YouTube has launched a feature that allows users to create an AI avatar that resembles them for use in Shorts. The new functionality, teased earlier this year, is now available through the main YouTube app and YouTube Create.
The avatar creation process involves users capturing a “live selfie” by recording their face and voice while reading several prompts. This results in a photorealistic avatar that can be used in YouTube Shorts, with each prompt resulting in an avatar generation lasting up to 8 seconds. Users can create multiple clips consecutively, with the initial setup needed only once.
This avatar feature builds on the existing Google Veo models integrated into YouTube Shorts. While a similar function for ingredient-to-video creation, which allows users to upload pictures, has been available since last year, the voice component is new to this feature.
YouTube stated that “the avatar feature gives users an easier way to include themselves safely and securely in videos.” The company emphasized that the selfie video and voice data will only be utilized for avatar creation, ensuring that no one else can use the avatar for original Shorts. Users can delete their avatar at any time, and YouTube will automatically remove inactive avatars after three years.
All videos featuring avatars will include watermarks and digital labels, such as SynthID and C2PA, along with disclosures indicating that the content is AI-generated. To access the feature, users can navigate to the Create ‘+’ section within the YouTube app and select “Create video” with their avatar. The feature is also available in the Remix menu under options like Reimagine.
The rollout of this feature began today globally, excluding Europe, and is available to users aged 18 and older, with full availability anticipated in the coming days. Users must have a pre-existing YouTube channel to create an avatar.
