YouTube Music launches new labels as AI-generated music floods streaming platforms

YouTube Music began rolling out a labeling system on Wednesday that marks songs created with generative AI, so listeners can see when artificial intelligence played a role in a track's composition or vocals. The label appears on the player page, and tapping or clicking it opens more detail about how AI was used. The same labels also appear in the main YouTube app, extending the disclosure beyond the music service itself.
To decide which tracks get flagged, YouTube said it worked with a range of music partners and also applied its own detection tools. Partners that distribute music through YouTube can declare how a work used generative AI by choosing one of three designations: Fully Gen AI, Partly Gen AI or No Gen AI.
The categories work as follows. Fully Gen AI covers tracks that were entirely generated by AI. Partly Gen AI covers works where artists mixed in AI-generated components but also added their own live elements. A track that was only edited with pitch-correction tools such as Auto-Tune falls under No Gen AI. That last distinction is notable for producers: corrective vocal tuning, a decades-old studio staple, is not being treated as generative AI use even though it is machine-assisted.
YouTube is framing the program around disclosure rather than restriction. AI involvement by itself does not make a track ineligible for monetization on the platform. YouTube has also been building on the creation side, including Dream Track, a tool that lets select creators generate soundtracks using AI-generated versions of participating artists' voices.
The launch lands in a streaming market where rivals are moving in different directions. Apple Music plans to add "Made With AI" labels as part of AI Transparency Tags later this year. Spotify already has AI labels and a music spam filter. Deezer has taken a stricter line: in July it reported receiving nearly 90,000 fully AI-generated tracks per day, and it labels detected AI music, keeps it out of algorithmic recommendations and editorial playlists, and withholds royalties from fraudulent streams. Deezer is also working on a filter that would let users exclude AI-generated music. TIDAL labels recordings it identifies as entirely AI-generated, lets listeners disable them, and excludes them from royalty attribution.
The contrast is the core of the story. Deezer and TIDAL are treating detected AI music as something to suppress, demote or strip of royalties. YouTube is treating it as something to disclose while leaving distribution and monetization largely intact. This suggests YouTube is positioning itself as a neutral pipe: it keeps its catalog broad and lets listeners and rights holders decide what to do with the information, rather than acting as a gatekeeper on artistic provenance. A likely trade-off is that disclosure only works as well as the declarations behind it. Because partners self-report through the three designations, the accuracy of the labels depends on honest filing, supplemented by YouTube's own detection. Uploaders who fail to declare, or who misjudge whether a tool counts as generative, could produce gaps or mismatches that a stricter detection-first regime would not have.
How listeners see the labels is another variable. Placing them on the player page and in the main YouTube app makes them visible at the moment of playback, which is more prominent than burying provenance in metadata or liner notes. Whether that changes listening behavior, playlist curation or artist promotion is an open question. For now, YouTube is not saying it will exclude labeled tracks from recommendations, playlists or payouts the way Deezer and TIDAL do.
The declarations are also a signal to rights holders and distributors, who now have a formal field to fill in. Getting the designation right matters for transparency, and presumably for any future policy that builds on this data. The three-tier scheme gives the industry a common vocabulary, but the boundaries, such as where heavy AI-assisted production ends and Partly Gen AI begins, will be tested in practice. Auto-Tune's placement in No Gen AI is one early boundary marker.
Why it matters: artists, labels and distributors that release through YouTube now have a formal way to declare AI use, and a new step in their delivery workflow, even though AI tracks remain monetizable. Listeners gain visibility into whether a song was AI-generated, but the labels depend on partner self-reporting plus YouTube's detection rather than an outright ban. For product teams, this is a disclosure-first model, distinct from the demotion and royalty-exclusion approaches taken by Deezer and TIDAL, and it may set expectations for how other platforms handle AI provenance as Apple's tags arrive later this year.
Based on reporting from the original publisher. Visit the source for full context and later updates.
Publisher excerpt
The label will appear on the player page and include songs entirely generated by AI or partially, such as AI involved in composition and vocals.