I met Kenzo Soares through an online call, as he is currently a Resident Fellow at the Information Society Project (ISP) at Yale Law School. Kenzo is a carioca—a son of Rio de Janeiro—who has worked extensively on platform workers’ rights. He is also deeply knowledgeable about the political economy of Latin America from a Marxist perspective. During the call, which I took from the Fundação Getulio Vargas in São Paulo, we ended up passionately discussing platformization in Latin America, shifting from the intimate to the technological and the geopolitical.
In our conversation, digital sovereignty takes on a heavy, industrial form. We discussed Petrobras, the state-owned giant operating the most powerful supercomputers in the Global South, yet whose strategic oil data remains hosted on North American servers. This reality informs Kenzo’s concept of “survival infrastructures”: digital public infrastructures like Pix are not mere conveniences, but necessities for banking the poor and providing “white label” alternatives to the predatory gig economy.
Yet, a darker thread emerges regarding internal colonialism and the “black box” of the state. Kenzo acknowledges that Brazil is not just a victim of the North, but a regional hegemony that surveils its neighbors. He notes the deep-seated distrust of the demos, for whom the state is often a predator; here, the push for sovereignty is frequently met with the fear that public platforms are simply more efficient tools for taxation and control. How often happens with academics – to the degree that it is now a meme – we decided to write a paper together. 
***
What is your background and why are you interested in digital sovereignty? Was there a moment or a project that shifted your view?
You told me that you are interviewing activists, scholars and public officers. That is interesting, because I think at some point in my life I have been all three. For many years I was part of the Socialism and Freedom Party in Brazil as an activist. Then I spent ten years as a parliamentary adviser to a congressman, Marcel Freixo, first in the municipal council of Rio de Janeiro, then in the state congress, and finally in the national congress. This year I was in the Ministry of Science and Technology in Brazil for six months before I came here to Yale. At the same time, when I was still working in congress and later at the ministry, I was also doing my academic career: pursuing my master’s, then as a PhD candidate and finally as a lecturer at the school of communication of the Federal University of Rio de Janeiro.
I started by studying the political economy of platforms and digital ecosystems in general. But as I was in these policymaking environments, I was always concerned with the question: how can we go beyond the critique of capitalism? It is really important to understand these new ways in which capitalism is organizing new relations of production, all the debates around surveillance, and how we update historical and structural relations of power. But how can we go beyond criticism and start to build a program, to update the left’s program for society? I was at congress all the time trying to give answers to civil society. My work was to do this dialogue between civil society, NGOs, social movements and policymaking.
One thing that I think is really interesting in Brazil is that, outside the European Union, the Brazilian state is right now one of the main states that is struggling with big tech, especially big tech from the global North. If you think about disinformation and hate speech, and how to regulate that, Brazil is a good example. We even banned Twitter/X for some days to force Elon Musk to enforce Brazilian law about hate speech. I think it is natural for me to start thinking about digital sovereignty, considering where I am from. People always say that countries like the United States do not have to think much about digital sovereignty, because they have the platforms. For them it is natural that they exercise power all around the world. But for us, from a dependent country in the majority world, in a public university where we do our research, I was always thinking about how our meetings and our systems depend on big tech from the global North.
We are talking about Alphabet/Google, or Microsoft. Even in the Brazilian Ministry of Science and Technology, all the meetings and internal communication were done through Microsoft Teams. Now we have some public alternatives for calls like that, but still, the systems that we use every day are Microsoft products. We are talking about the ministry that supports the development of key technologies for Brazil. We are talking about the ministry that deals with nuclear energy, satellites, with projects of public cloud, and also very sensitive information for Brazilian sovereignty in general. All this goes through meetings, emails and documents that are hosted by Microsoft services.
I also interviewed some tech workers for my thesis. They were data scientists and software engineers, and some of them work for Brazilian state owned companies like Petrobras, the most important and biggest company in Brazil. They said to me: we have a specific API to access OpenAI’s ChatGPT. But at the same time, we use Microsoft Teams to exchange information between different departments. We are talking about a goldmine of strategic data, because we are talking about oil. That was a concern that I was able to map through my fieldwork. So I think that is my background and why digital sovereignty became such an interesting and obvious topic for research.
What do we talk about when we talk about digital sovereignty? How has the discourse evolved over the last years in South America and Brazil? Does Brazil have a privileged position, and why do you think that is the case?
