What is illegal in the real world is illegal online. Simple as that.
Por sete anos, o Supremo Tribunal Federal do Brasil contemplou uma questão que ressoa em todo o mundo digital: quando uma plataforma hospeda conteúdo que fere, humilha ou ameaça, quem responde por isso? Esta semana, em Brasília, os ministros deixaram claro que a resposta atual — ninguém — não é mais aceitável. O julgamento do Artigo 19 do Marco Civil da Internet representa não apenas uma disputa jurídica, mas um momento em que a sociedade brasileira tenta reconciliar a promessa de liberdade online com o peso humano da violência digital.
- Ministros do STF atacaram abertamente o Artigo 19 do Marco Civil, que protege plataformas de responsabilidade por conteúdo de terceiros enquanto não houver descumprimento de ordem judicial — uma imunidade que Toffoli chamou de escudo para a violência digital.
- O ataque à creche de Blumenau em 2023, que matou quatro crianças, tornou-se o símbolo mais doloroso do julgamento: nos dias seguintes, milhares de mensagens ameaçadoras inundaram as redes sem que as plataformas tivessem qualquer obrigação legal de agir.
- Alexandre de Moraes declarou que a autorregulação das big techs 'fracassou completamente', invocando até a escolha do Oxford Dictionary — 'brain rot' — como evidência de uma degradação digital global que o mercado não conseguiu conter.
- O Congresso teve anos para legislar sobre desinformação, mas o projeto de lei das fake news morreu em abril; o presidente Barroso anunciou que o prazo de cortesia ao Legislativo se encerrou — agora é a vez do tribunal decidir.
- O desfecho do julgamento pode exigir das plataformas uma postura proativa na moderação de conteúdo ou respostas mais ágeis a denúncias, encerrando efetivamente a era de imunidade quase total que moldou a internet brasileira desde 2014.
O Supremo Tribunal Federal brasileiro dedicou sete anos a uma pergunta que parece simples, mas carrega peso civilizatório: quando redes sociais hospedam conteúdo que machuca pessoas, quem é responsável? Esta semana, durante sustentações orais em Brasília, os ministros sinalizaram com clareza que a resposta vigente — ninguém — chegou ao fim.
O ministro Dias Toffoli, correlator do caso ao lado de Luiz Fux, expôs a contradição central do Artigo 19 do Marco Civil da Internet: as plataformas só respondem por postagens de terceiros se descumprirem uma ordem judicial de remoção. Enquanto essa ordem não chega, estão blindadas. Para Toffoli, isso significa que o que é ilegal no mundo físico pode circular livremente no digital. Alexandre de Moraes foi ainda mais contundente, declarando que a autorregulação das empresas de tecnologia entrou em colapso — e citou a escolha do Oxford Dictionary pela expressão 'brain rot' como sintoma de uma degradação que o mercado não soube conter.
O rosto humano do julgamento veio pelo relato do ministro Flávio Dino sobre o ataque à creche de Blumenau, em abril de 2023, quando quatro crianças foram mortas. Nos dias seguintes, milhares de mensagens ameaçadoras tomaram as redes sociais. Sob a lei atual, as plataformas não tinham obrigação alguma de agir. Dino descreveu aquele mês como um dos mais terríveis de sua vida.
O caso tramita no STF desde 2017 e foi adiado três vezes, a última delas a pedido da Câmara dos Deputados, que prometia votar uma lei sobre fake news — projeto que morreu em abril deste ano. O presidente do tribunal, Luís Roberto Barroso, foi direto: o Legislativo teve tempo razoável para agir e não agiu. Agora cabe ao tribunal decidir.
O que vier a ser decidido redesenhará o ambiente digital no Brasil. Os ministros parecem inclinados a estreitar a imunidade do Artigo 19, seja exigindo moderação proativa, seja impondo respostas mais rápidas a denúncias e ordens judiciais. De um jeito ou de outro, a era de proteção quase irrestrita às plataformas no país parece estar chegando ao fim.
Brazil's Supreme Court has spent seven years deliberating over a single provision of internet law, and this week, its justices made clear they believe the current rules are failing. The question at hand is deceptively simple: who bears responsibility when social media platforms host content that harms people? The answer, according to the court's leading voices, is nobody—and that is the problem.
