems and services.

Role of government and market in DPI development

Milton Mueller

Luca Belli

Henri Verdier

Risk of DPI leading to fragmentation if pursued with sovereignty-based approach

Bottom-up approach needed, with stakeholder engagement based on local realities

DPI as minimum necessary infrastructure to protect free internet, market and democracy

Speakers disagreed on the extent of government involvement in DPI development, with some advocating for a more market-driven approach and others emphasizing the need for government intervention.

Unexpected Differences

Global North-South divide in DPI discussions

Luca Belli

Milton Mueller

Unknown speaker

Global North studying Global South approaches to digital sovereignty

Institutional barriers to infrastructure development in some countries

Need to consider “leapfrog regions” lacking infrastructure in Global North as well

While most speakers focused on the Global South’s need for DPI, an unexpected perspective emerged highlighting similar infrastructure gaps in parts of the Global North, challenging the traditional North-South divide in digital development discussions.

Overall Assessment

summary

The main areas of disagreement centered around the definition and scope of DPI, the role of government versus market forces in its development, and the global perspective on DPI needs.

difference_level

The level of disagreement was moderate to high, with significant implications for how DPI might be conceptualized, implemented, and governed globally. These differences suggest that achieving a unified approach to DPI development and implementation may be challenging, potentially leading to varied strategies across different regions or countries.

Partial Agreements

These speakers agreed on the need for stakeholder engagement and oversight in DPI development, but differed on the balance between bottom-up approaches and regulation.

Luca Belli

David Magarde

Jyoti Panday

Bottom-up approach needed, with stakeholder engagement based on local realities

Need for regulation and openness in DPI development

Importance of institutional frameworks and oversight mechanisms

Similar Viewpoints

Both speakers view DPI as a tool to promote competition and protect democratic values in the digital economy.

Luca Belli

Henri Verdier

DPI can break monopolies and increase competition, as with payments in Brazil

DPI as minimum necessary infrastructure to protect free internet, market and democracy

Both speakers highlight the potential risks of DPI, emphasizing that its implementation can lead to centralization or be used for control depending on the context.

Jyoti Panday

Annelies

DPI can lead to centralization and compromise digital security

DPI is not neutral and can be used for control or public benefit depending on context

Key Takeaways

There is no clear consensus on the definition and scope of Digital Public Infrastructure (DPI)

DPI can potentially break monopolies and increase competition, but also carries risks of centralization and compromising digital security

The implementation and governance of DPI should involve multi-stakeholder processes and consider local contexts

There are differing perspectives on DPI between the Global North and South, partly due to existing infrastructure gaps

Funding and institutional frameworks remain key challenges for DPI development in many countries

Resolutions and Action Items

Work towards a narrower, value-driven definition of DPI

Analyze how DPI is proposed and implemented rather than labeling it as inherently good or bad

Pursue more bottom-up approaches to DPI development where countries are ready

Develop open standards and libraries for DPI that can be used by different stakeholders

Study successful DPI implementations in countries like India and Brazil to learn lessons

Unresolved Issues

How to define which digital services qualify as DPI

How to balance sovereignty-based approaches with the need for global interoperability

Where funding will come from for DPI development in resource-constrained countries

How to address institutional barriers to infrastructure development in some countries

How to ensure DPI promotes inclusion and public benefit rather than control or surveillance

Suggested Compromises

Focus on developing the ‘minimum viable infrastructure’ that protects free internet, markets and democracy

Combine regulation with openness and transparency in DPI development

Balance country-specific DPI approaches with efforts towards cross-border interoperability

Consider both software-based DPI and physical infrastructure needs in developing countries

DPI movement is one new world for various approaches. And you can connect it to infrastructures, to public service, to sometimes digital commons, to platform strategies.

speaker

Henri Verdier

reason

This comment highlights the broad and evolving nature of DPI, showing it’s not a single defined concept but encompasses multiple approaches and ideas.

impact

It set the stage for a nuanced discussion about what exactly constitutes DPI and how it manifests in different contexts.