I have some disagreements with other scholars, maybe even with people you are going to interview. I really think Brazil is not a complete digital colony in the same way as other countries in the global South. Of course we are dependent on the main layers of internet infrastructure: the domain name system, which the United States still controls through the State Department; and all the infrastructures from AWS, Amazon Cloud and so on. We know this. But at the same time, unlike almost every country in Africa, and unlike most Latin American countries, we are in better shape in some respects. In some specific fields we are even more developed than India, for example when we are talking about computational power, about tech workers, knowledge, and scholarly production.
We have scientists in Brazil who are doing research that is part of the state of the art of AI research in the whole world, in scientific research and in the volume of papers published. Petrobras has the biggest supercomputer in the whole global South that is entirely dedicated to AI development. It is bigger than what some European countries have. In that sense Brazil is in better conditions than Spain or Portugal. They are not main superpowers, but they are part of the global North. So I think, especially our tech workers and our tech community have the knowledge and the conditions to develop state of the art systems.
What Brazil does not have is capital. Even if we think that in Latin America almost all investment in information technology in general, and also in AI, is concentrated in Brazil, it is nothing compared to the United States and China. We do not have capital, especially venture capital. We have some Japanese funds, like SoftBank’s Vision Fund, investing in Brazil, but it is not comparable to China or the United States. Second, unfortunately, our national state does not have the kind of policies that other developing countries like India, China and even Russia have. We do not have data localization enforcement in Brazil. We do not have many public platforms. We do have a payment system, Pix, which was an advancement. I think Brazil is still dependent on global North infrastructures and most platforms.
I think Brazil also has some asymmetric relations in the digital field with its neighbors and with other countries from the global South. Most of the big platforms in Latin America are headquartered in Brazil. The big exceptions are Mercado Libre, the Argentinian marketplace, and Rappi, the Colombian delivery app. With these two exceptions, almost 70 per cent of the main platforms that are unicorns – valued at more than one billion dollars – are headquartered in Brazil, and these platforms are expanding to other Latin American countries. I always need to emphasize that Brazil is not competing like China to overthrow the United States. We do have specific power relations in our immediate field of influence, in our traditional and historical area of influence in Latin America.
How do you understand digital sovereignty as a concept? From which side do you approach it?
If you read the literature, digital sovereignty in general is still more a political claim than a concept that you can easily evaluate and measure. It emerges first in public discourse before it becomes a scholarly or theoretical concept.
Brazil certainly does not have the mechanisms to enforce its sovereignty in digital ecosystems that other developing countries have. It has less sovereignty than Russia and China for sure, and probably also less than India. But I think it has more potential to quickly develop these conditions. This is mainly a political issue in Brazil: whether the national state will take steps in terms of public policies, regulation, legal frameworks and public funding, because we already have the socio-technical conditions to quickly develop these solutions.
This is different from a country that does not have any local tech community, does not have public universities, does not have public IT companies. At the same time, Brazil does not fully have digital sovereignty when we consider its relations of dependence with the global market and the global North. But there is another question: is Brazil really respecting and helping the digital sovereignty of its neighbors? We just had a scandal of Brazilian surveillance over Paraguay, where Brazilian intelligence agents, equivalent to the NSA, were spying on Paraguayan officers through their digital devices. This was connected to the bilateral agreement that we have on energy generation. We have a binational hydroelectric plant on the frontier between Brazil and Paraguay, and we are renegotiating tariffs.
It is really interesting because Brazil was spying on Paraguay to have more information and a better position in tariff negotiations, at the same time that we are fighting against US tariffs that affect Brazil. That is why I say Brazil is not just a victim of the global North. It also exerts power, and it does not always help the sovereignty of its neighbors. It is not all roses and flowers between Latin American countries.
So do you see more unity or more fragmentation in Latin America regarding digital sovereignty?
I think Latin American political integration as a state agenda was stronger during Lula’s previous administrations. We had initiatives to create a continental parliament, inspired by the European Parliament. We had an expansion of the economic bloc Mercosur. In the current administration, Brazil is more enthusiastic about BRICS initiatives. I think Brazil should indeed develop connections with BRICS, but it should bring Latin America as a bloc to the negotiations. Otherwise, BRICS risks becoming just China’s area of influence. Then it is not real multipolarity; it is only a shift in which power is hegemonic, the United States or China. BRICS is basically a state process of integration through bureaucracies. We have no real public space for movements. This is different from twenty years ago, when integration between Brazil, Bolivia, Venezuela and Uruguay under leftist governments like Chávez and Lula was a process of integration through the state but also through social movements from below.
