(Credit: Dilara Irem Sancar/Anadolu via Getty Images)
Don't miss out on our latest stories. Add PCMag as a preferred source on Google.
YouTube has begun rolling out an AI-powered likeness detection tool to help creators spot videos where their face may be altered or generated using AI.
The tool is currently available to a limited set of creators in YouTube Studio's Content detection tab and will expand to all creators in the YouTube Partner Program over the next few months.
To sign up for the tool, creators will have to submit proof of their government ID and video scans of their face.
If a new video with their face gets uploaded, the tool will list them in the Content detection tab, where creators can view the flagged videos and request actions, such as copyright removal or likeness removal. They can also archive the listing if it seems harmless.
(Credit: Creator Insider/YouTube)Likeness detection works like Content ID, except that it looks for facial likeness instead of copyrighted audio or video, YouTube says. It is aimed at safeguarding a creator's identity and ensuring their audience isn't misled about what they endorse and what they don't, the company adds.
Setting up likeness detection can take up to five days and can only be done by the Channel Owner or users listed as Managers. Editors can take action on flagged videos, but they won't be able to set up the tool themselves.
In case a video with facial likeness fails to show up on the dashboard, creators can manually request a privacy review. YouTube says the tool is "still being tuned and refined" so there's a chance it won't pick up on all instances of your likeness being used.
The tool comes as video generation tools, including Google's own Veo 3 and OpenAI's Sora 2, have made it difficult for the audience to distinguish between actual people and their deepfakes. This tool was announced last year in partnership with Creative Artists Agency (CAA) and piloted in December to help actors and athletes take down their deepfakes.