Brazil orders Meta to stop training its AI on Brazilian personal data

Brazil’s data protection authority (ANPD) has banned Meta from training its AI models on Brazilian personal data, citing risks of serious harm and inconvenience to users.

The move follows an update to Meta’s privacy policy in which the social network granted itself permission to use public data from Brazil’s Facebook, Messenger and Instagram, including posts, images and captions, for AI training.

After a Human Rights Watch report revealed that LAION-5B, a large dataset of image captions used to train AI models, includes personal and identifiable photos of children in Brazil, putting them at risk of deepfakes and other forms of exploitation.

Brazil’s National Data Protection Authority (ANPD) said the policy poses an imminent risk of serious and difficult-to-repair harm to the fundamental rights of users in Brazil.

The agency’s notice gives Meta five business days to comply with the order or face daily fines of about $8,808.

Meta’s response to the suspension

In response, Meta said its updated policy is in line with Brazilian privacy laws and regulations, and described the decision as a step backward for innovation, competition in AI development, and further delays in bringing the benefits of AI to Brazilian citizens.

While Meta claims that users can opt out of having their data used to train AI, ANPD argues that there are undue barriers that make this difficult.

Other sources also show that Meta has encountered similar resistance from European regulators, prompting the social media player to suspend plans to train its AI models on European posts on Facebook and Instagram. On the other hand, Meta’s new data collection policies are already in effect in the United States.