On a cool, sunny morning I take an Uber to the USP campus to interview Professor Roseli Figaro, from the School of Arts and Communications at the University of São Paulo. Riding through the city, the difference between the more central neighborhoods and the wealthy southern ones is immediately visible: more greenery, bigger houses, less noise, fewer people. The flip side of greater affluence is a corresponding rise in the number of assaltos—robberies—and in the state of alert of those who live there, the Uber driver tells me.
He is incredibly kind. When I realized I had entered the wrong address, he offered to take me, free of charge, to the university. I have no cash, but I insist on paying and ask for his PIX number; later I send him the money through a friend. As we chat, I discover the man is, in fact, precisely the subject of the interview’s case study. After Ford shut down in São Paulo, where he worked for twenty-five years, he has been unemployed and cannot find a job, so he supplements his income as a driver. They pay poorly, he says—still smiling—and there is neither sick leave nor vacation.
I thank him and get out on campus, which is enormous. It is one of the most tree-filled areas of the city, and the various buildings are separated by green spaces where you can see long green corridors of tipuana, large trees with dark, deeply furrowed bark, very common along the city’s streets as well. After flowering they produce winged fruits like tiny propellers. Looking for the right building, I pass the stalls by the School of Psychology and Education, where they sell books by Lacan, Winnicott, and many others.
 
The trees in the USPI campus
The entrance to the School of Communication and Arts
The campus atmosphere is very different from the Fundação Getulio Vargas (FGV), which I visited the day before: a brand-new downtown building that feels almost militarized, with automated turnstiles, cameras, an excess of guards at the entrance, and elevators so technologically advanced that I had to ask how to call them. FGV’s interiors, inspired by “smartness,” reflect a neoliberal design, anonymous and uncannily similar to IKEA living rooms, which can only halt the development of any form of critical thought.
Roseli’s office is the exact opposite: green, modernist interiors, worn PVC steps, and books used as décor. She is a fascinating, courteous figure: a professor with Neapolitan parents who emigrated to Brazil, and an old-school Marxist. Before the interview she offers to have lunch with me at the university’s self-service cafeteria. During the interview, conducted in Portuguese, I cannot help being struck by how Marxist analyses are always sharper: starting from farther away, yet arriving closer to the mark – as the invasion of Venezuela in these days makes clear.
Roseli Figaro in her office
***
What is your trajectory and why are you interested in digital sovereignty? Was there any moment that changed your perspective?
Talking about trajectory and choices…This topic has never left me, because it is part of my story as a person born to working-class, poor parents. I was the first daughter in the family to go to university, the only university professor. I worked as a journalist in the trade-union press, in militant media, and my focus was always on understanding why workers do not understand their own situation of exploitation. That was the issue that brought me back to the university. I graduated in journalism at 21 and, after 10 years working, I returned to university to study the trade-union press, the discourse of the trade-union press that was being produced in the 1980s in Brazil.
After that research, I did my PhD, going deeper into that topic. I sought to understand how workers carried out their processes of reception and meaning-making based on news coming from different outlets: television news – Jornal Nacional was very important, especially in the 1990s – the trade-union press, the religious press, the grassroots neighborhood press and the corporate press. I went to study shop-floor workers at Mercedes-Benz of Brasil. My PhD involved 600 shop-floor workers. This study brought me to the understanding that it is in the world of work, in social relations, that meanings are produced. These workers, who had access to different media outlets and to their own trajectories and points of view, used the workplace to confront and discuss those ideas with other colleagues, building there their viewpoints and their ideological clashes.
This showed me that the world of work was a central mediation for communication processes. I deepened this study until 2005. I went to deliver my book to Armand Mattelart in Paris, because he was a very important interlocutor for me. Jesús Martín-Barbero was also very present and a crucial interlocutor in this process. Until 2007–2008, I deepened this study, observing that communication was a fundamental mediatior of the world of work, not only to produce communicational meanings and the construction of social interactions, but that work and communication have always gone hand in hand. When I went to study ergonomics and ergology in France, in Aix-en-Provence, I began to analyze concrete work situations as communication processes. I understood that it is not possible to work without communication. This was a quite innovative perspective on work, because in certain strands of Marxism, communication is seen as something alien to work.
