IGF 2024-Day 1-Workshop Room 4- WS43 States and Digital Sovereignty- Infrastructural Challenges

The following are the outputs of the captioning taken during an IGF intervention. Although it is largely accurate, in some cases it may be incomplete or inaccurate due to inaudible passages or transcription errors. It is posted as an aid, but should not be treated as an authoritative record.

***

 

>> MODERATOR: Hello, good afternoon. Good afternoon. Hello. One, two, three.

(Pause).

>> MODERATOR: Welcome, everyone, to the workshop states and the digital sovereignty infrastructural challenges.

I am Rodolfo Avelino, from the Steering Committee and I will be moderating this session.

We also have here Juliana Oms as online mediator and Korstiaan Wapenaar. Both are technical advisers for CGI and BR. I would like to thank the organization and everyone that presents today. My special thanks to our speakers that we will contribute to our debate.

The development of the Internet are marketed by the consolidations of large digital platforms and the growing use of artificial intelligence. This causes significant changes not only in social process, but also in public services such as health, education, communication, and the states capacity in general.

Despite technology advances, problems of the national security, surveillance, and alternative to implement digital policies arise. These issues can be framed under the conceptions of the digital sovereignty, a notion that can have multiple meanings and purposes, raises things such as the security of digital infrastructural, the security of the data and the data capacity to guarantee fundamentals rights.

This session aims at discussions policies to implement digital infrastructures in different regions and countries in light of different approaches to digital sovereigns.

I hope we have a great conversation and that which experience may serve as inspirations for others.

Now I would like to give the floor to our speakers. Our first presentation will be delivered by Min Jiang. We will start our conversations by the different conceptions of the digital sovereigns. Dr. Jiang, you have the floor for eight minutes, please.

>> MIN JIANG: Thank you, colleagues, at CGI Brazil for convening this session and for inviting me to join. Can you hear me all right? Just double-checking. Okay, great. Thank you so much.

And greetings also to participant from around the world. My contribution to the panel is based largely on a book I co-edited with Dr. Luca Belli of law school. Can I have the slides up please. Can you move to the next slide, please.

Hello, can -- move to the next slide please. Yes. Thank you so much. So the book is titled digital sovereignty in the bricks countries, how the Global South and emerging power alliances are reshaping digital governance coming out in two weeks through Cambridge University Press.

I will develop my remarks today in two parts. First I will trace the development of digital sovereignty and explain why digital sovereignty is gaining currency. Secondly I'll offer a framework for conceptualizing digital sovereignty between a traditional normative definition of sovereignty centered around nation states.

If fact, I would argue digital sovereignty is not something that belongs to the states only. Instead, digital sovereignty as broadly conceptualize is the stakeholder model by foregrounding the stakeholderism to be more widely adopted.

To start, next slide, please.

Given the IGF -- sorry, one slide back.

To start, given the IGF is a global forum under the auspice of the UN, it's appropriate to recognise that the UN is a post-World War II creation based on national independence and sovereignty. The idea of sovereignty which can be traced back to the 16th century as well as the 1648 piece of west Philia is foundational to the western states.

However, all of us also recognise that such normative and very idealistic notions of sovereignty are often good on paper, but not so good in practice. Sovereignty is frequently a function of power, strong states, for example, can invade other states, think Iraq war and the current war in Ukraine. Weaker states lack power to exert influence.

Next slide, please.

The problem of power imbalance is especially evident in the digital era where much of the world's digital infrastructure, data, services, and increasingly AI depends on a handful of Silicon Valley firms. Snowden's representation of the program in 2013 made it clear that the U.S. Government cannot be trusted.

Facebook and the general failure of U.S. big tech in the 2016 U.S. presidential election also made it clear that U.S. big tech cannot be trusted.

While sovereignty was once thought to be an authoritarian product shipped out of China, post Snowden it's not surprising why more and more countries including EU as a coalition of post -- of coalition of post nations are picking up the banner of digital sovereignty. And in fact, EU means by digital sovereignty self-determination to voice their dissatisfaction.

It's also not surprising why ICANN was pressured to move out of the U.S. commerce department in 2014. Whether the effort came from individual states, group of states, or multi-stakeholder fora, they share one thing in common. As the global consensus of U.S.-centered global digital order is breaking down national and international actors are in search of alternative to build a new digital order.

Next slide, please.

The moment we're in is not unlike the new world information and communication order debate in the 1970s.

Next slide, please.

Hello. Next slide, please.

>> RODOLFO AVELINO: One minute, please.

>> MIN JIANG: Thank you. The new Wicel debate in the 1970s published by UNESCO in 1980, at the time these were the exact phrases from the report, were of concern to Global South countries which were critical of the free flow of information agenda, championed by the U.S. and UK seen as an instrument or information colonization and cultural imperialism. In the end, the U.S. and UK pulled out of the debate arguing global south countries used information sovereignty and cultural sovereignty as a pretext for censorship at home.

The power symmetry mirrored the power symmetry today. While state perspective remains essential to understand digital sovereignty, many researchers including myself also recognise digital sovereignty is a signifier, is a term with many different meanings used by different actors to express their aspirations and also to assert control and power.

Thus, in the book volume, we adopted a more generative addition of sovereignty as an exercise of agency power and control in shaping digital infrastructure, data, services and protocols. The book's project bottom-up efforts led us to develop a broader framework of digital sovereignty mapping the following perspectives.

