Overview
- The pilot opens the AI deepfake detection tool to government officials, political candidates, and journalists after earlier rollouts to celebrities and roughly 4 million Partner Program creators.
- The system scans new uploads for AI-generated facial matches, alerts enrolled individuals, and lets them request removal through YouTube’s privacy process.
- Enrollment requires a video selfie and a government ID, with YouTube saying the data is used only for this feature, is not used to train Google’s generative models, and can be deleted if users withdraw.
- YouTube reports that removal requests have been relatively rare to date, and the company will continue to weigh exceptions for satire, parody, and political critique when evaluating flags.
- The company declined to name pilot participants, plans to expand access and pursue broader international rollout, is exploring voice and other IP detection and potential licensing or monetization options, and is advocating for the NO FAKES Act.