Meta has announced plans to utilize public content — such as posts and comments — from adults on its platforms in the European Union to train its artificial intelligence models. Meta aims to enhance its AI models so they can better understand European cultures, languages, and history. This follows the successful launch of Meta AI in Europe last month.
Starting this week, users in the EU will receive notifications on Meta's platforms explaining the data usage, along with enhancements to the user experience due to this data collection. These notifications will also provide a form for users to object to the usage of their data, and Meta has ensured the process to file an objection is straightforward.
Meta assures that private messages from users will not be used in this training, and data from users under 18 will not be included. This move aligns with Meta's goal to make AI more accessible and tailored for European users through its platforms like Facebook, Instagram, WhatsApp, and Messenger.
The initiative aims to train AI models to grasp the nuances and complexities of European communities, covering aspects such as dialects, colloquialisms, and humor. Meta asserts that its approach is transparent and consistent with European regulations, as it looks to bring the full potential of generative AI to the region.
Meta's method adheres to legal standards and follows precedents set by companies like Google and OpenAI, which have used European data for AI training. Meta's strategy was previously on hold due to regulatory clarifications last year, but recent engagements with European authorities have affirmed their compliance with legal obligations.