Looking at work as a communicational process was very important to me. And when I went back to study Taylorism, Fordism and, later, Toyotism – which I explored more in my thesis – I was able to prove that Toyotism, for example, does not introduce any new equipment or physical technology. The technology is social: reorganization of collectives, of work processes, of inputs and outputs of products.
That is why I formulated what I call the “binomial” of communication and work. In my post-doctoral research, I deepened this issue drawing on ergology, which studies the human being at work as a “body-self” that makes use of itself and “lets itself” be used by the other. It differentiates prescribed work – rules, procedures – and effective, real work, the work that is carried out in the unprecedented moment of action. The novelty of work requires communication, because it is in exchange, in interaction, that we recreate our work actions.
From then on, I focused my projects on the world of work of communicators, especially journalists, but I supervised countless studies on other categories: printing workers, call-center operators, book editors, advertisers, domestic workers and textile workers more recently, load handlers in commerce. I was always observing work as a communicational process. In this way we have followed the transformations in the socio-technical basis of work since the 1990s: electronic lathes, robots in factories, digitalization of processes and, in the case of communicators, the arrival of digital technologies and now artificial intelligence into the production process.
When we talk about digital sovereignty, what exactly are we talking about? How has this debate evolved in recent years in Brazil and in Latin America?
To talk about digital sovereignty, first we need to separate what happens in Brazil and in Latin America from what is determined by the hegemonic discourse. The word “hegemony” may even sound outdated, but it is central here. Sovereignty relates to a State. And for a State to be sovereign, it cannot be dependent. Since the 1960s in Latin America, we have had dependency theory. A critical, Marxist-inspired strand – Ruy Mauro Marini, Celso Furtado and others – addresses dependency as an obstacle to sovereignty and as something that blocks Latin America’s development.
There is also the post-war context, the creation of the UN, multilateralism: the idea of interdependent sovereignty, mutual respect among States, cooperation. That holds until the 1960s, until the Vietnam War, and then comes the wave of military dictatorships in Latin America, a U.S. project to keep these countries within its orbit of dependency. With Brazil’s re-democratization in the 1980s, this discussion about dependency and sovereignty does not re-emerge strongly; it seems “old” because, unconsciously, we have already incorporated dependency as something natural, in a non-critical framework. This framework said: “this is how we are going to develop capitalism here and allow the elite and middle class to access the goods that the North already has”.
With digital technologies and the shift from manufacturing industry to the data industry, the game changes. This new industry needs natural resources (water, energy, lithium, silicon, rare earths) and cheaper but qualified labor, which it seeks in the Global South. And it needs societies that accept that subordinate position. In this context, there is a strong push to redefine sovereignty: not as sovereignty of the State, but as individual autonomy, as “ownership” of personal data. This is profoundly damaging for a democratic society because it reinforces neoliberal individualism. “I am autonomous because I have my data”, “I fulfill myself alone”. But 80% or 90% of our population is poor and will never enjoy that kind of “autonomy” in the same way.
At the same time, a progressive-sounding discourse appears that bets on deepening the consumption of technology as a way out. Added to this is a vocabulary – post-human, post-industrial, actor-network theory – that, in my view, shifts the focus away from critical analysis. The center should be human activity in building the self and society, and these concepts end up obscuring the material means of that construction – in particular, the capitalist system. That is why I bring communication and work as the axis: we analyze working and communication conditions and how that builds society. Instead of this, we have a simulacrum of sovereignty, a simulation of individual autonomy. Sovereignty, in its full sense, is something else: it is the capacity of the modern democratic State, with popular participation, to preserve rights, natural resources, scientific and technological capacity.
And what about the discourse of popular digital sovereignty, such as some movements advocate here in Brazil?
Popular digital sovereignty, as formulated by sectors of the Homeless Workers’ Movement (Movimento Trabalhadores Sem Teto) and others, is close to what I am saying, but it removes the word “State”, because the liberal capitalist State is seen as irredeemably negative. This brings together anarchist, Trotskyist and other strands. I find it very positive to emphasize popular sovereignty. But we cannot mislead people: this sovereignty can only be built through a State. There must be a leading body. If it is a bourgeois State, of course it has limits. But what is the concrete alternative? A popular State. You don’t hold assemblies with 200 million people every day. You need organization, institutionalization of representation, networks, so that popular sovereignty can be exercised. The State, for me, is this form of organization and institutionalization.
