In response to growing concerns and recommendations from its own Oversight Board, Meta announced it will begin labeling posts containing AI-generated images on it’s platforms including Facebook, Instagram and Threads. This move is poised to enhance user awareness across Meta’s vast network, hosting over 5 billion active accounts, especially ahead of numerous global elections in 2024.

The technical solution for automatic labeling is still under development, expected to roll out in the coming months. Meta plans to use metadata from major AI and technology firms like Google, OpenAI, and Microsoft to identify AI-generated content. In the interim, a new feature will allow users to disclose when they post photorealistic videos or audio that has been digitally created or altered, with Meta applying labels accordingly. Failure to use this disclosure tool could result in penalties, underscoring Meta’s commitment to combating misinformation and manipulated media.

This initiative follows critique from the Oversight Board on Meta’s previous policies, urging for a more coherent strategy towards manipulated media. By addressing these concerns, Meta’s labeling policy represents a pivotal effort in fostering a more informed and secure online environment as digital and AI technologies continue to evolve.

Sources include: Axios