YouTube is taking steps to protect creators from misuse of their likenesses with the help of AI. The platform is teaming up with the Creative Artists Agency (CAA) to create tools that allow creators and celebrities to explore AI-generated content that uses their image, face or voice. This partnership aims to give individuals greater control over how their digital likeness is used, especially as artificial intelligence becomes more sophisticated and widespread.

Early next year, YouTube will test these tools with celebrities and athletes. The idea is to let them find videos on the platform that use AI to mimic their face, voice or other aspects of their identity and make it easier to request removal of such content. Once this initial phase is complete, YouTube plans to expand the program to include top creators, creative professionals, and other influencers on the platform. The move is expected to protect YouTube’s most prominent users, who often have to deal with issues like impersonation and misuse of their image.

In September, YouTube had already announced plans to develop tools that help creators manage AI-generated illustrations, including voices. Now, the company is delivering on that commitment by giving celebrities the ability to tackle these issues on a larger scale. This is particularly relevant because AI tools can now create surreal versions of someone’s face, voice or body, which can be used in ways the original person never agreed to.

A key player in this effort is the CAA. The agency represents many of the world’s biggest stars and is already preparing for such a situation. Last year CAA launched a system called CAAVault. The tool scans and stores detailed digital records of its customers’ likenesses, including their faces, bodies, and voices. By combining this technology with YouTube’s platform-wide tools, the partnership aims to make it easier for creators and celebrities to monitor and control how their digital identities are used online.

YouTube is also taking steps to address another growing problem – the misuse of AI to replicate the singing voices of creators or artists. The company is working on a new “synthetic-singing recognition technology” designed to detect AI-generated content that mimics someone singing. Music labels have already started requesting the removal of AI-generated songs that sound like their artists. YouTube’s new tools will further strengthen these efforts. Earlier this year, YouTube began requiring creators to label videos that contain AI-generated content to ensure transparency. This means that anyone uploading AI-generated work must disclose it in the video’s description or metadata.