Meta, which owns Instagram and Facebook, among other things, is pushing forward its plans to train artificial intelligence using content from its European users.
Starting at the end of June, the company would allow an AI model to use public posts, photos, and comments for so-called AI training, which has sparked widespread reactions from users. Now, the company's plans are being delayed, following a request from the Irish data protection authority.
Meta describes the decision as a "disappointment".
"We are convinced that our approach is compliant with European laws and regulations. AI training is not unique to our services, and we are more transparent than many of our industry peers," the company writes in a press release.
Meta has previously used material from its users outside Europe to, among other things, train the language model Llama. The company believes it needs to conduct AI training on European posts to "reflect Europeans' language, geography, and cultural references" in its AI services.
