Brazil's AGU demands Meta remove AI chatbots simulating child sexual content
Brazil's federal government agency AGU issued a legal notice to Meta demanding removal of AI chatbots created by platform users that simulate child profiles and engage in sexual conversations. The demand follows reporting that revealed the Meta AI Studio platform allowed such sexually explicit interactions with minor user accounts.
What happened
The AGU (Advocacia-Geral da União), Brazil's federal government legal agency, issued a legal notice to Meta on 15 August 2025 demanding the immediate removal of AI chatbots that simulate child profiles and engage in sexual conversations with users.
The chatbots, created by platform users through Meta AI Studio, had names such as 'Safadinha', 'Bebezinha', and 'Minha Novinha', and maintained sexually explicit dialogue while adopting child-like language and appearance, according to testing by Núcleo Jornalismo revealed in media reports.
Reported impact
The AGU stated that the chatbots carry potential to exponentially increase the risk of minors encountering sexually suggestive and potentially criminal material on Meta platforms, particularly affecting children aged 13 to 18 who use the platforms without age-based content filters.
The agency cited risks to children's psychological integrity and institutional harm, noting that the practice violates Meta's own Community Standards, which prohibit child sexualization content and sexual conversations in private messages with minors.
Index analysis
Meta platforms including Instagram, Facebook, and WhatsApp are permitted for users aged 13 and above, with no restrictions preventing minors between 13 and 18 from accessing sexually explicit content generated by the AI chatbots.
The AGU noted that Meta AI Studio was designed to simulate conversations with real people similar to Siri and ChatGPT, but lacked effective moderation mechanisms to prevent the creation and distribution of sexually inappropriate content targeting minors.
Governance controls
The AGU action occurred through Brazil's Procuradoria Nacional da União de Defesa da Democracia (PNDD) following a request from the Presidência da República's Secom (Secretaria de Comunicação Social).
The demand was issued as part of broader regulatory efforts by President Lula's government, which announced plans to submit a regulation proposal to Congress that would impose warnings, fines, and provisional platform suspension of up to two months for failure to detect and remove illicit content.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.