By André Lucas Fernandes, Beatriz Knippel and Leonardo Nogueira
In his first statement after winning the 2026 election, President Luiz Inácio Lula da Silva called for regulating digital platforms. He said he does not want Brazil to stay dependent on foreign artificial intelligence (AI) and that the country needs to train its own professionals to build a Brazilian AI.
Senate working group on artificial intelligence (Photo: Waldemir Barreto/Agência Senado)
Lula also named attempted interference in Brazilian democracy through digital platforms as one reason for regulation. What he said became plain during the campaign, after numerous complaints from the public, researchers, the United Nations and the Public Prosecutor’s Office.
Since before the election, the Rights in the Network Coalition (CDR) has warned about AI’s capacity for interference, influence and manipulation, and about the harm it can do directly in essential areas of the country, such as democracy, elections and sovereignty. The coalition has also said repeatedly that Brazil needs robust regulatory tools to hold AI systems accountable and to require transparency about decisions tied to how they operate.
Even so, for a year and seven months civil society has tried without success to engage with Rep. Aguinaldo Ribeiro of the Progressive Party in Paraíba state, the rapporteur of the AI bill (2338/23), which is moving through the Chamber of Deputies. Over that entire period, Ribeiro has met repeatedly with Big Tech representatives and lobbyists, and has even attended events with the same companies and groups.
It is worth recalling that the text now under review in the Chamber is the product of a long discussion that began in late 2020. While Brasília churns through its political maneuvering, the problems keep piling up. Disinformation, networks that produce echo chambers and AI facilitated gender violence, which has done massive harm to women and exploded this year, are all spreading. So is algorithmic racism, which attacks the integrity and lives of Black Brazilians. These cases become more common every day.
But what about the future? How can Brazil build guidelines for digital sovereignty and make sure the information economy produces wealth for society? The lack of a regulatory response weakens the goal of a sovereign country. The private sector keeps repeating its refrain of “innovation without permission,” even as it systematically disrespects rights and uses its logistical agents to erode the democratic process in Brazil. Against that backdrop, the regulatory structure proposed in Bill 2338/23 is a balanced set of obligations. It places no real burden on companies that are, in theory, able to develop software with basic documentation and governance processes.
If the burden is small for companies, the payoff for society could be structural. The measures would set clear procedures for civil liability and change how applications are developed, so that they stop treating our rights as something to trade away for the sake of development and an “AI miracle.”
Lawmakers seem to see their role less as rational mediators of the interests of all society and more as facilitators for lobbies of every kind. In civil society, we joke that every time deputies and senators attend a tech lobby event, a public statement violating rights and attacking collective interests follows.
In light of this, we need to propose something that may sound radical to some but is really just putting social democracy to work. Those who develop risky technologies, and the infrastructure behind them, must answer for the harm that results. In an era of global warming, there is no justification for failing to apply the precautionary principle when taking in data centers in Brazil. The multitrillion figure the Finance Ministry has been promoting cannot be eaten and does not turn into jobs. Without clear rules, the AI economy is just speculative money swirling around an economy that sells immateriality and spreads concrete harm.
The disorder caused by the lack of a regulatory program leaves Brazil as a mere consumer of technology, and its people as fuel for perverse systems of data processing, profiling and ultracapitalism.
Brazil built a multistakeholder culture, with meaningful public participation in the Marco Civil da Internet, its internet bill of rights, and in the General Data Protection Law (LGPD). Now the country stands idle in the face of clashes between the branches of government, depending on decisions by the Supreme Federal Court, whose legitimacy is being challenged. In this always boiling cauldron, it is impossible to build a regulatory environment that protects rights. It is also impossible to innovate against innovation rhetoric that merely reproduces Silicon Valley’s recipes.
To say yes to technology, a country has to be able to say no. As a sovereign state that aims to be developed, Brazil should steer technology toward the specific areas where it can act as a catalyst for positive social change. What we have today is the opposite. State investment is insufficient, nowhere near a “Big Science” project, as history shows, and it is trapped in a reductive discourse of fiscal responsibility that will never allow a Brazilian AI to develop. The state is subservient to lobbies and unable to build a framework of responsibilities that protects people and enriches the nation’s citizens, who hold true sovereignty through their democratic power. Its notion of development is limited to old ideas such as the triple helix of innovation, and its diplomacy is paralyzing. Taken together, they point to a great past ahead of us, not to a future.
The voice of the people should inspire Lula’s fourth government to turn its guidelines around 180 degrees, to rely on evidence and to build what Brazil, as a power built on solidarity, can offer that is different in the global geopolitical order of technology.
Leonardo Nogueira is a data journalist at data_labe. He previously covered politics, cities and human rights for O Globo, was a reporter at Agência Lupa and was a trainee in fact checking at Agence France-Presse (AFP).
André Lucas Fernandes is the founder of IP.rec, the Recife Institute for Research in Law and Technology. He holds a doctorate in law, focusing on AI and legal concepts, and teaches at the Federal University of Pernambuco (UFPE).
Beatriz Knippel is a lawyer with a postgraduate degree in Criminal Law and Criminology. She coordinates the Privacy and AI working group at Aqualtune Lab and is a consultant for Coalizão Direitos na Rede (the Network Rights Coalition).
This article first appeared in LeMonde Diplomatique Brasil, was translated to English with express authorization of its editor-in-chief, and can be read in its original Portuguese here.



