TikTok is trialling an opt-in tool, Likeness Detection, that searches the platform for AI-generated videos using a creator's face and lets that creator report them. Only a small group of US creators has it for now. Social media consultant Matt Navarra first spotted it, and a TikTok spokesperson then confirmed the test.
To take part, creators confirm their identity through the verification firm Jumio with a live selfie and a government ID. TikTok then looks for matching faces in AI-made posts, and the creator decides what to report. Matches go to the creator for review, and they can flag individual posts or whole accounts. The company says it does not keep ID documents and uses facial data only for matching.
The test comes on top of TikTok's wider labelling work. The platform says it has labelled more than 3 billion videos as AI-generated, has joined the steering committee of the C2PA provenance standard, and removed 86 million fake accounts in the first quarter. On the music side, its SoundOn distribution arm already screens uploads for unauthorised derivative works. It uses ACRCloud's detection service and sends flagged tracks for human review.
YouTube went first with a similar face-matching tool in late 2025. It later opened it to talent agencies, then to all adult users, and plans to add voice. The music business is pushing on the same problem: Sony Music has asked services to remove more than 135,000 impersonation tracks. Spotify is piloting a feature that lets artists approve releases before they appear on their profiles. The worry dates back to the 2023 fake Drake track, which gathered hundreds of thousands of streams before it was pulled. TikTok has not said whether or when its tool will reach more people.
Sources · 1