Meta Faces Roadblock as Brazil Restricts Use of Personal Data for Training AI Models
Brazil has blocked Meta from using its citizens’ Facebook and Instagram posts to train artificial intelligence models. Learn more about the development and what it means for the tech giant.
- Brazil’s national data protection authority has banned Meta from using personal data from the country to train artificial intelligence systems.
- The decision arises from Human Rights Watch’s recommendations on purpose limitation and data privacy for children.
Brazil’s data protection authority (ANPD) has barred Meta from training artificial intelligence models on Brazilians’ personal data. The body has cited major risks of user challenges and legal and social damage. The decision follows changes to Meta’s privacy policy earlier this year, allowing the company to use public Instagram, Facebook, and Messenger to train AI.
Following the policy change, a report by the Human Rights Watch found that some of the image-caption datasets used for training AI models contained identifiable, personal photos of Brazilians, including children who are at risk of being used for deepfakes and other forms of exploitation.
See More: US Supreme Court Directs Florida and Texas Courts To Reexamine Social Media Laws
Brazil is one of Meta’s largest markets, accounting for over 100 million user accounts on Facebook alone. The ban will take effect in five working days, after which the social media giant risks facing daily fines of around $8,000. The development comes even as Spain’s prosecutors have started investigating Meta over suspicions of data-protection violations over its AI training program.
Meta has responded to the ban, stating that it would hurt innovation and delay the benefits of AI services to users in Brazil. This development gains importance as regulators in the EU have also criticized Meta for planning to use Facebook and Instagram posts originating from the region. How governments regulate the use of personal data assets for AI tools can have major implications for tech companies in the future.
LATEST NEWS STORIES
- Google’s Greenhouse Gas Emissions Surge by 48% in Five Years Due To Artificial Intelligence
- Google Launches kvmCTF Vulnerability Rewards Program to Improve KVM Hypervisor Security
- Microsoft Customer Emails Breached in Midnight Blizzard Hack
- Critical CocoaPods Vulnerability Makes macOS and iOS Apps Susceptible to Supply Chain Attacks
