Interview with Sergio Amadeu. 
The introduction of Digital Tribulations, a series of intellectual interviews on the developments of digital sovereignty in Latin America, can be read here. 

I met Sergio Amadeu, a legend of the Brazilian free software movement, at a launch event hosted by the MTST [Movimento dos Trabalhadores Sem Teto / Homeless Workers’ Movement] in São Paulo in November 2025. I attended the event with a fellow researcher from Finland, my friend Darina. We initially struggled to find the venue, as the address was not clearly marked on Google Maps. When I asked for directions, a friendly woman opened a door to what looked like a refugee camp—something we hadn’t expected: a cluster of precarious small houses and tents built with aluminum. Though the camp seemed quiet when we arrived, it was clearly a living community, complete with shared kitchens, a vegetable garden, and other communal spaces.
The telecentro of the MTST.
Walking toward the edge of the site, we headed up a street and discovered a telecentro—a community computer lab designed for digital inclusion where residents of the periphery can access essential services, study, and learn technology. On the walls were quotes from Paulo Freire— the Brazilian pedagogue who famously said that “reading the world precedes reading the word.” Here, the slogan had been adapted to the digital struggle: “the reading of the world precedes the reading of the code.”
The sentence on the wall.
Nearby, a group of people had gathered for a conference set up under the shade of a tree. As the only non-Brazilians there, both blondes, we stood out; I overheard amused whispers from the crowd: “Olha, agora temos gringos!” (Look, we even have foreigners now!). We sat down and listened to the book presentation. Sergio spoke with charisma and intensity. When I approached him afterward, he graciously agreed to meet me later at a café on Avenida Paulista, in the heart of São Paulo.
What struck me most after my arrival in Brazil was the country’s staggering rate of digitalization. In many ways, it is far more digitized than Europe, though not always for the better. WhatsApp is not just a messaging app here; it is the primary interface for every person and every business. This is likely because free services offered in countries facing structural economic uncertainty quickly transform into essential infrastructure.
I was also struck by the lack of a widespread data protection culture. Personal data is collected aggressively by everyone, often unnecessarily. I once tried to use a laundromat, where the registration process was guided by the avatar of a black woman, but the system demanded so much personal information that I eventually gave up. Furthermore, biometric access is now standard for entering buildings in major cities. Without a CPF [Cadastro de Pessoas Físicas], the Brazilian individual taxpayer ID, life becomes incredibly difficult, from booking tickets online to accessing basic governmental services.
In this interview, we discuss the legacies of free software in Brazil and the competing visions and programs of digital sovereignty. We explore the challenges facing a country torn between the compromises of leftist governments and the looming risk of the far-right regaining power to use these very technologies for surveillance—a trend already visible in the policies of several Brazilian state governors.
***
What is your trajectory, and why are you interested in digital sovereignty?
I have been researching the internet since the last decade of the 20th century. I did my master’s on internet control and regulation and defended it in 2000. My background is in Social Sciences and Public Policy. I believe that public policy does not exist without consolidation into norms, rules, decrees, and ordinances. I have followed digital networks for a long time, both as a researcher and as an activist in the free software movement. From being a user, I became a contributor in open-source communities in 2003. I then coordinated the first free software implementation committee in the federal government during the first Lula administration. 
At that time, there were many clashes with large companies, especially Microsoft. The government was divided. Part of the government was interested in forging alliances with what we now call Big Techs. I advocated for the drafting of a presidential decree for the implementation of free software in the federal public administration, but this was not possible because it displeased a portion of the government, including ministers Palocci and Furlan.
I left the government, returned to academia, and continued researching digital networks. Currently, I research what is called artificial intelligence—which I prefer to call “actually existing artificial intelligence,” because there is a lot of mystification. By discussing the relevance of data for artificial intelligence, I began to frame the issue of digital sovereignty starting from data sovereignty. My relationship with the theme is as a researcher and as someone who comes from the open technology community, and who considers that technologies concentrate hierarchical, political, and economic determinations. 
When we talk about digital sovereignty, what are we talking about? Brazil has a privileged position in the discourse. How has this debate evolved in recent years?
