Social media giant Meta is resuming publicly available content on its platforms from European users to train its AI models. The company will use public posts and comments adult users share on its platforms, including Instagram and Facebook.
Meta had previously tried to use user data to train its AI models, but the process was halted after activists raised concerns about data privacy. Max Schrems, leader of Austrian NGO NYOB, raised alarms about Meta’s AI training programs and data usage, requesting various national privacy organisations to stop Meta from training the next generation of its AI models in the EU.
However, Meta claims in its announcement that a panel of EU privacy regulators confirmed that its original approach met all legal requirements. The company also launched its Meta AI assistant in March 2025 for European audiences, long after it debuted in other regions like Asia and the US.
The clearly worded announcement also promises that the company won’t use people’s private messages with friends and family to train its AI models. Public data from accounts of EU users under 18 will also not be used. Last but not least, users will have the option to object to their data being used for AI training, and Meta claims to honour all objections it has already received, as well as any newly submitted ones.
Meta isn’t doing anything; other giants like Google and OpenAI are already in Europe, as the announcement pointed out. It’s merely following the examples of its rivals and taking a cautious approach while doing so. Google and OpenAI have used EU user data to train their AI models, jumping through regulatory holes without much trouble. Meta is late to the party, but it’s good to see a big tech firm finally take a cautious stand for public privacy.
In the News: Cybercrime group rolls out major updates for on sale phishing tool