I have studied at length the genealogy of platformization. There is a corporate strand – Google, Toyota, etc. – but there is another that arises from attempts to improve State planning: giant computers in the Soviet Union, then Cybersyn in Chile, and today what happens in China. This makes me think that we need public digital infrastructure…
Exactly. That is where the “popular” comes in, in the sense of power. When we talk about popular digital sovereignty, we are talking about another State, about another hegemony. But to get there, we need to defend this current State against its full capture by the most reactionary forces. Today, the State often only interests the elites as a repressive apparatus against progressive populations and ideas. Transforming this State means changing its content: from a basically repressive State to a State capable of bringing together and representing other forces. It is a struggle for hegemony within the State, not its abstract negation.
Do you think platformization in Brazil has specific characteristics?
When we talk about platformization, we can understand it in several ways. In a more general sense, it is an attribute for companies that own digital technologies and present themselves as mere intermediaries: a “meeting place”, a “facilitator” of commercial exchanges through their proprietary technologies. This is the dominant idea here. Big Techs arrive in Brazil with this narrative and dominate the market. How? By platformizing the entire production chain, that is, subjecting other businesses to their logic. Look at the case of the restaurant and food-producer network in Brazil and iFood. What we predicted in 2019–2020 – that shopkeepers would complain about iFood – is happening. The company has appropriated the entire chain: restaurant customers, restaurants’ data, customers’ data, knowledge of the production chain. With that, it can regulate the price that the restaurant can offer and controls that relationship, creating enormous dependency. That is platformization.
This spreads throughout commerce. Any store with an e-commerce website depends on cloud infrastructure to store data, customer records, sales history. It needs to use software – this “toolbox” – to run its system. Who provides this toolbox are Big Techs. It is another production chain, but now subordinated to their logic. It is radically different from what we had until the 1980s–1990s with the metalworking industry, food industry, etc., which operated in parallel chains, connected to transport, but not subordinated to a single digital infrastructure. Today, all chains end up subordinated to the logic of the tools offered by Big Techs. We have become much more dependent.
There is also the platformization of work. These companies operate with a small core of highly qualified workers – algorithm, software and hardware developers – although even this group is now starting to face devaluation. Alongside them, there is an army of workers scattered across the world, service providers with no recognized employment relationship, performing fragmented tasks, paid per piece – a 19th-century logic revisited. Lower wages, intense competition, a huge global reserve army, with specialized niches: Venezuela, Brazil, Kenya, for example, working as data annotators for AI, detectors, moderators, etc.
This platformization brings a new form of precarization: segmentation, individualization, competition among peers, while at the same time hiding the boss. The worker competes with other workers, often on the other side of the world, without seeing who controls the platform. Platformization operates on these two fronts: production chains and work. I also really like the spider-web metaphor: the spider spins its web, captures insects and then consumes them. The platform acts a bit like this: it sets the web, captures data, work and relationships, and begins to continuously extract value from this entanglement.
Can we connect all this to the case of PIX? Why was it built in Brazil? What do you think about what the Brazialin government has been doing? 
Isn’t it great? This is national, popular sovereignty. I have not done a specific study on PIX, but the issue is very interesting. The Brazilian ruling class is truly terrible – that is the word. It knows how to take advantage of every situation. PIX, for me, is proof that there is technology, institutional capacity and qualified labor in Brazil. But it is also proof of how financial capital knows how to realize itself in an accelerated way. Capital is realized in circulation; the faster it circulates, the more it is valorized. For the Brazilian ruling class – bankers, the financial system – it was extremely interesting to accelerate its realization through PIX, without going through certain intermediaries and competitors. So PIX deliverz a public infrastructure, technical qualification, a very important popular demand, making people’s lives easier; but all this within the logic of capital, allowing money to be put into pockets faster. That is my thesis: it is technological sovereignty in a certain sense, but at the service of a financial system that remains hegemonic.
We are living through a moment in which these platforms begin to dictate, even to formally independent countries, what it means to “be sovereign”. They arrive selling digital sovereignty solutions to nation-states. What is happening in Brazil is emblematic: separate packages for Google, Microsoft, etc., handing over to these companies the data operations of 11 important institutions that hold population data – health, services, public policies, banking data. All this is wrapped in the discourse that Microsoft will only operate the infrastructure, will build data centers in Brazil, with our money, and that the cloud would be “sovereign”. Why? Because Serpro would be involved. But instead of strengthening Serpro directly – with investment, technology transfer, building internal capacity – the option is to outsource the core of operations to a Big Tech, calling it a sovereign cloud. It is not sovereign. How could it be, if another company controls the cloud? I like the rented-house metaphor: you bring your furniture, build your life there, but the house belongs to someone else. One day the owner knocks on the door and says, “I want the house back.” That is not sovereignty, it is managed dependency.
