Meta is once again using public posts from European Facebook and Instagram users to train its artificial intelligence (AI) tools. The company announced on Monday that it will begin training its AI models using content shared publicly by adults across the 27 member countries of the European Union (EU). Meta’s AI assistant was recently introduced to European users, following its earlier rollout in the United States and other global markets.
The Associated Press reports that Meta’s AI ambitions were previously stalled by the EU’s strict data protection rules, which give individuals significant control over their personal information. The company’s initial plans were met with resistance from NOYB, a Vienna-based privacy group led by activist Max Schrems, which urged European regulators to intervene.
Meta’s AI Training Approach
Meta emphasizes that private messages will not be used to train its AI systems. The company also notes that its competitors, such as Google and OpenAI, have used data from European users to train their AI models. To comply with transparency requirements, Meta will soon notify EU users about its AI training plans and provide a straightforward way to opt out.
Meanwhile, in the United States, Meta CEO Mark Zuckerberg is facing a landmark antitrust trial that could potentially force Meta to divest Instagram and WhatsApp. The Federal Trade Commission (FTC) alleges that Meta eliminated competition by acquiring Instagram in 2012 and WhatsApp in 2014, establishing an illegal monopoly in the social media landscape. The trial is expected to last two months, during which the U.S. government will attempt to prove that Meta used its dominance to suppress potential threats rather than competing fairly.
