The American social media giant Meta, targeted by complaints in 11 European countries, is suspending its plan to use the personal data of its users in an artificial intelligence (AI) program, the Irish Protection Commission said on Friday. of data (DPC) in a press release.
“Meta is suspending its plan to train its extended language model using public content shared by adults on Facebook and Instagram in the European Union (EU) and the European Economic Area (EEA),” announced the DPC, acting on behalf of the European Union (EU).
Meta, who could not immediately be reached, had been targeted since last week by the Austrian Noyb association, the bête noire of the tech giants, which had asked the authorities to intervene “urgently” to prevent the implementation implementation of this new confidentiality policy, scheduled for June 26.
“We welcome this development, but we will monitor it closely,” said the head of the Viennese NGO, Max Schrems, in a press release. “So far, there has been no official change to Meta’s privacy policy, which would make this commitment legally binding,” he added.
Noyb therefore does not withdraw the complaints at this stage.
If certain public data is already used to train generative AI models, Meta wanted, according to the Austrian association, to go further and “completely take” all the data of its billions of users collected since 2007.
The goal: to use it as part of an “experimental AI technology without any limits”, affirmed Noyb (for “None of your business”, meaning in English “It is not your business”), without asking the consent of the Internet user, although required by the reference European regulation on data protection (GDPR).
The Noyb association is at the origin of numerous complaints against web giants.
In the case of Meta alone, its actions led to “administrative fines of more than 1.5 billion euros”, she recalls.