Meta Gains AI Training Approval from UK

Meta has gained approval to use public posts from U.K. users to train its AI models, marking a significant step in its AI development efforts. After halting its program in July due to regulatory concerns, Meta negotiated with British authorities to resume using content shared by adults on Facebook and Instagram for this purpose.

This data will help refine Meta’s generative AI models by improving their understanding of language and human interaction, specifically incorporating British cultural nuances. This could improve a number of elements of Meta’s AI-focused technology, especially given the likely change that AI will become a much bigger part of the various social media platforms that Meta operates.

Meta has emphasized that only public posts, comments, and images from adult users will be used, ensuring that private messages or posts from users under 18 are not included in the training dataset. The company highlights that this process aligns with legal provisions under “legitimate interests” in U.K. law.

This approval follows similar negotiations with Brazilian authorities, although the European Union remains cautious. Meta had to introduce an opt-out option for E.U. users, per the bloc’s Digital Services Act, allowing them to prevent their content from being used in AI training. Users uncomfortable with the prospect of their content being utilized for AI training are encouraged to make their profiles private, and Meta plans to notify U.K. users of this development in the coming weeks.