I am absolutely critical of the way the Brazilian government has been handling this. I had the opportunity to speak briefly with President Lula. He listened to me for about 11 minutes, at a meeting of the National Council for Science, Technology and Innovation. I basically brought him the message I’m giving you here. He listened, thought it was great, told them to publish the text, but in practice what is done is to sign documents that go in the opposite direction. Politics requires negotiations, compromises; external pressures are enormous. In my view, a large part of policy implementation in the United States today is driven by the interests of Big Techs, and the most radical spokesperson for these policies is the side that supports Trump. When these companies cannot get their deals approved, or when countries start passing laws that regulate their activities, they put on pressure. The reaction comes as threats, trade wars, sanctions. It is a very ill-intentioned policy, to say the least.
And platform regulation? How is that legislative debate going?
Regulation is very stalled. We have Bill 2338, which deals with AI, approved in the Senate with several cuts and now stuck in the Chamber of Deputies. At the same time, there is a flood of other bills, many of them clearly written to order for the platforms, seeking “light”, business-friendly regulation. Bill 2630, which aimed to regulate social networks, was defeated in 2022 after a massive campaign by Meta and Google in the media and on their own channels. We received messages like “Careful, the government wants to censor”, that is, outright disinformation. They were fined, but they pay the fine and that’s it. We do not have a Congress committed to popular sovereignty – not even to minimal sovereignty. The portrait of Congress is, to a great extent, the portrait of the most aggressive Bolsonarism. And this is reflected even in issues such as taxation of betting, financial pyramids, “bets”, predatory fintechs. Regulation gets stuck there as well.
You talk a lot about ideological confusion, especially on the side that calls itself left-wing. What do you mean by that?
I use “left” in heavy quotation marks, because it has become too broad a label. In today’s Brazil, just by saying “human rights” you are classified as left-wing. Saying “racial rights”, same thing. Saying “we need to regulate platforms” already puts you on the “left”. So we have everything from a neoliberal left, which accepts the logic of the market with some social cosmetics, to more critical positions. This arc is too wide to solve our analytical problems, but it does show internal diversity – which is real. Within this arc there are various views on platformization, regulation, sovereignty, national development.
There are positions – which are not mine – that see technological development as always positive: “it is inevitable, we must adapt and make the best of it”. From there comes the idea of inevitability and adaptation, even when there is talk of preserving some rights. This generates what I call ideological confusion. It makes it difficult to build organized forces with more lucid diagnoses about what platforms are, what sovereignty is, what these technologies mean when we accept them in a subordinate position – the impact on natural resources, local populations, labor, science.
Sovereignty involves producing knowledge autonomously, and we are losing that. One of the platforms’ strengths is their monopoly over the production of information and knowledge. How can we do autonomous science oriented toward collective well-being if we are begging Big Techs for data, if we do not have sovereign infrastructure, if we do not have adequate budgets? Even while doing a lot with little – and we do – today we lack infrastructure to develop technology of public interest, via public policy, and to train the next generation of scientists. Brazil’s scientific future is, to a large extent, compromised.
Maybe a more cheerful question: do you see spaces of resistance? Unions, associations, platform cooperatives in Latin America seem to be organizing…
If I am here talking to you, I am not a spirit that descended from heaven, right? If I exist and think this way, it is because there is a social base that makes this possible for me, that sustains me. And just like me, there are other colleagues such as Rafael Grohmann, Leonardo Foletto, Sergio Amadeu and so many researchers and activists you will talk to. We are the fruit of resistance that exists in academia and in Brazilian society. It manifests in trade unions, cooperatives, social movements such as MST and MTST, women’s movements, anti-racist movements. There is a strong cultural movement, artists defending their intellectual production. There is a vibrant, productive, creative force.
