Meta to Begin AI Training on Public EU Content After Regulatory Green Light
Public Data from EU Users to Feed Meta’s AI Systems
Meta announced on Monday that it will start training its artificial intelligence systems using publicly shared content—such as comments and posts—from Facebook and Instagram users in the EU. The move follows a pause initiated last year after Meta faced regulatory challenges around privacy and data use in Europe.
In addition to public posts, Meta confirmed that interactions with its generative AI assistant, Meta AI, will also be leveraged to enhance model training.
Gradual Rollout Follows Delayed EU Launch
Meta AI, the company’s generative assistant, only recently became available in the EU after launching much earlier in the U.S. and other regions. The staggered timeline was largely due to the EU’s stringent privacy standards under the General Data Protection Regulation (GDPR).
While Meta has been training its models on U.S. user-generated content for several years, such practices in the EU required further regulatory scrutiny and a clearly defined legal basis.
Regulatory Challenges and Resolution
Back in June 2024, Meta halted its AI training plans for the EU and the UK after the Irish Data Protection Commission (DPC)—the body overseeing Meta’s compliance across the bloc—raised concerns. The DPC acted in coordination with several other European data protection authorities.
However, the company began to resume these efforts in September 2024, starting with public posts from users in the UK. Now, Meta is officially expanding this approach to the broader EU region.
“Last year, we delayed training our large language models using public content while regulators clarified legal requirements,” Meta said in a statement on its official blog. “We welcome the opinion provided by the EDPB in December, which affirmed that our original approach met our legal obligations. Since then, we have engaged constructively with the IDPC and look forward to continuing to bring the full benefits of generative AI to people in Europe.”
User Notifications and Opt-Out Options
As part of its rollout, Meta will notify EU users via in-app messages and email starting this week. These alerts will explain the use of public content for AI training and include a form for users to object to their data being used. The company has committed to honoring all previously submitted opt-out requests as well as new ones.
Importantly, Meta clarified that its models will not be trained using private messages or any public data shared by users under the age of 18 in the EU.
Building AI for Europe, with Europe
Meta emphasized that using EU public content is essential to building tools that reflect the region’s unique cultural and linguistic diversity.
“We believe we have a responsibility to build AI that’s not just available to Europeans, but is actually built for them,” the company stated. “That’s why it’s so 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. That means everything from dialects and colloquialisms, to hyper-local knowledge and the distinct ways different countries use humor and sarcasm on our products.”
Meta also pointed out that competitors like Google and OpenAI have already tapped into data from European users to train their own AI models.
Regulatory Oversight Continues
While Meta has now moved forward, scrutiny of AI training practices in the EU isn’t going away. Just last week, the DPC launched an investigation into xAI’s training of its Grok chatbot, signaling continued vigilance over how personal data is used in the development of large language models.
0 Comments