AI Training on User Data: Privacy Advocates Threaten Meta with EU Class Action Lawsuit
EU activists threaten legal action against Meta over alleged misuse of user data in machine learning training - Meta facing potential EU lawsuit over AI training using user data: Data privacy advocates warn of legal action against the tech giant
Hey there! Meta, the parent company of Facebook and Instagram, is causing a stir among privacy activists. They're planning to tap into user data from their platforms to train their AI models. But here's the kicker - they're doing it without explicit user consent, relying on a "legitimate interest" claim instead.
Max Schrems, founder of Noyb (None of Your Business), isn't buying it. He thinks Meta's actions violate European Union (EU) law. "The European Court of Justice has already called out Meta for using the 'legitimate interest' argument for personalized advertising," Schrems explained. "So why are they suddenly claiming it for AI training?" In his view, Meta is putting profits over user privacy.
Noyb believes that Meta won't be able to comply with other GDPR rules, like the right to be forgotten or the right to correct false data, while using user data for AI training. Moreover, Meta provides AI models like Llama as open-source software, meaning anyone can download and use them. This makes it tricky for Meta to recall or update models after publication.
Schrems highlights that Meta doesn't even need all user data for functional AI development. Companies like OpenAI and MistralAI, who don't have access to such vast amounts of data, have still produced AI programs superior to Meta's. On the other hand, a lawsuit could pose significant risks for Meta, including shutting down their AI system and paying damages to millions of users.
"We're considering our options for filing injunction lawsuits," Schrems stated. If Meta doesn't change its ways, they might face some serious consequences.
- *
EU* User Data* Facebook* Instagram* ECJ* AI Training
Insights:- Meta's current strategy for AI training is under scrutiny for potentially breaching GDPR, particularly in light of the ECJ's stance on "legitimate interest" and personalized advertising.- Legal experts and advocacy groups argue that explicit user consent is required for AI training using personal data, and Meta's "opt-out" approach may not be GDPR-compliant.- Users' rights, such as the right to be forgotten and the right to access personal data, could be compromised by using personal data for AI training without explicit consent.
- In the midst of controversy, Meta's use of user data from Facebook and Instagram for AI training is facing criticism from both users and data protection activists, as it appears to conflict with EU regulations, specifically the General Data Protection Regulation (GDPR).
- The European Court of Justice (ECJ) has previously contested Meta's use of the "legitimate interest" argument for personalized advertising, sparking conversations about the validity of this approach for AI training as well.
- The potential for an EU class action lawsuit looms over Meta, with Noyb threatening to file injunction lawsuits if Meta continues its current data practices, highlighting the intersection of technology, policy-and-legislation, politics, and general-news in this case.