The term digital sovereignty is under dispute. When I defend digital sovereignty, I bring it close to the idea of food sovereignty, a concept developed by popular movements, especially the peasant movement: the community has the right to choose what to plant and what to eat. By using the concept of digital sovereignty, we affirm the need for a minimum level of technological autonomy to define which technologies to develop and use, on what basis, and for what purpose. When we use only platforms and technologies from large North American corporations, we lose the ability to define basic elements of the development and use of these technologies. To be sovereign, technology needs to be appropriated by collectives—by national, local, and community collective intelligence.
And digital sovereignty involves data sovereignty. Data has high value in the digital economy and is a fundamental input for artificial intelligence. Our society should define which data will be created, its purposes, and, once created, how it should be used. When data goes to Big Tech data centers, the possibility of making decisions about that data is lost. Hence the term data sovereignty.
Looking at the trajectory of free software and open source, what worked and what failed? What can we learn, especially with the arrival of the smartphone?
There was a change in the technological landscape that made the use and development of free software more difficult. Mobile phones created devices that prevent you from using just any software. You are stuck with a certain hardware that requires a specific type of software. Very few people can remove what comes pre-installed and install other free software, and these often do not communicate with telecommunications operators. On the mobile front, this became complicated. Google, to face Apple, used Linux to create Android. It’s even a joke: the most used operating system in the mobile world is no longer free, because it is under Google’s control, even though it was born from free software and the Linux kernel licensed under the General Public License (GPL).
Another change occurred in the computer world: the computing paradigm migrated to the cloud. There was a massive outsourcing of infrastructure from governments, companies, and individuals to data centers controlled by large companies. You lose autonomy over the software because many systems are now used in the cloud. Curiously, clouds use free software. Most of AWS’s infrastructure runs on Linux; today, a large part of the services of Microsoft, which is the second-largest cloud provider in the world, does too. But all of this was captured by large companies. Despite having free licenses on the servers, they are under the control of corporations that transform collective work into private profit. It is necessary to liberate free software from big techs.
The lesson is that we need to act on several levels to build digital sovereignty. In Brazil, neoliberalism became the doctrine and logic of public managers, including part of the current government. It is not a minimal state; it is a state at the service of companies. To implement sovereign digital infrastructures, it must be done gradually, and free software is fundamental. Free software returns to the field of dispute allied with the struggle for digital sovereignty. It is no use saying we are going to remove all Big Techs from the game in public administration. There is no immediate alternative. We need to build concrete alternatives, which involves expensive infrastructure, technology policy, and the use of the state’s purchasing power in favor of digital and data sovereignty.
Large corporations know the strength of this agenda and launched the “Sovereign Cloud” product. They promise a “sovereign cloud,” but in practice, they install data centers in Brazil while maintaining control of the data, operation, and revenue. It is argued that, at least, the data would be safe in Brazil. It is not. The North American legal apparatus gives primacy to the U.S. state to control machines even outside Brazilian territory. In June 2025, the French Senate called Microsoft’s legal consultant and asked if the software and data running on the company’s computers in France would be subject only to French authorities. The Microsoft representative said: no! The Cloud Act obliges North American companies to comply with U.S. decisions wherever they are. If it is like this in France, it is no different in Brazil. Therefore, the “sovereign clouds” of the big tech companies and the products of Serpro and Dataprev that use Amazon, Microsoft, Google, and Oracle do not offer actual sovereignty.
What do you think of what the Federal government is doing to develop digital sovereignty?
The current government succeeded an administration full of neo-fascists who were destroying democracy—I am talking about the government of Jair Bolsonaro. The Lula government found an administration in a state of “scorched earth.” This created great difficulty. However, for the first time, federal government technology documents brought up the term digital sovereignty. But there is still more discourse than concrete projects. At the same time, Serpro operates as a broker, a reseller for Amazon, Oracle, and Microsoft. The discourse seems good, but the practice is the same as before. This needs to be changed. It is difficult to replace suppliers without concrete projects. Since they don’t launch projects, they fall into the neoliberal trap that everything has to be in real-time and requires immediate solutions: “let’s take everything off Amazon, and now what?” Now it is necessary to build free and sovereign infrastructures and management solutions here. Since they don’t build them, they stay trapped by the real-time requirement.
Ok, but I think also the idea of the Soviet Union projects or how Cybersyn worked was to improve planning with the use of real-time data.
