Home / AI / Meta Restarts AI Training with European User Data

Meta Restarts AI Training with European User Data

Meta Restarts AI Training with European User Data

After almost a year-long hiatus prompted by regulatory concerns, Meta is resuming the collection of public content from European users to train its AI models. This comes as EU officials are poised to impose the inaugural fines under the Digital Markets Act (DMA).

Meta announced on Monday that it will begin utilizing public posts, comments, and AI interactions from adult users on Facebook, Instagram, and WhatsApp within the EU. This initiative aims to enhance its generative AI systems and has received approval from the European Data Protection Board (EDPB).

In its official statement, Meta explained, “This training will better support millions of people and businesses in Europe, helping our generative AI models to have a deeper understanding and appreciation of their cultures, languages, and histories.”

Previously, Meta was restricted from accessing EU data. In a 2024 statement, Meta noted, “Without access to EU user data, we can only provide a suboptimal experience. This limitation means we are unable to launch Meta AI in Europe at this time.”

Opt-out Options for Meta Users

This week, European users will start receiving notifications via their apps and email, detailing the data collection process and its purpose. These notifications will include a link to an objection form allowing users to opt out if they choose.

Meta emphasized, “We have designed the objection form to be easily found, understood, and used. We will respect all objection forms already submitted, as well as any new ones,” as stated in Meta’s press release.

The company reassured users that “we do not utilize private messages shared with friends and family to train our generative AI models.”

Moreover, “public data from the accounts of users in the EU under 18 years of age is excluded from training purposes.”

Meta also highlighted that it is following a precedent set by other companies like Google and OpenAI, which have already used data from European users for AI model training.

Anticipated DMA Fines

Meta’s announcement coincides with the European Commission’s readiness to levy significant fines against both Meta and Apple for purported violations of the new Digital Markets Act.

Competition Commissioner Teresa Ribera emphasized the Commission’s commitment to enforcement on Tuesday, remarking to the European Parliament: “If there is no willingness to comply, we will not hesitate to impose the fines prescribed by the law.”

Companies found in contravention of the DMA face fines of up to 10% of their total global revenue, which can escalate to 20% for repeat offenses.

Although the EU is intent on sanctioning Big Tech, the prevention of model training using citizens’ data remains a challenge.

European users who prefer not to have their data used by Meta should be vigilant for upcoming notifications.

Deje un comentario

Tu dirección de correo electrónico no será publicada. Los campos obligatorios están marcados con *