TikTok is testing a tool that lets creators ask the company to look for AI-generated content that may be using their likeness without permission, TikTok US spokesperson Zachary Kizer told The Verge. The test is limited at launch to “some” creators in the United States, according to Kizer.
The feature was spotted by social media consultant Matt Navarra. For creators who make a living with their face, voice, and audience trust, the useful part is obvious: TikTok is offering a way to search its own platform for synthetic lookalikes instead of making people hunt for them one post at a time. The less fun part is the verification step, because of course the anti-deepfake tool starts with another identity check.
How the test works
Creators included in the trial have to opt in before TikTok runs the scan. Kizer said participants must first verify their identity through Jumio, an identity-verification company. That process requires a live selfie scan and an ID check.
Kizer told The Verge that TikTok does not keep users’ ID documents. He also said facial information is used only to match likenesses and to help identify possible unauthorized uses of a creator’s likeness. That is TikTok’s claim, and it is the core privacy promise attached to the test.
Once verification is complete, TikTok’s system searches for AI-generated material that may resemble the creator. The creator can then inspect the results TikTok surfaces and report posts or accounts they believe are unauthorized.
The mechanics described so far are narrow. TikTok is not saying, at least in the details reported by The Verge, that the tool automatically removes anything. The creator reviews the findings and may report them. That distinction matters, because likeness-matching systems can produce false positives, miss altered material, or fail when synthetic media is edited enough to dodge detection.
A familiar platform problem
TikTok is not alone in trying to build this kind of control panel for synthetic identity abuse. YouTube has worked on a comparable likeness-detection tool and recently expanded access to all adult users, according to The Verge.
The pressure is easy to understand. Generative AI tools have made it cheaper to produce convincing imitations of real people, and platforms that host creator economies now have to offer more than a “please report this” button buried three menus deep. TikTok’s test is a small step toward that, though for now it is available only to selected US creators and depends on TikTok’s own scanning and review flow.
For creators outside the test, nothing changes yet. For creators inside it, TikTok is asking them to trade an identity-verification process for a new dashboard of possible AI impersonations. Whether that becomes a broad protection or another platform feature with limited reach depends on how TikTok expands it and how well the matching actually works.
This story draws on original reporting from The Verge.