Cybersyn began to be implemented by Allende in Chile. Brazil has something that Chile and Argentina do not: state-owned public data companies, which have now been distorted and have started to function as dealers for big techs. But we could reverse this with a new management policy. The current government could have made this inversion if it had this clarity, but it doesn’t. Most managers and leaders of the current government do not understand that technology is also a geopolitical apparatus.
I speak from a perspective I call data colonialism. We use not only the anti-colonial struggle of Fanon but also the vision of decoloniality from Latin American sociology—for example, Aníbal Quijano, a Peruvian sociologist who stated the following: colonialism, as a political structure, no longer exists. Brazil is sovereign. But coloniality remains in the epistemes, in the way of thinking, in the culture. Many left-wing managers are subjected to coloniality; they identify with the colonizer.
The camp just outside the center of SP.
Decolonizing the imaginary?
That’s it, liberating the imaginary. There is no single way to make automated systems, called artificial intelligence. There are several others, but we are subordinated to an exclusive type of thought and approach. Even those who do not agree with the concept of decoloniality could think from Marx and observe the profound alienation that operates around technology. There is a strong alienation between managers and political leadership. They discuss philosophy, politics, geopolitics, cooking, fashion, sports. But when it comes to technology, they buy whatever works, as if technology were neutral.
It is necessary to break this idea and reposition technology in the field of economy and culture. Technology is one of the greatest cultural pressures on a society. It is not external; it is not just technical. Technique expresses ways of facing problems that society has. The economic power of big techs derives from technological dominance. North American economic-military power does too. There is no racial superiority; that is a colonizing invention that remains in the imaginary.
Which was more or less the Gramscian concept of hegemony. And how do you evaluate the results of popular movements like the Movimento dos Trabalhadores Sem Teto (MTST)? Are there Latin American specificities in the struggle for the commons?
Brazil has many popular movements. The MTST [Homeless Workers’ Movement] is one of them. Another is the MST, Movimento dos Trabalhadores Rurais Sem Terra [Landless Workers’ Movement], which has great strength in the countryside and a strong technology sector. They have a sovereignty agenda and are now advancing in the defense of digital sovereignty. Together with the MTST, they defend “popular digital sovereignty” to differentiate it from the false idea of digital sovereignty from Big Techs. Technologies need to be appropriated and validated by the communities.
The technology group of the MTST decided to use available technologies and attract young developers, UX designers, system administrators, and data scientists. They were successful. They are starting to discuss the adoption of free software. They don’t use it fully yet, but they are moving in that direction and want to bring more people into the free software community. Free software communities in Brazil are numerous. There is, for example, the Casa de Cultura Tainã, which digitally articulated a network of Quilombos. Quilombos were communities formed by escaped enslaved people in the 17th, 18th, and 19th centuries, and their remnants exist to this day. In Brazil, for a long time, the ideology of racial democracy prevailed, which denied racism. For years the Black movement said there was racism in Brazil. Rede Globo, the main TV network in Brazil, said there was no racism in Brazil. Academia did not recognize racism in Brazil. But the struggle of the Black movement changed that reality.
Walking in the camp.
And then what happened? The internet helped connect these quilombola communities. The Rede Mocambos, with quilombola leadership in Campinas, built a network using free software a long time ago. Many quilombos use free software. They coordinate with groups like the Coletivo Digital. São Paulo had the largest network of telecenters with free software on the planet: 400 free internet access units in the peripheries, all with free software. Coletivo Digital is a non-governmental organization that was born from this experience of free telecenters in the city of São Paulo.
What are the specificities of the city and to what extent has it become platformized? Is it a smart city?
There are groups that defend smart cities. We are critical. In practice, a smart city becomes a set of sensors and cameras to watch the population and interconnect some services. There are movements like Tire meu rosto da sua mira [Get my face out of your sights], against biometrics and facial recognition. This is an intense dispute. There are state governments linked to the PT [Workers’ Party], such as Ceará and Bahia, which, through political alliances, handed over the security area to the right and adopted biometrics and facial recognition a long time ago. Except facial recognition does not reduce crime. They claim arrests of people with warrants, but this does not increase security. What is seen is more fear and the possibility of mapping people and controlling unwanted movements, mapping favelas and peripheries.
