YouTube Music has started rolling out labels for tracks it says were created primarily with generative AI. The new notice appears directly on a track's player page, where listeners can tap an "AI" button for more information about how the song was made. Music Business Worldwide and Music Ally both reported that the labels went live on October 14.
The company's own example is a song made with a generative model handling both the composition and vocals. That matters because the rollout is aimed at recorded music, rather than the video disclosures YouTube has already used elsewhere on its service. Music Business Worldwide reports that YouTube plans to extend the music labels to the main YouTube app as well. In a separate May update on video disclosures, the company said it could automatically apply labels when its systems detected significant photorealistic AI use, including where creators had not made a disclosure.
The system draws on more than one source of information. YouTube says it is working with music partners while also using its own detection tools. According to Music Business Worldwide, partners delivering content through DDEX or CSV templates can make a declaration when they supply a track.
Those declarations have three categories: Fully Gen AI, Partly Gen AI and No Gen AI. A prompt-made track downloaded from an AI music generator is listed as fully generated. An artist using AI for a bassline or string part before adding live vocals and instrumentation is a partly generated example. Pitch correction and automated AI mastering sit in the no-AI category under the guidance cited by Music Business Worldwide.
A partner declaration is not the only route to a label. The platform says it may use additional signals alongside those disclosures, and may rely only on those signals where no generative-AI information has been supplied. That gives YouTube Music a role in deciding how material is classified, not simply displaying the information sent by labels and distributors.
Important details remain unaddressed. Neither the blog post nor the partner guidance reviewed by Music Business Worldwide says whether labelled music will be treated differently in recommendations or royalty payments. The reporting also does not establish whether a partly generated track will receive the same player-page label. Those gaps matter because the public-facing button is only the visible end of a broader classification system.
YouTube Music is joining a fast-moving set of platform responses, but not through one shared format. Music Ally notes that Apple Music, Spotify and Deezer have also set out their own approaches rather than using the two voluntary labels proposed in July by music-industry bodies. Music Business Worldwide reports that Apple Music plans "Made With AI" labels, while Spotify has tested disclosures in song credits and announced an AI Persona badge for artist profiles with AI-generated identities.
For now, the change gives listeners a new signal without resolving what that signal will mean across streaming. YouTube Music has described its goal as giving people clearer information about the music they hear. The rollout confirms that AI disclosure is becoming a product feature on major services; the treatment of labelled tracks, and the consistency of the labels between platforms, remain unsettled.
Read more on EDM Global News.