How To Use YouTube”s AI Deepfake Music Generator

Artificial intelligence has infiltrated the music industry. Although the software isn’t new, its acceptance by significant figures and entities such as the Grammy Awards is. Despite musicians like Drake and Bad Bunny’s public disdain for AI-generated music, YouTube supports the technical advancement.

Yesterday (November 16), Variety announced that YouTube revealed its partnership with Google’s DeepMind artificial intelligence lab had pumped out another feature for users. The new feature is titled AI deepfake music generator, and several entertainers, Alec Benjamin, Charlie Puth, Charli XCX, Demi Lovato, John Legend, Papoose, Sia, T-Pain, and Troye Sivan, have agreed to allow users to utilize their vocal likeness.

How To Use YouTube’s AI Deepfake Music Generator

So far, YouTube’s AI deepfake music generator has limited capabilities. To use the feature, users must upload a video within YouTube Shorts. Before publishing the clip, the user can add AI-generated music to their queued clip. To do this, users enter a prompt within Dream Track in YouTube Shorts that expresses a scenario they want to soundtrack. Then, from the list of approved artists (listed above), they can select the artist’s vocals that fit the aesthetics of the track best.

Before publishing the video to their subscribers, users will be allowed to preview the AI-generated track. It is important to note that, similarly to YouTube Shorts, there is a time restriction. AI-generated tracks’ maximum duration is 30 seconds in length. The feature has not been rolled out for full-length YouTube uploads or live broadcasts.

Some artists covered here are Warner Music artists. Uproxx is an independent subsidiary of Warner Music Group.