YouTube is rolling out a new AI-powered feature for Shorts that lets creators generate a digital avatar of themselves to use in videos. The tool enables users to create an AI-generated version of their likeness and voice, which can then appear in Shorts or be used to generate entirely new clips.
According to YouTube, the feature is designed to help creators insert themselves into content more easily without needing to film every segment manually. The company said avatars can be used in existing Shorts or in AI-generated clips built from text prompts.
How the avatar system works
To create an avatar, users must record a “live selfie” through the YouTube app or YouTube Create, capturing their face and voice from multiple angles while following on-screen prompts. Once the setup is complete, creators can generate clips featuring the avatar by entering a prompt and selecting the “make a video with my avatar” option. Generated clips are currently limited to eight seconds each, though multiple clips can be combined in a single Short.
YouTube said the avatar setup only needs to be completed once, though users can update or delete their avatar later if needed.
Restrictions and disclosure requirements
The feature is limited to users aged 18 and older who already have a YouTube channel. YouTube also said avatars can only be used by the creator who made them and cannot be accessed by other users to generate separate content.
All Shorts created using the avatar tool will include visible disclosures and digital markers identifying the content as AI-generated. YouTube said these include watermarking as well as SynthID and C2PA labels.
The avatar tool is the latest addition to YouTube’s growing set of generative AI creation features. YouTube said the feature will roll out gradually, with availability expanding over time.