TikTok is testing a tool for detecting deepfakes using the appearance of authors.

TikTok has begun testing a new feature that helps creators identify unauthorized AI-generated deepfakes using their likeness. The new feature will allow content creators to find AI videos featuring their face and report violations.

To use the feature, test participants must verify their identity through the Jumio service by undergoing document verification and a real-time selfie check. The company does not store users' documents, and facial data is used solely for matching likenesses and identifying potential violations.

After verification, the system automatically scans AI-generated content that may use the creator's image. Users can then review the found materials and, if necessary, report the posts or accounts.

The tool aims to help creators protect their digital identity amid the growing number of AI-generated deepfakes. YouTube previously introduced a similar feature, making it available to all adult users.

Source: The Verge