Meta has announced plans to start training its AI models using data from European Union users of its apps, including Facebook and Instagram, with notifications to users beginning this week.
The data Meta will use for training includes public posts, comments, and chat history with Meta AI, but excludes private messages with friends and family. This data usage applies only to users over the age of 18. According to Meta, the purpose of using EU user data is to create AI models that better reflect the nuances of the region. This includes capturing dialects, colloquialisms, local knowledge, and the distinct ways humor and sarcasm are used in different countries, which is particularly important for the development of multi-modal AI that produces text, voice, video, and imagery.
Meta will notify its EU users about the AI training program this week through in-app notifications and email. The notifications will include a link to an objection form for users who wish to opt out of the data being used for AI training. This move follows Meta’s decision to postpone similar AI-training plans in Europe last year after being asked to do so by Irish regulators. As of the latest update, Meta’s privacy policy still indicates that the company is delaying its plans based on regulator feedback, although it is now proceeding with the notifications.
This development is part of Meta’s broader strategy to enhance its AI capabilities using user data from regions with robust data protection regulations, such as the EU and the UK. Last year, Meta announced plans to train its AI models on data from British users, who, like EU users, have stronger data protection rights compared to users in the US. It’s worth noting that Meta has already used data from adult Facebook users, including all text and photos they had publicly posted since 2007, to train its AI models, as admitted by the company last year.
