In 2014, Brazil gave the world a radical idea – a constitution for the internet, written with society rather than merely for it. The Marco Civil da Internet (Civil Rights Framework for the Internet), signed after an unprecedented public consultation, declared that no one could delete what you write online without a judge’s order. It was an answer to the power of the platforms. Neither the State nor the company, but the law – spoken by the judiciary, case by case.
Twelve years later, the logic has flipped. Brazil has built one of the most interventionist content moderation regimes on the planet – not through a law debated in Congress, but through court rulings, presidential decrees, and principles drawn from a bill that died before becoming law.
The turning point was Bill 2630/2020, the “Fake News Bill.” Passed by the Senate in 2020, the text promised greater transparency from social networks, measures to combat bots and fake accounts, and liability for boosted content. But in the Chamber of Deputies, it became a battleground. Supporters saw it as a shield against misinformation; critics nicknamed it the “Censorship Bill.” Google warned that the law would create an “autonomous regulatory entity” capable of compromising the free flow of information. In April 2024, after years of stalemate and intense lobbying by big tech companies, the bill was shelved.
The legislative vacuum did not remain empty for long. In June 2025, the Supreme Court (STF) declared, by 8 votes to 3, the partial unconstitutionality of Article 19 of the Marco Civil – the very provision that required a court order for content removal. In one stroke, platforms could now be held liable without a court order in cases involving inauthentic accounts, paid content, and bot networks, with mandatory and immediate removal in cases of hate speech, terrorism, and attacks on democracy. The justices adopted the concepts of “duty of care” and “systemic risk,” borrowed from the European Digital Services Act. In May 2026, two presidential decrees (12,975 and 12,976) enshrined these rules and extended them to paid advertising, user traceability, and the protection of women online.
Who Pays the Bill
There is merit in the premise – the digital environment has already shown that, without counterweights, it can be weaponized against democracy itself. The January 8, 2023 attacks, for example, were largely organized on social networks. Protecting vulnerable groups and democratic debate is a legitimate function of the law.
But there is an invisible price. When platforms can be punished for omission, the mathematics of fear favors over-removal – political criticism, satire, and unpopular opinions may disappear before any substantive analysis, because the cost of erring on the side of removal is lower than the cost of erring on the side of retention. Researchers at FGV warn that, in the absence of clear parameters separating illicit speech from legitimate criticism, the regime risks being used by political elites to restrict lawful expression. Human Rights Watch has also recorded concerns that the judicialization of the issue – including the nationwide blocking of X in 2024 – may encourage platforms themselves to censor content out of fear.
The result is an institutional inversion. Instead of rules approved by popular representation, with clear procedures, appeals, and transparency, Brazil now has a regime shaped largely by the discretion of a handful of justices and the Executive. In this environment, freedom of expression risks depending less on clearly established rights and more on who interprets the rules and who supervises their enforcement.
The 1988 Constitution prohibits anonymity and punishes abuses of freedom of expression, but it never authorized prior censorship. The line between legitimate and illicit speech can and must be clearly drawn. The fair path, however, is to draw it through clear law, democratic oversight, and due process – not to replace the arbitrariness of algorithms with the arbitrariness of the State. The debate remains open, and the answer Brazil gives will help determine whether the next generation inherits a free internet or a managed one.
*Juliana Antonelli is the Director of Instituto Atlantos, an independent, nonprofit think tank based in Porto Alegre. She holds a Bachelor of Laws.
Source: We Are Innovation