The Brazilian experiment with PIX has been an evolution of what has been copied de facto from the UPI in India and actually maybe the Indians will not disclose that they have copied this from Russia because Russia introduced his system MIR after the invasion of Crimea in 2014, when it was sanctioned immediately the day after by the US with the prohibition of Visa and MasterCard.

speaker

Luca Belli

reason

This comment provides important historical context and shows how DPI initiatives have evolved and spread globally, often in response to geopolitical pressures.

impact

It broadened the discussion beyond just technical aspects to consider geopolitical and historical factors shaping DPI development.

DPI is not neutral. DPI and how it’s rolled out, how people experience it, will be shaped by the context. An authoritarian regime is going to use DPI to enhance control. And a revenue-focused, revenue-connection-focused regime is going to use it to enhance collection. And a public services, public benefit-oriented regime is going to use it to create more inclusion.

speaker

Annelies

reason

This insight highlights that the impacts of DPI depend heavily on the motivations and governance structures implementing it.

impact

It shifted the conversation to consider the critical importance of governance frameworks and oversight in DPI implementation.

We have now Bill Gates blogging about digital public infrastructure because of course this is a fashionable concept that’s been co-opted by Microsoft as usually happens with all nice concepts.

speaker

Luca Belli

reason

This comment raises important questions about the co-opting of DPI concepts by large tech companies.

impact

It prompted discussion about potential conflicts between public interest goals of DPI and commercial interests.

One of the concerns I have with this notion of sovereignty when it comes to DPI is that you could very easily recreate that situation in which local actors within a state essentially, if the DPI is not open and not standardized in a way that is facilitating competition, it can very easily lead to recreation of national oligopolies or national monopolies.

speaker

Milton Mueller

reason

This comment highlights a key tension between national sovereignty and open, competitive systems in DPI development.

impact

It sparked debate about how to balance national control with openness and competition in DPI.

Overall Assessment

These key comments shaped the discussion by highlighting the complex, multifaceted nature of DPI. They moved the conversation beyond technical definitions to consider historical context, geopolitical factors, governance implications, and potential pitfalls. The discussion evolved to grapple with tensions between national sovereignty and global interoperability, public interest and commercial motivations, and the need for both government involvement and market competition in DPI development. This nuanced exploration revealed DPI as a concept with significant implications for digital governance, economic development, and global power dynamics in the digital age.

Which digital services qualify for the label of digital public infrastructure?

speaker

Jyoti Panday

explanation

This is important to clarify the scope and definition of DPI, as there is currently a lack of consensus on what should be included under this label.

What values should translate into something being labeled as DPI?

speaker

Jyoti Panday

explanation

Understanding the core values and principles that define DPI is crucial for developing consistent frameworks and policies.

How can we ensure interoperability between different DPI systems across countries?

speaker

Henri Verdier

explanation

Interoperability is key for enabling cross-border use of DPI and preventing fragmentation.

How can we balance the need for digital sovereignty with international cooperation in developing DPI?

speaker

Luca Belli

explanation

This tension needs to be addressed to prevent fragmentation while allowing countries to develop context-appropriate solutions.

How can we ensure transparency and accountability in DPI systems?

speaker

Luca Belli

explanation

Transparency is crucial for building trust and enabling proper oversight of DPI systems.

What role should regulation play in the development and governance of DPI?

speaker

David Magarde

explanation

Understanding the appropriate balance between regulation and market-driven approaches is important for effective DPI development.

How can we address the digital divide and ensure meaningful connectivity alongside DPI development?

speaker

Luca Belli

explanation

Ensuring widespread access to connectivity is crucial for the success and equitable impact of DPI initiatives.

How can developing countries take control of their own digital infrastructure development?

speaker

Audience member

explanation

This is important for addressing concerns about external influence and enabling context-appropriate solutions in the Global South.