AI’s social media training: Meta Platforms, the owner of Facebook, announced plans to incorporate social media content from Europe to train its generative artificial intelligence models. The company will use publicly shared content from EU users on platforms like Instagram and Facebook to train its Llama large language models, as stated in a blog post on Monday.

The approach aligns Meta’s data usage in Europe with its practices elsewhere, despite stringent EU privacy and transparency regulations. Meta's policy executive mentioned in a September interview with Reuters that the company trains its Llama models using public posts while excluding private posts and messages.

In April, Meta's chief product officer indicated that the company was still determining the best way to implement this in Europe when releasing the latest Llama versions. Last month, Meta began notifying Facebook and Instagram users in Europe and the UK about its use of public information to improve AI.

However, the advocacy group NYOB has filed complaints across Europe, arguing that these notifications are insufficient. According to EU privacy rules, Meta must obtain opt-in consent from users for such data usage.

(With Reuters inputs)