Meta Will Use EU Users’ Public Content To Train AI

Meta will train its AI using public posts and chatbot interactions from adult EU users, while offering an opt-out to protect user privacy.

Meta, the parent company of Facebook and Instagram, has announced that it will start using public posts and comments shared by adult users in the European Union, as well as interactions with its Meta AI chatbot, to train its artificial intelligence models.

After launching Meta AI in Europe last month, the company is now pushing forward with plans previously paused due to privacy concerns. Reuters points out that Meta’s rollout had been delayed after Ireland’s Data Protection Commission intervened last June, and advocacy group NOYB filed complaints urging regulators to block the company’s data use plans.

The Irish Data Protection Commission also imposed a €251 million fine on Meta last year because the company exposed 29 million user records through a 2018 data breach which affected 3 million EU users. Meta also received a $101.5 million penalty as part of a separate password security violation case.

Additionally, Meta received negative reactions from the public after its AI-generated profile feature appeared on Facebook and Instagram which led to negative feedbacks.

Meta announced that EU users will start getting notifications about their data usage and how to opt out in the near future. Users will have access to a form which enables them to object to their information being used for AI training.

The company emphasized that it “won’t use private messages or content from users under 18,” and WhatsApp will not be affected by this change. “People’s interactions with Meta AI — like questions and queries — will also be used to train and improve our models,” the company said in their blog post.

Meta added that the move is meant to improve AI tools that understand Europe’s many cultures and languages. “This training will better support millions of people and businesses in Europe, by teaching our generative AI models to better understand and reflect their cultures, languages and history,” the company said.

Although Meta originally avoided using European data, it now says it is simply following the lead of Google and OpenAI, who have already used such data for training their AI systems.

Meta claims it has made the opt-out form “easy to find, read, and use” and promised to respect all objections. France24 points out that the company is investing up to $65 billion this year on infrastructure to support its AI push.

However, critics remain concerned about the ethical risks and environmental costs of these powerful technologies. Reuters reports that the European Commission has not commented on Meta’s latest move. not commented on Meta’s latest move.