Next slide, please.

In the state of digital sovereignty perspective, nation states exert control over digital architecture, data, protocols and services. It can be both positive and active. While Brazil, for example, build the digital payment system during the payment and India build the UPI digital payment system as digital financial infrastructure to increase independence and inclusion, Russia on the other hand built the RUNET for digital isolation.

In the sovereignty perspective, regional alliances like the EU develop unified digital policies as well as legal and digital infrastructure. Former German chancellor Angela Merkel in fact gave a speech at the 2019 IGT defining EU's digital sovereignty as a form of self-determination. So for EU countries being sovereign doesn't mean -- doesn't mean, you know, working alone, but working together.

Network digital sovereignty is something ICANN as an organization might endorse, and emphasizes decentralized control, network interoperability, freedom from nation states and global coalition.

Corporate digital sovereignty tends to endorse the freedom of tech giants in driving digital economies and shaping digital norms, something scholars have critiqued to be a form of surveillance capitalism or data colonialism.

Personal digital sovereignty exercises empowerment of individuals to control their digital ideas. Postcolonial sovereignty is efforts by former colonized nations to reclaim autonomy access in the digital space. Finally, common sovereignty, exercises community-driven governance of shared digital resources and production of public digital goods.

This is embodied in the open source free software movement as well as international digital solidarity and labor movements.

So for us, digital sovereignty is not something that belongs to the nation states only. Broadly conceptualized, digital sovereignty is not on the stakeholder model, it foregrounds the underlying power issues. Global sovereignty shouldn't be determined by national governments alone, but needs a global coalition to address pressing issues of global divide and digital infrastructure, entrenched challenges of censorship as well as structural monopoly and digital inequality.

Thank you for allowing me to share my perspective. Look forward to further exchanges and discusses.

>> RODOLFO AVELINO: Thank you. Thank you, Dr. Jiang.

Now producing the commissioner of the International Communication Commission since 2021. The commissioner has 15 years of professional experience in the telecommunications field.

Mrs. Catherine, you commented on the [?] government adapted to undergoing digital transformations and how these projects relate to broader European context.

>> CATHERINE: Thank you very much. I think my presentation will be put on now.

I want to thank hosts of this very interesting, very important, very, very, very much relevant topic of workshop. This is information in Coordination Center of Brazil, also I think everybody will join me thanking the Host Country, Saudi Arabia, Riyadh, for this amazing venue for the IGF. And it's a great pleasure to share a perspective from the other -- another part of the world about how the state can see very small, tiny state in the south caucuses can see the challenges and overcome the challenges related to digital sovereignty and previous presenter very in best manner outlined what are the layers of the digital sovereignty and how it has evolved.

And I will be speaking about the very specifically telecom layer, the most upstream perspective of the infrastructure and challenges related to my country, Georgia, and it is also related to the region itself. The south caucuses.

So my perspective as representative of the telecom regulator and state representative will be specifically to the very upstream layer of the infrastructural challenges. I'm trying to now move on.

Thank you. So my friend from south caucuses is helping me as we do usually. Thank you.

Oh, thank you. Thank you. Thank you so much.

So we all know that most part of the data as we speak about digital, most part of the data traffic goes through the submarine cables, you know, and how important submarine cable resilience and resilience levels are.

We see that this information was kindly provided by our -- one of our important partners which is World Bank. So I'm allowed to show this information. But it's actually this first slide is available information you can find and it's updated on a daily basis almost.

And you see this is a most upstream layer of the Internet. And this is a value chain. This is the growing market. And this is the very -- how to say -- the basis of our connectivity that enables us to exchange data, to protect data, and this many cables are also created to ensure that the data exchange is resilient.

So the sovereignty information on the infrastructure level is pertained.

If we can go to the next slide. It's a bit of challenge. Thank you.

So now let's speak about the very tiny segment of this big map of worldwide connectivity. Which is South Caucuses where my country Georgia is located. And you can see the map, you can see the geography, and you can see that from this, like, almost 500 infrastructure connectivity routes, only one route is connecting Georgia and South Caucuses with Europe.

And this connection is very important. Direct connection, you understand how important it is if we speak about kind of the infrastructure level resilience and sovereignty of data based on that.

And there are -- the aspects of resilience which is the -- providing direct international access, not through some other jurisdictions, but through the sea. This is what makes the submarine cable so important nowadays. And development of more interregional networks is obvious, necessity of development of more interregional networks is absolutely necessary for our region.

This is the challenge we are faced now. And fortunately, we have partners who are supporting this with making this connectivity and resilience really to work.

I will speak later on that.

And what else is happening on our region is that there is a project of trans-Caspian cables under discussion which will connect us further to the East Asia or Central Asia, which is we know that digital is interconnected.

So we need to be part of global, resilient connectivity paths.

And with this, not only the infrastructure level resilience and connectivity -- direct connect corridors come to our mind, but also the layer that we're (lost audio).

So what else comes to mind after seeing that there is definite need of expanding the infrastructure level interdependence of the region, it comes also the layer of services, software, and data protection. With data related to resilience layer, we will speak about the digital concept which is also crucial for our region.

Which is.

>> Hello. Check.

>> CATHERINE: Some kind of check. Next slide, please. Thank you.

