Meta Secures Approval to Train AI with User Data from UK Facebook and Instagram Users

Meta has announced its intention to commence the training of its artificial intelligence models using publicly available data from Instagram and Facebook users in the UK. Initially, the company intended to leverage this data three months ago, but this initiative was paused due to regulatory issues. Now, Meta has indicated that it has acquired the necessary approval to move forward with the AI training process. The organization asserted that this step will allow the AI systems to better mirror the cultures of British users.

In an update released through its newsroom, Meta mentioned that it had previously suspended its plans to utilize user data in the UK after concerns were voiced by regulatory authorities regarding data usage and the opt-out capabilities for users. However, it expressed that constructive discussions have occurred with the Information Commissioner's Office (ICO), which has led to the resurgence of the initiative. The company communicated that it will start the training of AI utilizing this data in the upcoming months.

The reasoning behind this move, as stated by Meta, is to more accurately reflect British culture, history, and vernacular when individuals engage with the AI. Furthermore, the organization conveyed its commitment to showcasing the diverse cultures globally by relying on localized user information.

Meta elaborated on its data collection approach, emphasizing that it will refrain from obtaining information from users in the UK who are beneath the age of 18 years old. The company will not utilize private data, such as direct messages between friends and family, for AI training. The focus will be on using public information—like visible posts, comments, and public photos from adult users on Instagram and Facebook.

Beginning next week, adult users residing in the UK will receive notifications on Facebook and Instagram outlining what Meta is planning and how it intends to execute this initiative. These notifications will also inform users about how to access a format to contest the application of their data for training generative AI models. Users who have previously opted out will not receive these notifications, ensuring their data remains off-limits for this purpose.