The short video platform TikTok has begun testing a new tool called Likeness Detection. The feature, currently available only to a limited group of American creators, is designed to identify content created using artificial intelligence that uses a blogger's face without their consent.
The tool's mechanism works as follows: the system automatically scans the platform for AI-generated videos featuring a specific creator's face. If such content is detected, the creator can promptly file a complaint against the deepfake or fake account.
To activate the feature, users must verify their identity through a third-party service called Jumio. The process involves taking a real-time selfie and uploading a scan of an identifying document. TikTok representatives emphasize that copies of documents are not stored on the platform's servers, and biometric data is used solely to find matches with suspicious content.
In my opinion, this is an important step in combating reputational risks for creators. In an era when synthesized videos are becoming indistinguishable from real ones, such tools are not a luxury but a basic necessity. However, the effectiveness of Likeness Detection will directly depend on the scale of its implementation: as long as the feature is only available to a select few, most bloggers remain vulnerable. I hope TikTok scales the solution globally in the coming quarters.