Effective moderation is possible, it's not difficult, and these platforms must do it
Em novembro de 2023, uma juíza do Rio de Janeiro ordenou que plataformas digitais utilizassem seus próprios algoritmos para localizar e remover ativamente conteúdos que expunham uma menor vítima de crime violento — contrariando o Marco Civil da Internet, que exige a identificação prévia de links específicos para que haja obrigação de remoção. A decisão, mantida em segunda instância, revela uma tensão crescente entre a arquitetura jurídica da era digital e a realidade de vítimas cujas imagens circulam livremente, ignoradas pelas ferramentas de denúncia das próprias plataformas. No horizonte mais amplo, o caso levanta uma questão que transcende o Brasil: até quando as leis concebidas para proteger a liberdade na internet poderão ser invocadas para justificar a inação diante do sofrimento humano concreto?
- Imagens de uma menor vítima de crime violento continuaram circulando nas redes mesmo após múltiplas denúncias formais — o TikTok chegou a concluir que o conteúdo não violava suas políticas internas.
- A juíza Fabelisa Gomes Leal rompeu com o Marco Civil ao ordenar que X, Facebook e TikTok usassem seus algoritmos para buscar e remover ativamente o material, sem depender de URLs específicas indicadas pela parte.
- O TikTok recorreu alegando que a lei exige identificação precisa do conteúdo, mas a desembargadora Marília de Castro Neves Vieira manteve a ordem, apontando que as plataformas já exercem moderação ativa quando lhes convém.
- O caso aguarda julgamento definitivo no tribunal, sem data marcada, enquanto abre uma fissura no principal marco regulatório da internet brasileira e pressiona por uma legislação mais abrangente sobre redes sociais.
Em novembro de 2023, a juíza Fabelisa Gomes Leal, da Vara Cível de Campo Grande, no Rio de Janeiro, emitiu uma decisão incomum: determinou que X, Facebook e TikTok utilizassem seus próprios algoritmos para localizar e excluir conteúdos que expunham uma menor vítima de crime violento — sem que fosse necessário indicar cada link individualmente. A ordem contrariava diretamente o Marco Civil da Internet, a lei de 2014 que estabelece que plataformas só são responsabilizadas quando descumprem ordens judiciais referentes a conteúdos especificamente identificados.
O caso remonta a 2022. A advogada Gabriella Ventura apresentou denúncias pelos canais oficiais das plataformas, todas rejeitadas. Ao recorrer à Justiça, pediu a remoção de trinta URLs e, além disso, que as empresas fossem obrigadas a buscar ativamente conteúdos idênticos ou relacionados. A juíza inicialmente ordenou apenas a remoção dos links indicados, mas, diante da insistência da advogada, ampliou a decisão: as plataformas deveriam rastrear e remover o material por conta própria, sob pena de multa diária de quinhentos reais.
O TikTok recorreu, argumentando que o Marco Civil exige especificidade. Em 7 de dezembro de 2023, a desembargadora Marília de Castro Neves Vieira negou a suspensão da ordem e a manteve integralmente, observando que as plataformas já realizam moderação ativa de conteúdos quando escolhem fazê-lo — tornando frágil o argumento de incapacidade técnica.
O caso expõe uma lacuna real: o Marco Civil foi concebido para evitar que plataformas se tornassem censores, mas acabou criando um escudo que, na prática, tem protegido empresas que ignoram reiteradamente ordens de remoção em casos de abuso sexual e exploração infantil. Para Ventura, a tecnologia evoluiu além do que o legislador de 2014 podia prever, e a interpretação da lei precisa acompanhar essa realidade. O desfecho do julgamento definitivo permanece incerto, mas a disposição judicial de tensionar os limites do marco vigente já sinaliza que o atual modelo pode estar chegando ao seu limite.
In November 2023, a judge in Rio de Janeiro issued an order that broke with Brazil's foundational internet law. Judge Fabelisa Gomes Leal, sitting in the civil court of Campo Grande, commanded three major social media platforms—X, Facebook, and TikTok—to do something the law said they didn't have to do: use their own algorithms to hunt down and delete harmful content without being told exactly where it was.
