Yesterday, Brazil’s National Data Protection Authority issued a preliminary ban on Meta’s use of personal data of users based in Brazil to train its artificial intelligence (AI) systems.
The decision stems from “the imminent risk of serious and irreparable damage or difficult-to-repair damage to the fundamental rights of the affected data subjects,” the agency said in announcing the ban.
The news follows Human Rights Watch reporting in June that personal photos of Brazilian children are used to build powerful AI tools without their knowledge or consent. In turn, others use these tools to create malicious deepfakes, putting even more children at risk of harm.
The National Data Protection Authority’s decision included two arguments that reflected Human Rights Watch’s recommendations. The first is the importance of protecting children’s data privacy, given the risk of harm and exploitation that results from their data being scraped and used by AI systems. The second centers on purpose limitation, and that people’s expectations of privacy when they share their personal data online—in some cases, years or decades before these AI systems were built—should be respected.
Meta has been using its US-based users’ publicly-posted personal data to train its AI models since last year. Last month, Meta paused its plans to do the same in Europe and the United Kingdom after objections from 11 data protection authorities. Yesterday’s decision effectively bans this practice in Brazil and imposes a daily fine of 50,000 reais, or about US$9,000, for failure to comply within five working days from notification of the decision.
Following the regulator’s decision, Meta said that it “complies with privacy laws and regulations in Brazil,” and called the decision “a step backwards for innovation [and] competition in AI development and further delays bringing the benefits of AI to people in Brazil.”
The Brazilian government’s decision is a powerful, proactive move to protect people’s data privacy in the face of swiftly evolving uses and misuses of AI. Yesterday’s action especially helps to protect children from worrying that their personal data, shared with friends and family on Meta’s platforms, might be used to harm them in ways that are impossible to anticipate or guard against.
 systems.){kind=link}