Meta says it will resume AI training with public content from European users

London (AP) – Social media company Meta said on Monday that it will start using publicly available material from European users, which trains their artificial intelligence models, starts the work again put on hold After last year, activists expressed concern about data secrecy.

The company, who owns Facebook and Instagram, said it would train its AI system using public positions and comments shared by adult users in the 27-Nation European Union.

The company said, “People’s interaction with Meta AI – such as questions and questions – will also be used to train and improve our model.” blog post,

Meta is taking this step after launching its Meta AI assistant last month for European users, after a long time it rolled into the United States and other major markets.

The company’s AI training efforts were interrupted by strict European Union data privacy laws, which control people how their personal information is used. Vienna -based Group Noab under the leadership of activist Max Sheremes had complained to various national privacy guards about Meta’s AI training schemes and urged the company to stop their next generation of AI model.

Meta said that in December, a panel of the European Union’s secrecy regulators “confirmed” that its original approach fulfilled the legal obligations.

The company said it would not use private messages to train its AI model and reiterated that it is only following the example of rivals from Google and Openai, “both have already used data to train their AI models from European users.”

Meta said it would start informing users in the European Union about training, and will include a link to a form where they can object at any time.

“We will respect all the objection forms,” ​​the company said.