Metaās New AI Move: Powerful, Aggressive, and Under Fire. As of this month, Meta is now training its AI on public posts, likes, and comments from EU Facebook and Instagram users ā unless you opt out. No, this doesnāt include private DMs. Yes, the default is opt-out, not opt-in. And yes, regulators and privacy groups are already going after it. Why Meta says itās doing this: š§ To make their models more āculturally intelligentā in Europe š To compete with OpenAI and Google, who already leverage global datasets š To localize AI outputs ā humor, language, nuance Why privacy advocates are pushing back: š GDPR was designed for explicit consent, not buried opt-outs āļø āLegitimate interestā isnāt a free pass for mass data harvesting š« Training AI wasnāt the original reason users gave Meta their data Groups like noyb argue this crosses legal lines ā and theyāve already filed to block it across multiple EU nations. France, Belgium, and Hamburg are also stepping in. Germanyās courts, though? Theyāre letting it slide (for now). Hereās the bigger play: Meta is betting that public content = fair game ā legally and strategically. But the line between āpublicā and āpersonalā is thinner than ever. And as AI models train on billions of user interactions, the question becomes: At what point does optimization cross into exploitation? If youāre in the EU and donāt want your posts feeding AI, youāll need to manually opt out ā through a multistep form buried in Metaās settings. Ethical? Strategic? Inevitable? Drop your thoughts below š