Our problem is that, because of all this theoretical and political confusion, we do not have a single channel for that power. In critical moments, we need, even with differences, to build that channel: a national instrument, with a few shared slogans, a minimal line of action. I lived through the struggle for re-democratization. In the 1970s, as a student, I saw the importance of organizing, learning from mistakes, building unity to win the amnesty, to push for direct elections, to win the 1988 Constitution. It was not the Constitution of our dreams, but it was what was possible.
Now it is similar: we need a political platform of unity around sovereignty, national and popular development, defense of natural resources and sustainability. We are going to host COP30, spend fortunes to bring people to Belém, and at the same time we are handing over natural resources to Big Techs so they can build data centers that consume water and energy, connected to the exploitation of rare earths, lithium, silicon, etc. It is time to clearly put that on the agenda and negotiate from there, not hand it over on a plate. If we do not have a government and leaders with clarity, who can explain this to the population, we will remain mere consumers of cell phones, thinking that this is a gift.
Can things be done differently? Yes, they can. We have scientists, resources, a huge country. What we lack is political strength. The Lula government is the product of a great alliance to defeat fascism in 2022. That means a government that is internally contradictory. Unity was built to win the election, but not to formulate a great sovereignty plan in this conjuncture. My hope is that, through international politics, especially in the BRICS, space will open to build sovereignty – not because I idealize China or Russia, but because they are examples of countries that, with all their problems, have managed to build more sovereignty.
And, in a pragmatic way, what steps could Latin America take in the next five years?
First, political action. We must escape the right’s tricks to create conflicts between Latin American countries – Milei is one example, Paraguay, etc. If we can strengthen networks with Mexico, Colombia, Chile – and today we are somewhat distant from these countries, which I do not fully understand –, that will be important. We have similar issues, and they can play a key role in building a more integrated Latin America, even if Mexico is, geopolitically, glued to the United States. It is also essential to deepen articulation with BRICS, seeking technology transfer. The Chinese are not “nice guys”; if you do not stand firm, they do not hand anything over. So we must negotiate hard.
Within Latin America there are big asymmetries. Bolivia is very rich in minerals, but has less scientific infrastructure and qualified labor than Brazil, Argentina, Chile, Mexico. Central America also faces many shortcomings in terms of scientific infrastructure. We, compared with the big powers, are small, but we are something. A policy of mutual assistance, of building things together, is fundamental. The very case of PIX is an example: an infrastructure that can be shared without turning it into an instrument of Brazilian imperial projection. Because, yes, many neighbors see us as a small regional empire. Stronger exchanges in the cultural, scientific and economic fields can reinforce Latin-American ties and increase our capacity to exert pressure vis-à-vis the United States and Big Techs.
One question on gender and race. How does the precarization of work, especially under platformization, relate to this?
My perspective is class-based. I am not identitarian, I am not post-structuralist. I respect those who work that way, but I consider it a serious theoretical mistake in the Brazilian case. We are a structurally racist society in which 60% of the Black and brown population is working class, living from selling their labor power – formally or precariously, informally. Brazil has never had 50% of workers in formal employment with rights; that has never happened. Our legacy of colonialism and slavery is very present. If we do not understand gender and race within the logic of colonialism crossed by class struggle, we will get nowhere. Otherwise, everything becomes a moral problem – “good” or “bad behavior” – and not a civilizational problem, a problem of power structure. Talking about “intersectionality” helps to a certain point, but for me it is the question of class that structures the others. The question of the feminine, for instance, goes beyond that of racism when we think of the Black working woman.
Black women in Brazil are on the last rung of the social hierarchy: below white men, below white women and, often, below Black men. What does this mean in practice? That she can be beaten by her husband, suffer harassment from her employer, be beaten in the street, raise children alone, be treated as marginal. She is the last. How can we treat this woman only from a gender perspective? It is enough to re-read Casa-Grande & Senzala with a critical eye to see how the feminine in the “big house” and in the “slave quarters” was a fundamental dividing line in the exploitation of bodies. Platformization and digital technologies have not changed this structure. That is the tragedy: we have advanced technologically, we demand complex cognitive skills, but we cannot, because of the power structure, change values so deeply rooted in our culture.
Don’t you think capitalism has changed so profoundly that we would already be in a state of techno-feudalism?
No, I do not. We are still in capitalism. Capitalism reinvents itself every day, and now it has reinvented itself in a way that, in my opinion, will still take about two centuries for us to find solutions to the problem we have got ourselves into. And I hope I am wrong.
The Machine that makes and Remakes