So why we need some kind of interregional digital hub. If we speak about the concept of protecting our data, we see that concerning the geopolitical situation around south caucuses, you can see on the map how important it is to have the connectivity -- some kind of interregional connectivity hub that will enable us having the transfer and data protection frameworks which is -- which is aligned with the EU -- GDPR's EU's data protection legislation, and that will allow to us have some kind of protected sovereign data transfers throughout the -- throughout the region.

Another important aspect why we need to have data perhaps in region is that upcoming technological demands related with AI definitely requires that information should be brought closer to the customers. So this is another challenge, and this is another condition to building interregional connectivity hubs. Which will involve our regional countries, but which will create some kind of alternative to overcome choke points like you see -- that you see in the Red Sea region.

And when we say about the challenges, challenges bring usually the opportunities, so my presentation here was brought to show you that we're trying to turn these challenges into the opportunities for our region. And as I mentioned in the beginning, if we can move to the next slide, thank you.

So there are some kind of articles about what are the challenges, how the politics and geopolitics are shaping the Internet now and what is going on around the world, it's very much relevant to our region.

And as I mentioned to this concept of south caucuses digital hub to make our data more resilient is kind of the answer to the question how we can build more robust digital service. So digital layers starting from the upstream infrastructure up to the software and to the data protection layers.

And you can see this Baltic Highway project which is supported by European Union, and we also are supported by World Bank and by European Union to build similar regional connectivity corridors that will enable countries in the region to be connected safely with other worlds and to bring the data closer to our subscriber and to ensure that policies and regulations that are adopted amongst the European Union to be transposed to the region at data hubs.

This is how we see answering to the challenge. With that, I've seen of course this is now in projection stage. Projection stage means that we have some concept how this data center regional hub concept should work. And we really hope that it will be continued and it will turn, because the adoption of AI and growing demand on the machine learning or bringing the more, like, content into the digital space gives us understanding that this project should be elaborated as soon as possible.

So this is what I want -- I wanted to share with you. And I'm happy to answer later the questions. Thank you.

>> RODOLFO AVELINO: Thank you. And now introducing Korstiaan who is a principal at the center of excellence in Johannesburg. He develops strategies to Africa's developmental challenges. Korstiaan will comment on the capacity limitations in Africa, different steps on local/private, and the international players connecting to the tradeoffs of sovereignty.

You have the floor.

>> KORSTIAAN WAPENAAR: Thank you very much, colleagues, for having me. And for the opportunity to participate.

Can I ask -- okay. Great. Thank you very much.

There have been a couple version changes this morning, so there might be a couple of edits that might not have come through, but we can run with it.

So I think the point of departure to start is that digital transformation of the public sector is a prerequisite for socioeconomic development in Africa. African states have struggled to deliver services to people in organizations at scale, and subsequently these technologies allow them to reach people in need at scale when they need.

Unfortunately, despite this prerequisite, let's call it, African countries have largely struggled to deliver on the opportunities of digital transformation. The government development index is a useful proxy for that, and we see that there are only four African countries that have managed to achieve above the global average.

And so there are some critical underlying drivers of this kind of performance. In particular, one is acknowledging that many African countries have significant fiscal constraints and significant capacity constraints in terms of their expertise. And that subsequently this impacted the rollout of quality of hard infrastructure if we can provide data centers and the like as well as soft infrastructure being the services that -- the technologies that are used to deliver services through this hard infrastructure.

Next slide, please.

If we look at data centers as a proxy for the availability of infrastructure across the continent, we see that there is a rapidly growing demand for more physical infrastructure. Apologies that there's no access on this graphic, is that African countries will need to more than double their data center hosting capacity by 2030. At present, though, a number of these countries are underdeveloped, there's not a lot of digital activity, and so localization requirements are hard to meet through local data center because it's economically infeasible to host this domestically just to meet those requirements.

Next slide, please.

Subsequently -- one more slide, please. Yeah.

Subsequently, we've seen that a number of markets across the country governance have started where experiments were deploying their own data centers to manage their own data and operate their own technologies and infrastructures. In many cases, though, due to the capacity limitations, these are fully managed. They're underutilized, and they have become what is termed economic drains, maybe one might call it a white elephant or the like.

And it leaves a little bit of a quandary for African countries that are trying to achieve localization requirements independently and autonomously.

Next slide, please.

So subsequently what this means is that there's an inherent dependency in Africa on maybe inherent is a bit of a strong term. But there's a mutual benefit between the state and the private sector in delivering this hard infrastructure where in many cases players such as your hyperscalers are supporting governments in the operation of their own technologies.

Next slide, please.

And so subsequently, as the value of digital public infrastructure is better understood and is gathering steam across the globe, we likewise see increasing adoption in Africa.

As we saw before, the EGDI and the like, that there is -- this adoption is slow or slower across -- across the continent.

So these principles Dr. Min was talking about, open source and the like, these principles are arguably key mechanisms that will allow service delivery at scale by allowing governments to adopt these technologies, lead with their own interests, and operate them independently and autonomously.

Next slide, please.

If we break away from -- we'll start to unpack the debates within the DPI realm around what is public, we see that there is room to explore the role of the private sector in supporting the design and delivery and operation of services through technology and government.

And so we see in Africa that the private sector has a key role to play in many of these -- in many countries and service delivery around that P and DPI and whether that P needs to be big P or small P for those that are participants in the debate. We know who that financial services players, telecoms, retailers, vendors, and community are all supporting of bolstering government in its delivery of services. So arguably, if we think about sovereignty, this is maybe baked in the definition a little bit, the sovereignty in terms of government's ability to deliver services independently or its requirements to engage the support of the private sector. And we see that in Africa that participation of the private sector may be a requirement and is not detrimental inherently.

