Meta has announced that it will start training its AI models on public content from Facebook and Instagram users in the European Union (EU), after previously pausing plans due to data privacy concerns.
The company had faced resistance in the EU due to strict privacy laws, including the General Data Protection Regulation (GDPR). However, following clarification from regulatory bodies, Meta has resumed its efforts to train AI models using public posts and interactions with its users.
Starting this week, users in the EU will receive notifications explaining that Meta will use their public data and interactions to train its models. Users can opt out of their data being used by submitting a form through the app or via email.
Meta emphasizes that it only uses public data from users over 18 years old and does not utilize private messages. The company aims to build AI models that are tailored to European communities, understanding nuances such as dialects, humor, and hyper-local knowledge.
This move follows similar practices by companies like Google and OpenAI, which have used EU user data to train their AI services. However, the Irish Data Protection Commission (DPC) continues to scrutinize how large language model creators train their AI services, with a recent investigation into xAI’s training of Grok.
Source: https://techcrunch.com/2025/04/14/meta-to-start-training-its-ai-models-on-public-content-in-the-eu