YouTube launches AI likeness detection to stop deepfake misuse
YouTube has launched a likeness detection tool for creators within its Partner Program. The technology is engineered to identify AI-generated content using a creator’s face or voice without consent, enabling them to request its removal from the platform.
This system aims to protect creators from the misuse of their identities, such as in videos creating false product endorsements, spreading misinformation, or causing reputational harm. The launch follows specific incidents, including a case where a company used an AI-generated clone of YouTuber Jeff Geerling’s voice to promote its products without his authorization.
To access the feature, creators must first navigate to a new “Likeness” tab in the YouTube Studio dashboard and agree to the platform’s data processing terms. The identity verification process requires the creator to scan a QR code with a mobile device. This action leads to a verification page where they are prompted to upload a photo ID and record a short selfie video for confirmation.
After successful verification, a dashboard displays videos that the system has detected as containing the creator’s likeness. From this interface, creators have several courses of action available. They can submit removal requests under YouTube’s privacy policies, file a formal copyright complaint, or choose to archive the video notifications for their personal records without taking further action.
The current availability of this likeness detection tool is part of an initial rollout phase. YouTube has stated plans to expand access to the feature to a wider pool of creators over time as the program develops.