And so this is necessity in terms of given limitations.

Next slide, please.

Likewise, when we look towards the emerging DPI ecosystem, we see that there are a wide variety of nongovernment players that are offering technology and support. A couple of them are on screen there. So these entities will help governments identify what technologies to use. They will help them roll their technology out and optimize it for their local environment.

Again, this is contrary to a hard line view of an independent economist state by joining in the participation of these entities and their support. So these role players arguably as nongovernmental affairs are particular to capitalizing digital transformation across Africa. Likewise, equivalent to the requirement of private sector players.

Thank you very much.

>> RODOLFO AVELINO: Thank you, Korstiaan.

Now, we have a policy adviser at the digital impact alliance. His area of working includes research and [?] digital public infrastructure. In his previous role at the Ministry of electronics and ITGOI, he worked on DPI negotiations at G20. Tackling questions of why, what, how of DPI.

Ritul, given your experience in this field, who do you share with -- who do you share your -- with us your thoughts on the connections between digital sovereignty and DPI. .

>> Ritul Gaur: Hi. I wish I was there in person. But there's a gentlemen in the room who are my colleagues, so if I say anything which is controversial, they are my lawyers. So I want to start, and I also have a great job to explain something which has been a lot of time theorizing, which is digital public infrastructure.

To think about digital public infrastructure, think of society in the digital age. Now, what is it that is absolutely required? It is an identity system which is secure, which is -- which can be authenticated against something which can truly prove that you are you. And in a unique way.

So an identity is an important benefit. A fast payment systems which allows you to transact both B to B and person to business, person to person, et cetera. And then data which allows you to both store and share your data across both public and private services in order to access different services.

Now, it's not restricted to this because DPI's still an evolving concept and there are already new DPIs in climate, commerce, et cetera. But proudly think about why are we doing this as infrastructure is because it lays down just the common minimum rails as in the 19th and 20th centuries, roads, rails, dirt, and then you build on top of it for other services.

Now you could ask me a question there is how digitization happens, what makes it new or why are we saying this or calling this as an approach?

A simple answer to that is think of DPI from three common aspects, which is tech, governance, and community. Now, when you think of technology, it is ab evaluation of open source technologies using open standards and open specs to build the deck that's required.

So essentially for your critical national digital infrastructure, you are not going to big window contract, but you're using a lot of open source tools, using open standards. You're not going for big brand, you're using a lot of open technologies.

The governance of DPIs, first is the deck, the second is the governance. The governance is multilayered. There's a governance embedded in the protocol itself, which is safety by design, security by design. And there's a governance of specific aspect of DPI, let's say if it's an ID, there will be an ID regulation or legislation. And of course your broader umbrella data protection GDPI which also applies. There's the tech, then there's the governance and then the most important part, DPI is nothing without its community.

So to borrow a phrase from a professor, DPI allows you to have shared means to many ends, because essentially it's laying out the most common rail. But then allowing others to build a market economy around it, allowing others to build that payment service app to then offer other things. So that's the amalgamation of these things.

And there are two most important things of digital public infrastructure is it has to be open for all to access. And it needs to be interoperable. It needs to be interoperable across different systems and the country, et cetera.

Now the element of sovereignty as to what is the role of DPI and sovereignty is linked. I believe DPI does empower governments and countries to exercise more sovereign control over their critical national digital assets.

And we've seen this in the case of India. India gradually moved away from -- 80% of the financial transactional goes through DPI. They are not going through MasterCard, but national ID data, everything is homegrown, it's out of India, so I think in a lot of sense, both India, Brazil, Singapore, Togo, we've seen that how DPI has been a critical enabler of sovereignty.

But now I think at this stage we need to take a step back and actually analyse what is sovereignty. What is digital sovereignty and what does it mean in this context?

I'm going to break it down in three aspects. One being the data part of it. The other being the hardware part of it. And third thing, the software part of it. The software is the most easiest part. Why? How does DPI enable sovereignties? Essentially you're taking something from GitHub, contextualizing it, making it in a way for your population, contextualize for your population so it essentially becomes your own source code with the moderation, with the moderation that is required.

We've seen this in the case of an ID provider, a government to person service provider, a wallet, a fast payment system, open CI, there's a lot of these software which are out there and open as open digital assets are adopted by the country to then contextualize in their own economy and then in the software is housed and hosted within the premise of that country. It's only by that country.

So software is one aspect of DPI sovereignty. The second is the hardware.

Now, a lot of DPI related stuff requires you to have scanners, cameras, et cetera. I think in this case the sovereignty is a bit malleable because you still require international companies. They are still running through companies that make these standardized hardwares, et cetera, which are required to enroll large swaths of population. And there are solutions that are required based on your contextualized population. In this case we have something called voice pops which pops every time you make a payment.

So a lot of vendor management procurement happens in the hardware part of it, which could be both domestic but international.

And then finally, the data aspect of it. So data is something which happens in both ways in terms of DPI. It also stays with country if you have localization norms. But also a lot of them go for cloud-based data because it's cheaper, you can switch clouds, et cetera.

So I think how to and sort of summarize this and say how can DPI enable sovereignty is, of course, using open technologies. That's the most important thing. But true regulation, use data localization norms, get better deals with vendor, and make sure that for poorer countries particularly that a vendor does not come and harass you.