There is another important thing: we need to think about digital sovereignty beyond the state. Our discussions of sovereignty in general are deeply connected to states, since the Westphalian paradigm became the normality of international relations. But one thing that I emphasize in my research is the idea of popular or grassroots digital sovereignty. I studied some movements in Brazil that bring this debate. I was recently at the launch of a book that argues that if states are not promoting integration, then social movements and workers should articulate Latin American solidarity relations and some kind of workers’ organization on a continental scale. That is one thing.
For sure, Brazil has a responsibility, as the biggest economy on the continent, to support its neighbors with technology transfer, infrastructure sharing, a pool of resources. And not only with Latin America, but also with Africa. A real historical challenge that maybe can explain why Brazil does not have a stronger connection with Latin America is that, in the past, Brazil was more connected to Africa than to its Latin American neighbors. Unfortunately, that was largely because of the slave trade, but not only because of that. Brazil was more of a transatlantic community. I am from Rio de Janeiro. The elites and intellectuals from Rio de Janeiro were more connected to Luanda, to Angola, to Mozambique, than they were to Asunción or La Paz. Buenos Aires was an exception because of the River Plate, but not so much to other Latin American capitals. So I think Brazil can also help to develop digital solidarity relations with at least the Portuguese speaking countries in Africa.
I was thinking that BRICS is mainly a geopolitical and economic attempt to bring states together, but it is lacking the element of the demos, that shared cultural component which is so important for democracy. In Europe we keep referring to an idealized European essence. It is a myth, but it kinda works. But I wanted to ask about Pix, the best example of digital public infrastructure from Brazil.  I think it is a scandal that we do not have that in Europe.
Pix is a really interesting example. I think it is an example of digital public infrastructure, maybe not a completely comprehensive one yet, but it is powerful. And it is going to be extended to other Latin American countries, which is also important. It is also an interesting example of why the European Union probably does not have a similar system. Technically Pix is one hundred per cent Brazilian technology: Brazilian technicians and tech workers developed the system. But it was inspired by the Indian model, and the Indian model was inspired by Russia. Russia needed such systems because of sanctions. When the credit card duopoly, Mastercard and Visa, stopped operating in Russia because of sanctions, Russia needed to develop its own payment system. Since the 1990s, Russia and China had been pushing a cyber sovereignty agenda, but it was the sanctions that really made them develop their own payment systems. India was inspired by that, and then Brazil learned from India. We actually sent technical people on mission to India before developing Pix.
I think this is a good example that BRICS is not just speeches, but in this case there was real cooperation. If not a full technological transfer, it was at least shared inspiration from the global South. I would say that the European Union does not need or does not have such a system because it has never been cut off from the international payment system. There was no similar pressure. I think Pix is revolutionary. One example that we always give: Brazil is a very poor and unequal country. The first time ever in Brazilian history when most of the population was part of the financial or banking system was after Pix. The pandemic also really helped, and so did all the Brazilian programs of cash transfers from the state. But Pix was the moment when most of our population started using banking systems, and not only online banking but banking in general.
I know that because I interviewed a tech worker who worked in a Brazilian startup whose business model, before Pix was launched, was to try to make the poorest part of the Brazilian population use banks. It was a fintech. They failed. The startup closed because Pix solved the problem. So it was a public solution. That is an example of national sovereignty, and it can also be the basis for mechanisms of international solidarity. Pix also shows the contradictions when we talk about sovereignty and we do not distinguish between the state and the people. One of the biggest political defeats of this Lula administration was when the government tried to expand the powers of the tax authority. The federal government published a normative instruction that lowered the limit over which Pix transactions started to be monitored for fiscal reasons, especially income tax. There was a huge popular revolt. The far right tried to capture this popular uprising against the government.
The government’s argument was: why are you fighting against having more data about your financial transactions if the idea is to fight organized crime and black markets? But the thing is that, for people, they love Pix, but it is a black box. They do not know what the state does with their data. For very poor workers, the idea is that if the state wants to have more data about them, it is because it is going to raise their taxes. So I think this is an example of how Pix is controversial. People thought that more data in the hands of the state was not necessarily a protective move from their point of view. They consider the state an enemy. And that is the contradiction of every capitalist state.
Sometimes legal scholars distinguish between internal and external sovereignty. That is also related to the United States.  Digital platforms act as the long arm of US foreign policy abroad, but internally they are also forms of government that can oppose the state. People sometimes prefer to be subjected to platforms, like Apple’s encrypted data, than to the state. Historically, the state is at the same time a driver of interstate competition that produces inequalities and, in some places, a welfare state that tries to reduce the very inequalities this system produces. It is very contradictory. This is also the legacy of Cybersyn: digital public infrastructure needs good government, otherwise it turns into a totalitarian nightmare. Is relation between platforms and the state is always ambivalent?
