Data without a yes: Schrems wants to stop Meta's AI plans

“We are updating our privacy policy as we expand AI in Meta,” the email said. Meta relies on a “legitimate interest” in being able to use user data for artificial intelligence. This also means: Any person who does not object will automatically provide personal data to Metas AI.

And the data collection plans appear to be comprehensive: “Only after close examination of the notification did it become clear that Meta wanted to use personal posts, private photos or data from online tracking for unspecified ‘artificial intelligence technology’. This aims to record information from any source.” And possibly share it with unknown “third parties,” said a statement issued by noyb.

Personal chats are excluded

“The only exception appears to be conversations between individuals. However, conversations with companies are not protected.” The association has filed complaints in eleven European countries and is calling on the authorities to initiate emergency action to immediately stop this use.

Noyb wants to legally force Mark Zuckerberg's company to first ask its users for permission before training AI models with posts from European users.

Objection is possible via the form

According to data protection activists at noyb, Meta violates the European General Data Protection Regulation (GDPR), even if personal chat messages are excluded from the use of artificial intelligence. In the email from Meta, users are also informed of their right to object. The form asks why data is not being collected – if you object, you will receive confirmation within a short period of time that no data will be collected.

But instead of simply offering such an opportunity to object, under current law, Meta should have sought active consent from users before using AI, Noebe says. “Meta’s privacy policy would theoretically allow any intended use. All of this is very worrying because the personal data of some four billion people is at risk.”

The European Court of Justice has ruled against Meta in the past

Complaints have now been filed in Austria, Belgium, France, Germany, Greece, Italy, Ireland, the Netherlands, Norway, Poland and Spain. This is not the first action taken by the NGO Schrems against Facebook's parent company. The fact that Meta relies on “legitimate interest” has already been an issue in the European Court of Justice (ECJ). At that time it was about the use of personal data for advertising – and the European Court of Justice blocked this measure.

“With regard to the declaration, the European Court of Justice has already made it clear that Meta has no legitimate interest in ‘overriding’ users’ right to data protection,” Schrems said. “However, the company is trying to use the same arguments to train unspecified AI technology. It seems as if Meta is once again vehemently ignoring the rulings of the European Court of Justice.”

Meta rejects criticism

Meta rejected Noib's criticism, pointing to a blog post in mid-May in which the company said it used publicly available online and licensed information to train its AI, as well as information that people had shared publicly through Meta's services.

“We are confident that our approach complies with data protection laws, and our approach is consistent with how other technology companies are developing and improving their AI experiences in Europe (including Google and OpenAI),” Meta told Reuters news agency.