YouTube’s likeness detection tool shows suspected AI deepfakes uploaded on the platform.
Creators can use the tool to request the removal of deepfakes
Photo Credit: Reuters
YouTube released its artificial intelligence (AI) likeness detection tool for creators on Tuesday. The deepfake capturing tool is aimed at protecting the content creators on the platform from a third party using their likeness or voice via synthetic means. To get access to the tool, eligible creators will have to go through an onboarding process that will require them to submit an ID card and a video selfie. Once the process is complete, these creators will be able to see all the videos that YouTube deems to be an AI-generated deepfake in a dashboard.
In a video on its Creator Insider channel, YouTube announced the release of the likeness detection tool and detailed its working. Currently, it will be available to creators in the YouTube Partner Programme, and the access will be expanded over time. Once a creator has access to the tool, they will be able to see it in the Content ID menu, where users can monitor copyrighted content as well.
YouTube's likeness detection dashboard
Photo Credit: YouTube
The onboarding process to get access to the tool is long and thorough, likely to avoid instances where a scammer registers under a creator's name. The onboarding requires the user to consent to data processing, submission of a government-approved ID card, and upload a video selfie.
All of this will be stored on the Google server, and once the creator's identity can be verified, they can access the tool. YouTube will begin showing the videos where it suspects AI was used to create a deepfake. The video streaming giant will also categorise the videos based on priority to bring attention to the user.
YouTube highlights that since the tool is in its initial days, it can show the user their own videos alongside AI-generated videos. The company first piloted the feature in December 2024.
Creators can request the removal of the video or ask it to be archived. Once a complaint has been raised, YouTube will review the video and take appropriate action.
Notably, users will have the option to stop access to the tool whenever they like via the manage tool option on the dashboard. In case one chooses to disable the tool, YouTube will stop processing their data to scan for deepfakes after 24 hours.
Catch the latest from the Consumer Electronics Show on Gadgets 360, at our CES 2026 hub.