The Austrian privacy advocacy group Noyb has issued a formal notice to Meta, challenging the company’s plan to use European users’ data for training its artificial intelligence models.
Last month, the US social media giant, which owns Facebook and Instagram, announced that starting 27th May, it would use all public content from its users, except WhatsApp, for its generative AI models, unless users opt out.
Max Schrems, founder of Noyb, criticised Meta’s approach. "Meta claims it needs to steal everyone’s personal data to train AI, which is absurd and laughable. Other providers manage without this and produce models that surpass Meta’s."
The launch of this feature had been delayed for over a year due to European regulations on new technologies, including the General Data Protection Regulation (GDPR) and the laws governing digital markets and AI.
Previously targeted by Noyb’s complaints in 11 European countries last year, Meta had to pause its project in June. This recent change of direction is justified by what the company calls “legitimate interest.”
“What legitimate interest justifies harvesting data” from everyone who has used Facebook or Instagram for 20 years? “It is neither legal nor necessary,” Schrems asserts.
Given Meta’s estimated 400 million active subscribers in Europe, Schrems believes that agreement from 10% of them would be “clearly sufficient for learning EU languages and similar idioms.”
Noyb emphasises that this notice is “the first step” towards a potential injunction or even a class action, warning Meta of “significant legal risks” if it continues its current strategy.
