TLDR

• Meta received approval from EU regulators to train AI models using public content shared by adult users
• Private messages and content from users under 18 will not be used for AI training
• Users can opt out via an easily accessible form that will be sent via app, email
• Meta delayed this training last year after privacy advocacy complaints
• Meta says other companies like Google and OpenAI already use European user data for AI training

Meta has been given the green light by European Union data regulators to train its artificial intelligence models using publicly shared content from its social media platforms. This approval comes nearly a year after Meta paused these efforts due to privacy concerns raised by advocacy groups.

Meta announced on April 14 that it will begin using posts and comments from adult users across Facebook, Instagram, WhatsApp, and Messenger to improve its AI models. The company will also use questions and interactions with its AI assistant as training data.

The tech giant emphasized that certain data remains off-limits. “We do not use people’s private messages with friends and family to train our generative AI models,” Meta stated. The company also confirmed that public data from EU account holders under 18 will not be used for training purposes.

User Notification and Choice

Starting this week, EU users will receive notifications both in-app and via email explaining what data Meta plans to use and how this will improve its AI services. These notifications will include a link to an opt-out form.

“We have made this objection form easy to find, read, and use,” Meta said in its announcement. The company promised to honor all objection forms, including those previously submitted when the initiative was first proposed.

This move follows Meta’s launch of its AI assistant in the European Union last month across Facebook, Instagram, WhatsApp, and Messenger platforms.

Why Meta Wants European Data

The company argues that diverse training data is essential for creating AI that works well for European users. “We believe we have a responsibility to build AI that’s not just available to Europeans, but is actually built for them,” Meta explained.

Meta points out that its AI needs to understand regional dialects, colloquialisms, local knowledge, and cultural nuances like how different countries use humor and sarcasm. This becomes even more important as AI models advance to handle text, voice, video, and images.

“It’s important for our generative AI models to be trained on a variety of data so they can understand the incredible and diverse nuances and complexities that make up European communities,” the company stated.

Regulatory Journey

Last July, Meta delayed its AI training plans after privacy advocacy group None of Your Business filed complaints in 11 European countries. The Irish Data Protection Commission then requested Meta pause its rollout until a review was completed.

The complaints alleged that Meta’s privacy policy changes would have allowed the company to use years of personal posts, private images, and online tracking data to train its AI products.

Meta has now received confirmation from the European Data Protection Board that its approach meets legal requirements. The company says it has “engaged constructively” with the Irish Data Protection Commission throughout the process.

Industry Context

Meta is not alone in using user data for AI training. In its announcement, the company pointed out that it’s “following the example set by others including Google and OpenAI,” noting that both companies have already used data from European users to train their AI models.

“This is how we have been training our generative AI models for other regions since launch,” Meta added, suggesting that European users have been the exception until now.

The approval comes as the EU has started implementing its AI Act, which establishes a legal framework for artificial intelligence technology that includes provisions for data quality, security, and privacy.

Other tech companies have faced similar scrutiny in Europe. X (formerly Twitter) agreed to stop using personal data from EU users last September after facing regulatory pressure. Google is also under investigation by Irish regulators regarding its AI development practices.

The post Meta Receives EU Regulatory Approval to Train AI on Social Media Content appeared first on Blockonomi.