The case began in 2022, though details remain sealed. A minor had been victimized in a violent crime, and images documenting that crime were circulating on social media. The victim's attorney, Gabriella Ventura, filed complaints through the platforms' official channels. All were rejected. TikTok even reviewed the request through its dedicated channel for child safety violations and determined the content didn't breach company policy. Ventura then asked the court for emergency intervention, requesting removal of thirty specific URLs containing the images, and—crucially—asking that the platforms be forced to use their technological capacity to find and delete identical or related content on their own.
The initial ruling only partially granted the request. The judge ordered removal of the thirty links but said nothing about proactive searching. When Ventura pushed back on that omission, the judge issued a new order: the platforms must actively locate the content using their own technological means and delete it. Failure would cost them five hundred reais per day. This was a direct departure from the Marco Civil da Internet, Brazil's 2014 internet constitution, which shields platforms from liability for user-generated content except when they ignore specific judicial orders to remove identified posts. The law assumes platforms need to be told what to take down, not that they should hunt for it themselves.
TikTok appealed, arguing exactly this point: the Marco Civil requires a specific link, not a general mandate to search. But on December 7, 2023, appellate judge Marília de Castro Neves Vieira denied the request to suspend the order and upheld the active search requirement. She noted that platforms routinely identify and remove unwanted content when they choose to, so claiming inability to find related material rang hollow. The case now awaits full appellate review, with no date set.
The decision has opened a fault line in Brazilian internet law. The Marco Civil was designed to protect platforms from becoming de facto censors, requiring them to act only on specific, judicially identified content. But it has also created a gap: platforms have repeatedly ignored removal orders in cases involving sexual abuse and child exploitation, leaving victims and their families with limited recourse. Ventura argues that technology has evolved beyond what the 2014 law anticipated. "Effective moderation of harmful or illegal content is possible, it's not difficult, and these platforms must do it," she told reporters. "The evolution of technology demands an interpretation of the law that doesn't privilege big tech over citizens' rights."
The case reflects a global tension. Governments worldwide are grappling with how to hold platforms accountable without turning them into publishers. Brazil has no comprehensive social media regulation—the Marco Civil remains the primary framework—and this decision suggests the courts may be willing to push its boundaries when child safety is at stake. Whether the appellate court ultimately sustains this order, and what it might mean for future cases, remains uncertain. But the judge's willingness to depart from established law signals that the current framework may no longer be seen as adequate.
Citações Notáveis
The evolution of technology demands an interpretation of the law that doesn't privilege big tech over citizens' rights. Effective moderation of harmful or illegal content is possible, it's not difficult, and these platforms must do it.— Gabriella Ventura, attorney for the victim
Platforms routinely identify and remove unwanted content when they choose to, so claiming inability to find related material is not credible.— Appellate judge Marília de Castro Neves Vieira, paraphrased
A Conversa do Hearth Outra perspectiva sobre a história
Why did the judge feel she had to break from the Marco Civil? Wasn't the law clear?
The law was clear, but it wasn't working. The platforms had rejected every removal request through official channels. Even when the victim's attorney pointed them to thirty specific URLs, they complied but then stopped. The judge faced a child whose images were still circulating, and the law offered no tool to force the platforms to look for copies.
But doesn't the Marco Civil exist precisely to prevent platforms from having to police content on their own?
Yes. It was designed to shield them from liability for what users post, to keep them from becoming censors. But that shield has become a wall. Platforms ignore judicial orders in abuse cases and face no real consequence. The judge decided the law's logic didn't fit the reality of what was happening.
What did TikTok argue when it appealed?
That the law requires specificity—a link, a URL, a clear target. You can't order a platform to search for "similar content" without defining what similar means. It's a reasonable argument about how law and technology intersect.
And the appellate judge disagreed?
She said platforms prove every day they can find and remove content they want to remove. If they can identify policy violations when they choose to, they can identify this. The claim of inability didn't hold up.
Does this mean the Marco Civil is effectively dead?
Not yet. This is one case, and it's still on appeal. But it signals that judges may not accept the law's limits when a child's safety is at stake. That's a pressure point.
What happens next?
The full appellate court will eventually rule. If they uphold the active search mandate, it could reshape how platforms operate in Brazil. If they reverse it, the Marco Civil holds. Either way, the case has exposed a gap in the law that Brazil will have to address.