If you are going for a vendor, make sure there's a high degree of vendor interoperability in your case. If you want to move your data from Google cloud to AWS or to  Oracle, you can do it. Use open source, open standards. Pick domestic standards as much as you can. Don't rely on the big vendors because they have a lot to service and you will be undertier, last tier list to be serviced.

And I think the final and most important thing is the funding. Try to get neutral owners who do not try to push you a certain kind of technologies. Try to find partners who are invested in the longevity of the system and not the constituents back home that wants to sell you a certain type of software which then the servicing of will be super expensive.

I think with that I'll conclude my statement. There's a big link and we've seen it in countries in Brazil, et cetera, where DPI is enabling high degree of sovereignty. But there are multiple phases to that sovereignty that still needs to be figured out and still needs to be tweaked and managed.

Thank you.

>> RODOLFO AVELINO: Thank you, Ritul.

Now we have the journalist in the coordinator Internet Steering Committee. She is also a special adviser to the minister of the science, technology, and innovation.

Renata, can you comment on how the Brazilian plan for artificial intelligence relates to the challenges around digital sovereignty.

>> RENATA MIELLI: Thanks, Rodolfo. Thanks for IGF for this workshop. I think Dr. Min, Catherine, Ritul bring a very broad perspective about sovereignty, infrastructure. Dr. Min brings some concepts. Catherine brings the concepts of connectivity, and Korstiaan brings another aspect, he told us about DPI.

I will answer your question, but as I'm the last one to speak, I'm going to bring some -- some rise specs about how Brazil and Brazilian government are seeing this -- this broader challenging regarding sovereignty in infrastructure.

Well, I started pointing something that's very obviously, but nowadays we need to tell the obvious. That we live in a world where every day all human activity, social, economic, and cultural relations are [?] by communication economists by a broad technology system. The massive of the new technologies reshapes the international political board and redefines groups of countries that are producers and consumers of the digital technologies. And this is our main concern is this.

And we bring this debate to -- during our presidency under the AI priority we led on the digital economy working group this year. Besides AI, we had another -- three other issues, meaning connectivity, DPI, and information integrity.

All starting from this perspective about how we can move and address the challenges of asymmetries we have in terms of technology and emerging new technologies.

Well, these issues relate to the symmetries between countries exist in many areas and have been present for a long time. This scenario has deepened significantly with larger digital platforms which, in a way, [?] society and set now forms of [?]. I feel large companies, the big techs operating in various areas of the economy, but have platforms at the core of their operation that may data authorizations, the flow of information, the provision of services, and in this very moment, all the -- the infrastructure and knowledge about the development and the employment of AI in world.

So regarding AI and other emerging technologies, what we have now, we are facing a scenario that at least to this moment we are facing a deepening divide in equality, particularly in the Global South. A sense, the digital sovereignty is increasing.

In term refers, as Dr. Min said, among other things, to a national technology autonomy or tools using its own infrastructure, datasets, workforce, and businesses.

Furthermore, it involves the ability to regulate and decide on its own digital AI path in a quest to ensure inclusive growth and sustainable development.

The digital sovereignty refers to the ability of the states to control their own infrastructure, emphasizing the position of each country in controlling ICTs that is a greater or lesser degree of autonomy to make choices and decisions in the field of technology.

And of course, in the field of cooperation within other countries, this is very important also.

So for us in Brazil, we have some key perspectives. The role of the states. So emphasizing the importance of government action through public policies that support the development of technological infrastructures, science, and technology initiatives and policies to foster innovation and reduce dependency. This includes encouraging and promoting the use of national technologies and regulating the use of technological tools.

I'm going to say nice precisely the AI plan. But to have another public policies regarding industrial economic development and other initiatives that compose a very large umbrella of public policies regarding investments and in technology.

The second point, sovereign digital infrastructures, developing infrastructure, digital infrastructures. We also face profound challenges in the meaningful connectivity and access. This is how to reduce the prices of equipment, for example, cell phones and computers for the population.

We are not -- we cannot see only the access aspect. We need to see more broader aspect when we're talking about meaningful connectivity. After all, if we are talking about leaving no one behind, we need to develop the capabilities locally to offer better services to society.

For this, there must be meaningful, connective, and we need to strategically think about how to be [?] digital training process for the entire population from young people to the elderly, for people living both urban and rural areas, in addition to connectivity, we need to think about the equipment, especially cell phones which is the most used device to access services that have the quality and minimal capacity to run applications and tools that use AI.

Governance and regulation and now in Brazil we're discussing a bill regarding regulation on AI. And for us, it's crucial to rate frameworks for that governance and platform regulations also. That ensure accountability, transparency, and [?] of technology.

In this scenario, where the digital system is available in the country, platforms in the AI tools are mostly international, it's necessary to discuss regulatory mechanisms that establish rules for the operation of these companies in the country with transparency obligations about their systems, learn information about specs that they have economic, socio and political impacts, conduct adjustments, among many other regulatory aspects related to social and human rights.

Security and privacy and also develop sovereign and security and privacy technologist.

For AI to aim at the [?] and other social issues, its developments must be guided from this proposal from the outset. This includes the use of training dataset of structures to enable AI to achieve its goals with a [?] and geographical diversity. Otherwise, AI could become yet another driver of inequality.

