Meta has announced that it will soon begin training its artificial intelligence models using content shared by European adult users on its Facebook and Instagram platforms. The training data will include posts, comments, and interactions with Meta’s AI assistant, but will exclude private messages and data from users under 18.
European users will receive in-app and email notifications this week explaining the data that will be used and will have the option to object to their data being used through an online form. Meta has stated that it will honor all objections submitted through this form.
Data Usage and Objections
The company claims that this training will improve its generative AI models by better understanding and reflecting European cultures, languages, and history. This development comes after Meta delayed its AI training in June 2024 following complaints from NOYB, an Austrian nonprofit organization, in 11 European countries.
Regulatory Compliance
Meta has worked with the Irish Data Protection Commission and received approval from the European Data Protection Board (EDPB) for its approach, which meets the EU’s stringent data privacy laws. The EDPB’s approval came after Meta engaged constructively with the IDPC and addressed concerns raised by other EU data protection authorities.
Related Developments
In a related move, Meta has started rolling out its Meta AI assistant in WhatsApp across Europe. While personal chats will not be used to train Meta AI, conversations with the AI assistant may be used to improve future versions.
