TikTok trials tool to help creators find AI copies of themselves
TikTok is testing an opt-in system that scans for AI-generated likenesses and lets selected US creators report possible misuse.
By James Whitfield · Staff Writer
2 min read
TikTok has begun testing a tool meant to help creators find AI-generated videos or accounts that may be using their likeness without permission, according to The Verge. The test matters for creators because it gives some of them a direct way to look for possible deepfakes and flag them to TikTok.
The feature is opt-in and is being tried first with “some” creators in the United States, TikTok US spokesperson Zachary Kizer told The Verge. Social media consultant Matt Navarra first spotted the test, The Verge reported.
How the test works
Creators included in the trial must verify their identity before using the tool, according to The Verge. That check is handled through Jumio, a company that provides identity verification services.
The process requires a live selfie scan and an identity document check, Kizer told The Verge. Kizer said TikTok does not keep users’ ID documents and said facial data is used only to match likenesses and help identify potential unauthorized uses of a creator’s image.
Once verification is complete, TikTok’s system looks for AI-generated content that may feature the creator’s likeness, according to The Verge. Creators can then examine the matches TikTok surfaces and may report posts or accounts they believe are unauthorized.
The company is not opening the feature to all users at this stage, based on Kizer’s description to The Verge. TikTok has described the current rollout as a test, and The Verge reported no timeline for wider availability.
Platforms add deepfake reporting tools
The trial comes as major video platforms add tools aimed at AI impersonation. The Verge reported that YouTube has been developing a comparable likeness-detection system and recently expanded access to all adult users.
TikTok’s test focuses on creators rather than the broader user base, according to the details provided to The Verge. The reported workflow centers on identity verification, automated scanning and creator review before a possible report to the company.
The feature also places TikTok in the role of screening for possible AI misuse after a creator opts in. The Verge reported that the system scans for AI-generated material that may use a verified creator’s likeness, but creators decide whether to report the posts or accounts TikTok finds.
Kizer’s comments to The Verge addressed how TikTok says it handles sensitive verification data during the process. The company’s stated limits on retaining ID documents and using facial information are part of the test as described by TikTok’s spokesperson.
This story draws on original reporting from The Verge.