This is why data sovereign is central to the development, implementation, and use of AI by countries that aspire to any degree of self-determination. Those was the main focus that Brazil brought to the digital economy working group this year. We produce as Brazilian president a contribution a toolkit AI assessment in partnership with UNESCO that was our knowledge partner. We deliver the potential of enabling our approach to the responsible development, deployment, and use of AI technologies.

And also a mapping AI induction for public service, we think sites into systematic operation and supporting applications within and by governments. Consider that DPI is, as Ritul said, a very strong and important tool to inclusive and sovereignty digital for countries.

In terms of public policies, the perspective I bring to this debate are reflecting our Brazilian AI plan as Rodolfo said, led by minister of science and technology of innovation this year.

A plan that forecasts an investment of 23 billion reals, around 4 billion dollars, for Brazil is a huge amount of money.

In terms of infrastructure and sovereignty, I highlight some investments from the Brazilian artificial intelligence plan, such as national infrastructure program for AI, around $105 million, sustainability in the renewable energy program from AI, around $83 million. Data and software ecosystem infrastructure program for AI, $165 million. Research and development program in AI, $870 million -- no, $144 million.

And the perspective to achieve in the AI super computer, that's put Brazil's on top of five super computers in the world. This is just some highlights on our AI plan that has five axes, regarding governance, private sector, investments, reskilling, and capacity from workforce and also infrastructure investments.

That's it for now. Thank you very much.

>> RODOLFO AVELINO: Thank you. Thanks a lot for the very relevant points.

Now we are going to open the floor to questions of the audience on-site and online.

First we can start on-site.

>> Hello, good afternoon. So very happy to hear that the research I've been conducting with professor Min Jiang has been presented here. I was happy to see that the points we raised in research are now well integrated.

But maybe some of them are not so well integrated.

Let me give you a good example, because we have been doing a lot of research on AI sovereignty over the past couple of years, and of course connectivity is one of the points that we stress is essential to achieve AI sovereignty.

Let me give you a complete example that also speaks to the debate on DPI that was brought here. Most of the south countries, including Brazil, do not have meaningful connectivity. We have most the population connected to most selection of apps, including mainly the Meta family of apps. To give you concrete details that friends from city here can [?] good study that they've done on many connectivities here, 78% of people in Brazil do not have meaningful connectivity.

What does it mean concretely? I think the Brazilian government is putting a lot of money, we are analysing this primarily in software and data with the AI plan, but even if we have the best possible language models trained with Brazilian data, if all Brazilians only access Meta AI through Whatsapp, zero rated where no one else will be able to access the new fantastic domestic models created thanks to the plan, that is not a very best way of directing the public investment.

And this is due to the fact that access is an incredibly relevant variable in this context. As you were saying, we have been doing the research, the fact that we access AI and will never access Brazilian technology because they will keep on having -- not having money to pay for full Internet connectivity and only being directed to Meta Facebook, Whatsapp, that is an enormous improvement from national innovation. It frustrates a lot of the logic, very good logic of putting public money to improving national research and development, but at the end of the day, the consumers will not use it and keep on not only using another non-Brazilian technology, but also train it.

For free of course. So I think the entire logic here is a little bit frustrated the let me give you a good example of an institution in Brazil who understood this logic very well. The Brazilian central bank when they introduced peaks, our Brazilian digital public infrastructure for payment, Whatsapp, one to two payments, but they blocked it and they suspended it and the rationale was precisely it had been launched before, everyone in Brazil would have used it only for payment, and we would not be here praising peaks as an example of a success story if the Brazilian central bank hadn't blocked Whatsapp payment and enforced it until the entry of PICs. So I think these are points that if not considered, I know that very well that the Brazilian AI plan does not consider connectivity, but I think it's a mistake and I think that this, as you were saying, it's an essential point and should be brought into the picture. Otherwise we risk putting a lot of public money for nothing.

Thank you very much.

>> RODOLFO AVELINO: Thank you for the question, Luca. Now let's go to online -- okay, okay.

>> Okay, thank you very much. I am a researcher at the University of sustainable lab and cofounder a nonprofit in Brazil. Amazing insights. I was wanting to ask the presenter and speaker from Georgia, I'm sorry I didn't get your name, I really apologize for that.

>> CATHERINE: No worries, it's Catherine.

>> I was wondering if you could talk more about the data center initiatives in Georgia and also if you could also share how are you thinking of embedding this within an energy infrastructure, water infrastructure as this has been a very hot topic in the last few months I would say. If you could share some thoughts about that and the Brazilian government is thinking about these also. So it would be interesting to hear.

>> CATHERINE: Actually you pointed out in the question the [?]that I missed and wanted to share about.

So the data center topic, what we have now under discussion is, like, first of all, as a regulator, and as a state representative, we are working a lot about enabling access to the existing Internet infrastructure, opening markets, building them IXP and neutral exchange point. This is the first step we are seeing to be. It's an ongoing process and somehow already almost done and this is the first step to enabling the real data center.

Another topic is resilience of infrastructure and finding out that we're quite small country but still find the geography work, the data center is best to be located from the also the energy point of view.

The good side of the story is that we have -- Georgia is the green energy producing country. So we can ensure that the energy produced locally is -- will be the green energy, which is very important how to say component of bringing the right data center and bringing the investors to be interested in this kind of projects.

And also related to this one thing is producing green energy. Another point is having the geographic location where the energy efficiency is -- will be the best with its -- actually with support of actually Amazon did some kind of research that in Georgia this energy efficient locations are present.

