Meta Gets EU Regulator Nod to Train AI With Social Media Content

Meta Cleared by EU to Train AI Using Public Social Media Content

Meta has received approval from the European Union’s top data regulator to train its artificial intelligence models using publicly shared posts and interactions from adult users across its suite of platforms, including Facebook, Instagram, WhatsApp, and Messenger.

Announced in an April 14 blog post, the tech giant confirmed that only publicly available content from adult users will be used. This includes posts, comments, and questions submitted to Meta’s AI assistant. However, private messages and all data from users under 18 remain strictly off-limits.

Meta emphasized the importance of this data for building culturally relevant and effective generative AI models. “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. This includes everything from local dialects and colloquialisms to region-specific humor and sarcasm.

Opt-Out Options for Users

To respect user privacy and comply with EU laws, Meta said users will be given a clear and accessible way to opt out of having their public content used for AI training. The company plans to notify users via in-app messages and email, providing what it describes as an “easy to find, read, and use” form.

Meta has a green light from data regulators in the EU to train its AI models using publicly shared content on social media. Source: Meta

Regulatory Hurdles and Privacy Pushback

Meta’s plan had been on hold since July 2024, following privacy complaints filed by advocacy group None of Your Business in 11 European countries. In response, the Irish Data Protection Commission (IDPC) requested that Meta pause AI training activities until a thorough review was conducted.

At the heart of the complaints were concerns that Meta could potentially use years of personal posts, private photos, and behavioral data to power its AI systems — claims Meta denied. Now, after a detailed legal review, the European Data Protection Board has cleared Meta’s current approach as compliant with EU data protection laws. Meta confirmed it is continuing to work closely and “constructively” with the IDPC.

“This is how we have been training our generative AI models for other regions since launch,” the company said. “We’re following the example set by others, including Google and OpenAI, both of which have already used data from European users to train their AI models.”

Broader Industry Under Scrutiny

Meta is not the only tech firm facing intense regulatory oversight in the region. In September 2024, Irish regulators opened a cross-border investigation into Google Ireland Limited to assess whether the company’s AI training practices adhered to EU privacy laws.

Social media platform X (formerly Twitter) also faced regulatory pressure. The company agreed to halt the use of EU user data for training its AI chatbot, Grok, after being scrutinized under the same framework.

These developments come in the wake of the EU’s landmark AI Act, which came into effect in August 2024. The legislation outlines comprehensive rules governing the development and use of artificial intelligence, including strict guidelines on data usage, privacy, transparency, and safety.

With regulatory approval now in place, Meta appears poised to push forward in the generative AI race — this time, under closer European scrutiny and a more privacy-conscious framework.