The Oversight Board announced that it has overturned Meta’s decision to allow anAI-generated, sexualized Instagram video impersonating a woman to remain on the platform. The Board found that Meta lacks adequate protections for non-public figures targeted by deepfakes, and highlighted the practical barriers to reporting these people face compared with public figures.
To reduce the burden on victims of AI-generated sexualised impersonation, the Board calls on Meta to let users designate trusted accounts to report violating content on their behalf, to create a separate reporting category for this type of imagery distinct from “harassment” or “nudity,” and to work with other AI developers to embed safeguards, content credentials, and watermarking at scale.
Mateusz Łabuz comments on the verdict:
“This is an important legal, social, and political signal that violence in the digital world should be treated exactly the same as violence in the real world. Pressuring tech giants makes sense as is often the only path that leads to change. We are glad that in this case we could contribute to a positive outcome.”
You can read the full decision here.