Yes, because we are fighting a neoliberal agenda that says: open markets, privatize all your state-owned companies. We keep saying: we need public clouds, in the sense of state-owned infrastructures; we need public servers; we need public data centers; we need public platforms. Now I am researching Brazilian state policies for platform cooperativism. Brazil is developing a white-label platform for platform cooperatives. A public university, Santa Catarina Federal University, is developing this platform with funding from the federal government, and the idea is that each cooperative can use the platform for free.
But the question is: who is going to access this data? Will society in general have access? Will it be an open platform? Or will only the state have access to this data? Next year we have elections, and the far right can become the administration. We are talking about putting all the data from these workers and workers’ organizations in the hands of the state. That is the contradiction.
Here in the United States we have a similar discussion. I attended a presentation about how the First Amendment makes it difficult for the US state to regulate hate speech, and how alternative models could be developed, for example by enforcing European law. The conclusion was that we should look to European style regulation. But the thing is: when we are talking about Nazis, of course we want to control hate speech. However, when there is a Trump administration, do we really want to abolish, or seriously restrict, the First Amendment? Who is going to define what hate speech is?
Let us return to platform labour. One question about the imaginaries of platform workers that you worked on: what did you find? Are they different from what we see in other continents? Have you seen some Latin American specificities of the organization of work?
You know dependency theory. I had a paper that tries to connect dependency theory to platform labour. The thesis is that Latin America has a structural overexploitation of labour. You have structural conditions that make labour cheap and vulnerable, and that creates natural conditions for platform labour to grow.  The data shows that Brazil, for instance, has some of the worst conditions for platform labour in the whole world. Every year the Fairwork reports show Brazil at the bottom, competing with some countries in Asia.
You can actually see it in the street. It is not only the prices of deliveries, which are absurdly low, but also working time and intensity. I have this joke. I had an Italian girlfriend and she had just arrived in Brazil. It was kind of late at night, around one or two in the morning, and she said: “I am hungry, but I do not want pizza.” I said: “Let us order something.” She replied: “But it is two in the morning. I do not want pizza.” And I said: “We can have anything.” She asked me what I meant. For me it was so natural that I could order anything at 2 a.m. in Rio de Janeiro. When I was in Europe, I realized that you cannot. Why? Because even the immigrants in Europe do not work twenty four hours a day for delivery platforms to the same extent and intensity that people work in Brazil.
I was also thinking about the permeation of Uber in Latin America. In most places in Europe we do not rely on Uber in the same way. We use public transport or our own car, while here it is all about using Uber, because Uber becomes part of the infrastructure of safety. So there are already existing structural conditions of inequality. Then the platforms come and “save” people in some way, and they benefit from these inequalities that capitalism has produced. But moving to the theoretical part, What is your position on techno-feudalism and on the idea that we moved beyond neoliberalism?
I come from media studies, where the way capital permanently accelerates social life also affects academic production. Twenty years ago we were talking about the “wealth of networks”: “network” was the keyword. Then it became “surveillance”. Then it became “platforms”. Now it is “AI” and “techno-feudalism”. So what is the difference between platform capitalism, surveillance capitalism and techno-feudalism? These are interesting concepts that focus on and illuminate specific dimensions of our contemporary society. But I think sometimes people feel the need to launch huge new paradigms because, as intellectual workers in the attention economy, they need to make noise. There is a political economy of critical studies. We are in the business of producing ideas.
I do not think the basic conditions that we call capitalist have been erased. Most of the social wealth is still produced by the process of people working in exchange for salaries. You know the argument that we live in a postindustrial society? In fact, we have more industrial workers on the planet now than at any other moment in human history. The problem is that most factories are now in Southeast Asia. We are not talking only about robots. We are talking about people who work in conditions very similar to those in Manchester when Engels was writing about the English working class. We still have people working in mines for rare earth minerals in Africa. We cannot make iPhones, or Nvidia GPUs, without kids working in almost slave-like conditions in Africa.
My position is that we are not just in a rent economy. You can say that platforms specifically work largely through rent: that is the way capital associated with platforms takes its share of global wealth. But this is not the majority of the global economy. People confuse the financial valuation of companies with the mass of value actually produced. We criticize this by calling it fictitious capital. People say: these are the biggest companies in the world. Yes, in market valuation. But if you think about the mass of value, it is still produced by labour.
