Meta Restarts AI Training Using Public EU User Data Amid Regulatory Scrutiny

Meta Resumes AI Training Using European Public Data

After almost a year of pause due to regulatory concerns, Meta has resumed collecting public content from its European users to train its AI models. This move coincides with the European Union preparing to impose the first fines under its new Digital Markets Act (DMA).

Meta announced that it will use public posts, comments, and AI interactions from adult users on Facebook, Instagram, and WhatsApp within the EU to enhance its generative AI systems. The European Data Protection Board (EDPB) has approved this rollout.

How Meta Plans to Use the Data

According to Meta, this training will help its generative AI models better understand and reflect the cultures, languages, and histories of millions of people and businesses in Europe. Previously, Meta was barred from using EU user data and stated that without it, it could only provide a "second-rate experience," which prevented launching Meta AI in Europe.

User Control and Privacy Measures

European users will receive notifications through apps and email explaining the data collection and usage. These notifications will include a link to an objection form, allowing users to opt out if they choose. Meta assures that it has made the objection process straightforward and will honor all objection requests, both past and new.

The company also emphasized that private messages with friends and family are not used for training generative AI models, and data from EU users under 18 years of age is excluded from training.

Industry Context and Regulatory Challenges

Meta highlighted that it is following the example of other tech giants like Google and OpenAI, who have already used European data to train their AI models.

The announcement comes as the European Commission plans to issue significant fines against Meta and Apple for alleged violations of the DMA. The Commission has reiterated its commitment to enforcing the law, with fines reaching up to 10% of global turnover, increasing to 20% for repeated offenses.

Despite regulatory efforts, preventing AI training on citizens’ data remains a challenge. European users who wish to avoid having their data used by Meta should watch for these notifications and opt out accordingly.