YouTube is utilizing AI to combat the abuse of AI technology. Participants in YouTube’s Partner Program, who have achieved either 1,000 subscribers with 4,000 legitimate public watch hours in the past year or 1,000 subscribers with 10 million legitimate public Shorts views in the previous three months, are now able to use an AI tool intended to limit the proliferation of deepfakes. This likeness detection tool, introduced at Made on YouTube in September, seeks to recognize and control AI-generated content that features an individual’s likeness.

In a video shared on its Creator Insider channel, YouTube clarified that the tool “enables you to effortlessly identify, manage, and request the elimination of unauthorized videos that may have altered your facial likeness or been created using AI—a vital measure to protect your identity and ensure your audience is not misled.”

Creators must initially verify their identity by uploading a photo ID alongside a brief selfie video. They can subsequently examine flagged videos within the Content Detection tab in YouTube Studio. If a video is confirmed to be AI-generated, they have the option to request its removal.

“Creators can currently request the removal of AI impostors, including alterations of face and voice, through our established privacy protocol. What this new innovation does is enhance that safeguard,” stated Amjad Hanif, YouTube’s vice president of creator products, to Axios in September.

The tool is currently accessible to certain creators within the YouTube Partner Program and will gradually be released to others in the coming weeks.

“YouTube’s aim is to develop AI technology that responsibly fuels human creativity, including the protection of creators and their enterprises,” remarked the company in its video. “We created this tool to assist you in monitoring how your likeness is represented—recognizing if others are making videos that use your facial likeness—to protect your identity.”