Even within platforms, if we talk about platform labour: iFood alone has almost 200,000 workers in more than 1,100 cities. That is just one Brazilian labour platform. How many workers does Uber have globally? And these are classical capital–labour relationships. Of course you have this new dimension of data extraction, but the economic exploitation is the same. So I think techno-feudalism may end up being the same thing that “cognitive capitalism” was ten or fifteen years ago: an attempt to describe changes inside capitalism that sometimes overstates the rupture. What these concepts are trying to capture, and where they have a point, is the growing power of private companies compared to national states, especially in the United States. This is the story where Elon Musk or Mark Zuckerberg appear to have more power than Trump in some domains. But the military-industrial complex always had this kind of power, so I am a bit skeptical.
Of course we have new regimes of production. That is not the same as a new society. For most of capitalism’s history, most people were engaged in farming. The moment when most of the global population started living in cities is right now. This happened in the last ten years. Until fifteen years ago, most humans still lived in the countryside. This transition is largely because of China. But even this huge transformation did not mean that capitalism as a system changed into something else.
With the rise of ubiquitous computation, I think one difference is the combination of the platform way of organizing people, money and services in space and time: capitalism always relied on monopolies, but platforms now have new ways of maintaining monopolies.
Of course. But when was capitalism not based on monopolies? The idea of pure competition is a fiction. All the infrastructure of communication has basically been monopolistic. In Europe, historically, how many television channels did you have? In Italy, RAI; in the United Kingdom, the BBC; in Brazil, Globo as a de facto national monopoly; in the United States, three big networks. Now we have Meta and Alphabet as a duopoly of attention on a global scale. This is a case where quantitative change becomes qualitative, for sure. It has huge impacts. My father is eighty years old. He was born in the 1940s in the countryside of Brazil, and now he watches South Korean soap operas. Of course this is different.
At the same time, I do not know if this configuration is going to be permanent, because the Chinese model of the internet, or “internet with Chinese characteristics”, is expanding. Many countries in Africa are buying digital sovereignty as a service from China instead of from the United States. The servers are from Huawei. What backdoors the Chinese state may or may not have into servers in Nigeria, I have no idea. People are afraid of the NSA, and now they give control to Huawei. We also have to ask how many countries in Africa do not access Netflix but do access WeChat or Chinese platforms. I do not know the exact numbers, but they are expanding a lot.
Let us end with this last question. What steps could Latin American countries, or Brazil in particular, take to enhance digital sovereignty in the next five years?
I think Brazil, and maybe other Latin American countries, have some advantages compared to the United States. I mentioned that we are developing and funding, through the federal government, public platforms for workers. I am from Rio de Janeiro, a city that has a state owned platform for taxi drivers, with more than 15,000 taxi drivers and more than one million users every year. In the United States it is almost impossible to even talk about state owned digital infrastructures for workers. So I think one thing is to develop public infrastructure and public platforms to offer workers an alternative to private digital monopolies. This is a way to make people have real sovereignty, because they have alternatives.
In a paper that we just submitted, we are calling them “survival infrastructures”. People need them. My sovereignty as an individual cannot be only a checkbox on a website as it is in the GDPR model, when I have no alternative, like Margaret Thatcher enjoyed to say, to accept the Platform Data Regime. If I have a public platform that has a better data regime, this means that I actually can deny Google access to my data, because I am going to use the public platform instead. To have an alternative offered by the state is a fundamental dimension to reinforce rights for our population and citizens. Of course this demands that the state has a data regime that is different from private companies. It means enforcing privacy. At the same time, we need to calculate the tradeoff between individual rights to privacy and what I call the social function of data.
For example, I have a smartwatch. I know Samsung is going to access all my biometric data. But if I want to share this with the Brazilian public health system, I cannot. If I had public software or a public app in my smartwatch, I think the Brazilian state should have access to its population’s health data, of course with the consent of citizens. So I think another dimension is to balance privacy with the wealth and public knowledge that we can create with an open data future, if we start having public platforms.
We also need to have public funding to develop our own AI solutions. I think there is an AI bubble. I think the world economy is going to go into a huge depression in the next years because the AI bubble is going to collapse. But anyway, beyond all the hype, technology is important. We have to develop our own models based on our own languages. We must develop a really strong regulation to ensure data localization and to ensure access for public universities to data. We need to do this as a bloc, to have more power to fight against the huge pressure, the lobbying, and the technological dependence on companies from the global North.
—
Drawing by Carlotta Artioli, @charl.art.