YouTube expands AI likeness detection to entertainment industry

On Tuesday, YouTube announced that it is expanding its likeness detection technology to the entertainment industry. The tool is designed to identify deepfakes—which are AI-generated content simulating a person’s likeness—allowing talent agencies, management companies, and the celebrities they represent to detect unauthorized uses of their faces. To develop and refine the tool, the video platform secured feedback from major talent agencies and management firms, including CAA, UTA, WME, and Untitled Management.

The technology functions similarly to YouTube’s existing Content ID system, which is the platform’s system for detecting copyright-protected material in uploaded videos. While Content ID allows rights owners to request removals or share in a video’s revenue, the likeness detection tool performs a similar function specifically for simulated faces. The system scans for visual matches of an enrolled participant’s face to help combat privacy policy violations and prevent public figures from being featured without permission in scam advertisements. Notably, entertainers do not need to operate their own YouTube channels to utilize the feature. Once the system flags a match, users can choose to request a removal for privacy policy violations, submit a copyright removal request, or take no action. However, the platform notes that it will not remove all flagged content, as its rules permit parody and satire.

The rollout follows a pilot program launched for a subset of YouTube creators last year, which expanded more broadly this spring to include politicians, government officials, and journalists. Although the technology is now reaching the broader entertainment sector, its current footprint remains limited. In March, the company noted that the amount of removals of AI deepfakes managed by the tool was very small. However, the company plans to expand the capabilities of the technology, stating that it will support audio in the future.

Why it matters

This expansion marks a significant step in YouTube’s effort to protect creators and public figures from having their identities exploited in AI-generated content without permission.