YouTube Music begins labelling AI-generated tracks with its own system

YouTube Music has started labelling songs made primarily with generative AI, with the system live since 14 October 2026. According to Music Ally, which reported the launch on 15 October, the label appears as an “AI” button that gives listeners information on how a track was made.

The platform describes the covered material as songs “created primarily using generative AI”, as quoted by Music Ally from YouTube Music’s blog post. Music Ally places the launch among a series of AI-labelling announcements from the biggest streaming services.

How tracks are flagged

Music Ally reports that the blog post names two elements: collaboration with music partners, and YouTube’s own detection tools, both used to identify AI-generated tracks. The post also says the labels apply the same labelling principles that YouTube already uses elsewhere on its platform.

Music Ally’s report does not say how the detection tools work, how many tracks have been labelled so far, or what detail sits behind the button. It also does not set out whether the rollout differs between countries.

Not the industry’s framework

Music Ally points out that YouTube, like Apple Music, Spotify and Deezer, is not using the labelling system that a group of music-industry bodies proposed in July 2026. The publication’s view is that it is hard to see a way forward for the industry’s preferred system. It adds that the bigger win may be that streaming services are committing to labelling at all, which it says is good for listeners and the industry alike.

What it means for listeners

YouTube Music is available to UK listeners, so the labelling is relevant to British users and artists, although this is a global platform policy rather than a UK-specific measure.

In practical terms, the label is YouTube’s own design rather than the industry’s proposed framework, and it is framed around songs made primarily with generative AI. The report leaves open how tracks that use AI only in part are handled.