And this is -- so this is kind of projected level. But there is a lot of stuff still ongoing to be done. Security aspect, fiscal security aspects are very important, but still needs to be resolved. And another part is also energy prices. We're quite competitive. Electricity prices. So this is -- those are the different, how to say, different components of the project we need to solve and put together. And most importantly financing model which will work best for this to make not only the local, like, Georgia-specific project, but regional project.

So investment options are there on the table, whether their state should be part of it or it should be totally public or it should be public-private partnership, et cetera. Those are kind of points to be resolved still.

Thank you so much.

>> RODOLFO AVELINO: Thank you. Let's do a round of two more questions. And the speakers will answer.

>> Yes, just make the online questions and I think I have one more from the audience here and then we make a round of the speakers answering, okay.

From on the Zoom questions, we have someone asking, he likes to learn about this cable that Meta is investing. So maybe again to Eka about her platforms investing in cable.

Another question, he asked, digital public infrastructure if translated to other language can be translated as a straight infrastructure. And this would be controlled or owned by them. Is interest a widely understanding of what AI is.

And finally to Dr. Min, how does [?] type influence collaboration, are democratic regimes more likely to cooperate than utilitarian ones or it's an outdated assessment? I think maybe we can do another -- one more here, okay. I think this is the -- oh. I think it's the last one because of time.

>> It was the same, so I had two questions there. So we have discussed -- it was my question that he answered. Yeah.

One, do you hear me? Okay.

So here are two questions. So we have discussed sovereignties for infrastructure development but they all are almost all for regional connectivity. So the question is, that how can the digital sovereignty on digital infrastructure that has original impact be used as a weapon against other countries? And if yes, how it can be eliminated.

And a small comment that it may be -- many regional projects requires cell [?] to get engaged. How can we secure it taking into account the challenges we have discussed today about digital sovereignty and digital public infrastructures and also (away from mic).

>> RODOLFO AVELINO: Thank you for the question. Now answering the question, please, give him your closing lines and the final comments.

Can we start by the same order. With Dr. Min.

>> MIN JIANG: Sure. Thank you for the great question. And I think it's a tricky question. I'll make two points in relation to your question. First of all, traditions -- traditional notions of definitions of sovereignty are usually predicated on nation states having a form of autonomy or self-determination. But not -- but do not take into account in reality the very notion of power, right?

So small nations and small states know this very well, especially in the digital age, big tech have power and, you know, financial power and that can easily eclipse those of small nation states.

In fact, if one examines the telegeography map that our previous speaker referred to early on about global undersea cables, companies like Amazon, Google, Facebook, all have their own dedicated infrastructure at that level, right?

So small countries not only in the Global South, but also, for example, EU, they recognise that in order to be sovereign, they must also cooperate and build alliances, okay.

So this is an important to recognise. Also in a previous speaker Ritul's account will public digital infrastructure, you know, he makes the case that nation states did digital development, especially those in the Global South, have to draw upon open source and free software, which are very, very important notions to common digital sovereignty.

So I think we need disrupt how we think about, you know, sovereignty to begin with.

And second, the question that is about regime types, it's very, very important notion for sure. But we also need to recognise, you know, the regime types are labels that we attach to nations, but nations also change and evolve. The political system as we had seen in the United States, my country, has evolved a lot. We just elected Donald Trump for a second term, right? How do we label countries and what type of regime they are is becoming more and more challenging. In the United States is a country with great power and with great power comes great responsibility, right? And what NSA, for example, implemented you know, for a long time and what the big tech are doing, perhaps, challenge this very notion of what it means to be democratic.

And I think we're at an age where the older sort of conceptualization and infrastructure and legal regimes to think about democratic is somewhat breaking down. And that's why we're seeing this resurgence of claim to digital sovereignty and the different actors, national or international, are hoping to gain more independence, autonomy, and self-determination.

So yes, I'm happy to carry on the conversation through some other means, but I'll restrict my comments for now. Thank you.

>> RODOLFO AVELINO: Thank you very much. Catherine.

>> CATHERINE: Yeah, very challenging questions, let me put it this way. And exactly the -- echoing what are the underlying challenges with sovereignty and starting from infrastructure level up to the service and data protection level.

What I wanted to outline that, yes, on the one hand side the sovereignty can be used as some kind of weapon, some kind of strength from the one country having the totally sobering kind of infrastructure and giving the access and kind of isolating from one side, conceptually isolating the country.

On the other hand side, it requires a lot of effort when we speak about the regional perspective putting the regional concept of sovereignty. So countries with very different political views should -- should stay together and agree on the major terms. But I think that this debate of digital sovereignty, why it is open debate and why it is an evolving debate that countries still are trying to understand what aren't this basic and minimal concept of, on the one hand side, independence of data and infrastructure, and at the same time the protection of data without this kind of touch points among countries -- between countries with very different political or views or geopolitical locations.

It's impossible to let this very interconnected world work. We will need more and more interconnected data centers, otherwise, it will not work at all. But at the same time, countries and regions are required to protect themselves by owning some kind of the infrastructure.

So I think that this is a thin line where we need to all agree and we need to introduce some kind of frameworks. For Georgia, what I can answer is that the EU framework we decided to go with, existing EU framework of this sovereignty concept on data-level GDPR that is provided by framework for data protection is one that is acceptable for us. And we think this is the best model we can introduce.

And it should work for our region as well. For the current situation. This is my answer.

