TikTok is testing AI likeness detection for some US creators only

TikTok is testing an opt-in tool that searches for AI content using a creator's likeness and lets them report suspicious posts or accounts, according to The Verge. TikTok US spokesperson Zachary Kizer says the test initially covers some US creators.

Likeness detection starts as a limited US pilot

Creators in the test first have to verify their identity through Persona. The process includes a real-time selfie scan and an ID check. TikTok says it does not retain ID documents and uses facial information only for likeness matching and identifying potential unauthorized use.

After verification, the system scans for AI-generated content that may use the creator's likeness. The creator can review matches and report them to TikTok.

Platforms are moving deepfake disputes into identity workflows

For creators, this is practical. A deepfake likeness is not only a reputation problem, but also a commercial risk: fake ads, scam accounts or content that damages audience trust.

For the platform, it is a sensitive shift. To protect a creator's likeness, TikTok first needs enough information to verify identity. Protection from AI misuse creates a new trust-heavy process inside the same platform.

Protecting a face depends on who receives the face

The weak point is not only detection, but trust in the whole data flow. TikTok says it does not store ID documents and limits facial information use. For a ByteDance-owned social network, that will still be politically and reputationally sensitive in the US.

It is also a pilot for selected creators, not a general protection for every user. People outside the program remain dependent on normal reporting and platform rules.

Response speed and expansion beyond the pilot will matter

The key signal is not just how many deepfake videos the system finds. It is how quickly TikTok acts on reports, how many false matches it creates and whether the tool expands beyond some US creators.

The Verge notes that YouTube has been working on a similar tool. If likeness detection becomes a standard feature, platform competition shifts from generating AI content to protecting the people whose faces make the platform valuable.

Lilith's verdict

TikTok is framing the detector as a camera at the door of creators’ dressing room. A camera can protect, but the first question is who holds the footage and how fast they open the door when a stranger walks in.