Minister Dias Toffoli laid out the contradiction on Wednesday during oral arguments in Brasília. Under Article 19 of Brazil's Marco Civil da Internet, passed in 2014, platforms face liability for third-party posts only if they ignore a court order to remove the content. Until that order arrives, they are shielded. "We live in a world of digital violence," Toffoli said from the bench. "And digital violence is exactly what Article 19 protects, as long as there is no violation of a judicial decision." He repeated the point with visible frustration: what is illegal in the physical world must be illegal online. The law, as written, seemed to him to say otherwise.
Toffoli's co-rapporteur on the case, Minister Luiz Fux, is examining the same text. But it was Minister Alexandre Moraes who has emerged as the court's most forceful critic of the platforms themselves. In previous sessions, Moraes has argued that self-regulation by social media companies has collapsed entirely. He cited the Oxford Dictionary's choice of "brain rot"—digital burnout caused by information overload—as the word of 2024, a symbol of the degradation he sees online. "The problem of low-quality discourse, hate speech, violence, bullying—it's not just in Brazil, it's worldwide," Moraes said. "Self-regulation has failed. We must preserve human dignity and honor. We must protect democracy and the rule of law."
The human stakes emerged most starkly through the testimony of Minister Flávio Dino. In April 2023, a man entered a daycare center in Blumenau, a city in southern Brazil, and killed four children. In the days that followed, thousands of threatening messages flooded social media platforms. Dino described that month as one of the most terrible of his life. The violence was not confined to the physical attack; it metastasized online, reaching into homes across the country through feeds and notifications. The platforms, under current law, bore no responsibility for any of it.
The Marco Civil da Internet has been waiting for this judgment since 2017. The case entered and left the court's docket three times in recent years. The most recent delay came after the Chamber of Deputies requested a postponement, hoping Congress would pass its own fake news legislation—a bill that died in April. The court's president, Luís Roberto Barroso, acknowledged the long wait. "The tribunal gave the legislature a reasonable period to act," he said. "It did not. Now it is our turn to decide."
What the court decides will reshape how platforms operate in Brazil. Moraes has already cited the Marco Civil in his decision to suspend X, the platform formerly known as Twitter, citing its refusal to comply with Brazilian court orders. The justices appear poised to narrow the immunity that Article 19 grants. The question is how far they will go—whether they will require platforms to police content proactively, or merely to respond more swiftly to complaints and court orders. Either way, the era of near-total platform immunity in Brazil appears to be ending. The court has waited long enough.
Notable Quotes
We live in a world of digital violence. And digital violence is exactly what Article 19 protects, as long as there is no violation of a judicial decision.— Minister Dias Toffoli
Self-regulation has failed. We must preserve human dignity and honor. We must protect democracy and the rule of law.— Minister Alexandre Moraes
The Hearth Conversation Another angle on the story
Why has this case taken seven years to reach a decision?
Congress was supposed to legislate on this. The court waited for them to act, but they never did. A fake news bill died in April. At some point, the court decided it couldn't wait forever.
What's the actual disagreement? Is it about who should remove harmful content?
It's deeper than that. The law says platforms only owe money if they ignore a court order. The justices are saying that's not enough—that platforms should be accountable for the harm they enable, not just for defying judges.
But platforms host billions of posts. How could they possibly police all of it?
That's the tension. The justices aren't saying platforms must catch everything. They're saying the current rule—immunity unless a judge says otherwise—is backwards. It puts the burden on victims to sue, then sue again when platforms ignore the ruling.
The Blumenau attack—why did that matter so much to the court?
Because it showed the real-world consequence. Four children died in a physical attack. Then thousands of threats appeared online. The platforms hosted all of it without consequence. That's what the justices mean by digital violence being real violence.
What happens if the court rules against the platforms?
It depends on how far they go. They could require faster responses to removal requests, or they could impose liability for algorithmic amplification—content the platform actively promotes. Either way, Brazil's platforms will face stricter rules than they do now.
Will this affect how X operates in Brazil?
It already has. Moraes suspended X partly using this law. A ruling that narrows platform immunity would give him and other judges more legal ground to enforce compliance.