The issue of privacy here is very important because of the bias and the context of police violence. I’ll give an example. I live in Sumaré, a middle-class neighborhood. If a camera gives a false positive, a lot of police will come, but being an older, white man in the local context, probably nothing will happen. In the periphery, the outcome can be fatal. The police kill many Black people every day in Brazil.
When I leave home, I pass several totems with cameras. They can control my steps through the city. It is a society of distributed, totalitarian surveillance, unacceptable in any democratic regime. I don’t want to be watched by enemies who might be in power. Here in Brazil, there are dangerous groups like Bolsonaristas, the extreme right, and neo-fascists. The government of São Paulo belongs to Tarcísio, aligned with Bolsonarismo. In general, they link up with evangelical sectors to get votes and want cameras in schools. 
In Paraná, the governor says that teachers are untrustworthy people. He implemented an app—which should be a worldwide scandal—that forces teachers to take photos of children for attendance records via facial biometrics. Biometrics is sensitive data. Instead of a roll call, the system recognizes who was present. This reveals an authoritarian and technocratic vision. Neo-fascism is not just a regime; it is a process. And in Brazil, there is a junction of ruralism and financial capital behind it; although they present themselves as “against the system,” they are the worst of the system.
Let’s go to a more positive example: PIX. Do you see PIX as a virtuous case of public infrastructure? Why did PIX happen in Brazil?
PIX was already being studied, but the decision to launch it accelerated due to a specific context. And my doubt was always why the bankers agreed to PIX, since they earned from transfer transactions between banks. It’s because something very interesting existed in Brazil. Brazil is perhaps the first or second in the relative number of WhatsApp users. This is due, among other factors, to zero-rating. What is that? Most poor Brazilians do not have monthly paid phone plans. They have prepaid plans; they buy 20, 30, 50 reais of connection. Facebook made deals with operators so that the use of WhatsApp would not consume the data allowance/cap of prepaid users. Facebook pays for you.
WhatsApp is omnipresent: more than 90% of Brazilians use it. If you want to hire a house painter or service worker, you do it through WhatsApp; everyone gives you their WhatsApp. Thus, the Meta Group, owner of WhatsApp, thought: now that I have all the small businesses in my hand, I’m going to dominate the currency. The tendency was for WhatsApp to integrate payments and currency within the app. But the Central Bank launched PIX.
So it was against WhatsApp?
Actually, this is not talked about. But this is a hypothesis. WhatsApp was about to launch its own means of payment, and then the bankers who were studying PIX accepted it. The bankers accepted PIX also because traditional bank transfers were less used by poorer people. Still, many people today use PIX for large transfers, which impacts revenues. Who was most annoyed by PIX? It wasn’t Visa or Mastercard, but the Meta group, which went to ask Trump to attack PIX because Zuckerberg lost the goose that lays the golden eggs here.
But the idea of PIX was not born inside the banks; it was born in the Central Bank to speed up transactions, and this time it got off the drawing board. PIX became public infrastructure, widely adopted. It is unthinkable to reverse it. Lula would not end PIX; on the contrary, Lula’s decided support for PIX in the face of Trump increased the popularity of the Brazilian government.
Sergio and other speakers at the book launch.
What pragmatic steps can Latin America take in the next five years?
It is necessary to bring together structures, entities, movements, and democratically elected governments to build sovereign public digital infrastructures, starting with the universities. Brazilian and Latin American universities have their data in Big Techs. We need to start at the beginning: create sovereign digital public infrastructures, support community providers, and technological arrangements between public entities, not just federal ones, to develop regional solutions. This is possible despite the political instability of the region. 
Today, for example, under the government of Javier Milei in Argentina, there is great difficulty in participating in collective projects, as public infrastructure is being destroyed. This ideological project of the extreme right, inspired by authors like Nick Land and Curtis Yarvin, aims to implode the State from within and replace it with a privatized techno-management, with no space for democracy. They want an entrepreneurial monarch, a CEO, and the State should be a share of stock as a company is. In this context, regional cooperation in South America is shaken by the advance of the extreme right, but this cooperation is fundamental for us to build collaborative, shared, distributed, federated, free, and independent technological infrastructures from Big Techs.
The central kitchen in the camp.