>> RODOLFO AVELINO: Thank you. Korstiaan, please.

>> KORSTIAAN WAPENAAR: I'll make my conclusion very short, maybe firstly just to say thank you to everyone, to the organizers for the participation. To my fellow panelist for the interesting discussion.

Without stepping ahead or two of the questions to yourself, maybe a couple of thought provokers. On the regional considerations for sovereignty, maybe just to propose the question around how one manages the aspirations of the EU to develop a continental identity system.

And how that would be governed and managed and the extent to which that is a risk of or how to prevent exclusion across different markets.

And then curious to hear from my fellow participants following me their view on multiple definitions of DPI and what that big and small P looks like as we think about our colleague. Thank you very much, everyone.

>> Ritul Gaur: Thanks, Korstiaan. I think the first question, how do we trust DPI, I'm not going to do each regional answer for that. But I can say just in terms of geography context, we need to ensure you have the highest [?] data centers, security assessments, you have regular audits, and you do similar things in cross-border context if you have an ID payment which is in regional context. We don't have it in this case, but as we build, I think these are the three metrics that we follow, but then there should be some nontech government side of it which will also follow.

Now answering the perplexing puzzle of the DPI which is what is the P, should DPIs be controlled or owned by state? There's no clear definition of DPI. It's an evolving stage. The G20 definition is as confusing as clarifying as it is. And I take some blame for it. But if you think about it, it has to be understood from a different grade perspective.

Something like an identity is a very sovereign function. To say you are you is -- is -- can be trusted by -- by a sovereign state more than any other entities. So in this case, ID sits out of the Ministry of electronics it's an organization backed by our constitution law and it's backed with civil servants, very state-driven function.

On the contrary, payment system is rather fluid. It's a nonprofit structure, it's denomination of different banks and the central bank coming together and just building the protocol. The rest of it is actually controlled by different banks who come and participate on top of it. But the role of the state in that case is the regulator.

The state is only the regulator in these payments scenario. And similarly, the document -- the data sharing wallet as well, the state, again, Section 8 company has created in which the state only regulates in terms of how you can share your credentials, et cetera.

I think it will defer on a country to country basis. I remember sometime back I was in Ghana and talking to the bureaucrat there. He said in our country everything is [?] so it be a country to country phenomenon. But in my limited experience, most ID systems, I think there's more, we met last week, most ID systems are part of either the home industry or the I.T. ministries, et cetera. So you will see a lot of identity function, which is center to any targeted beneficiary delivery, it is essentially establishing your relationship with the state is done by the state.

But other DPI functions can be performed by different partners. In fact, same pass pay now, prompt pay, et cetera, these are some payment systems and other systems which are cleared by the private sector in conglomeration with the state. But the state's role is at least in this case to be a regulator, to be an observer that nobody creates disproportionate amount of monopolies, that nobody's not playing by the rules. To set the broad rules of the game and let them come and build. If it it's something that requires a great degree of trust, this is the best entity do it. If it's something which can be created by a different marketplace coming together, state can be an observer or regulator.

Finally on sovereignty, I think there's an important link to be made there. My only concern is that as most countries grow in the quest to build their DPIs, we should not lose sighted of cross-border interoperability. The interoperability of those DPIs. As we all go towards moving our own payment systems, making our ID systems, et cetera, we also need to be cognizant enough that we are also thinking of regional blocks, we're also thinking of cross-border pay, et cetera.

So to not lose that. Otherwise, I think in a broader scheme of things, DPI is -- is a big-time [?] of sovereignty. Thank you.

>> RODOLFO AVELINO: Thank you very much, Ritul.

Renata, your final answer.

>> RENATA MIELLI: Yes, we need see it as complementary as with cooperation. Such each country face different realities in these areas in digital areas, cooperation will be fundamental. Without cooperation, we are not going to achieve sovereignty.

Establishing mechanism for original cooperation that create complementary may be a faster path in achieving greater autonomy for our nations.

I think we have to keep this in mind. Regarding the question that Luca made about connectivity in Brazil, he knows I'm a profoundly and deeply critical of regulating. But it's important to say in Brazil we have a huge public policy in terms of strategic investments of government that calls back -- how I can say, back growth, facilitation program and the connectivity policies are in the pact. And we've 28 billion reals, something around 5 -- $5 billion to invest in building connectivity and technologies, 5G, 4G, building backbones, school connectivity, health -- health system connectivity. So there is a public policy that are being made inside the communication ministry, Ministry of Communications. And we -- as I see and as the government see as my minister science and technology innovations see, we don't -- we cannot wait to solve the problems regarding connectivity. And I'm completely agree with you, Brazil doesn't have meaningful connectivity for all population.

But we need to start to build expertise and investments in infrastructure and in all the economic chain of the AI, because we need to start from some point.

So these are two policies that need to be put in movement with each other.

Communications are dealing with connectivity, doing the investments, and we, as minister of science and innovation and other governments and other ministers are focused on thousand built capability in terms of reskilling, in terms of infrastructure, and building applications, AI applications.

So that's my point. We need to do the both things together. So if you want to achieve some autonomy sovereignty in Brazil regarding technology, digital technology, and AI.

So thank you very much for the opportunity and that's it.

>> RODOLFO AVELINO: Thank you to our speakers for their great contributions and to everyone in the audience. This has been a very good workshop.

We appreciate the IGF organizations for hosting this valuable discussion. Thank you, all.