YouTube Music began rolling out labels for songs created primarily with generative artificial intelligence on October 14, putting the disclosure directly on the track’s player page.
In its official announcement, YouTube said listeners can tap the label for more information. The company said it is combining music-partner disclosures with its own detection tools. It plans to bring the labels to the main YouTube app but did not give a date.
The change gives artists, labels and distributors an immediate reason to check how their releases are classified. YouTube’s music-delivery guidance distinguishes fully generated content, partly generated content and content without generative AI. Partners can provide declarations through delivery metadata or, where applicable, YouTube Studio.
Those categories require care. YouTube’s examples put standard pitch correction and automated mastering in the no-generative-AI category. A song with entirely human-performed audio but AI-generated music-video visuals can fall under partly generated content. A partial designation therefore does not, by itself, establish that the recording’s performance was generated.
The announcement does not explain how each partner designation will appear to listeners. It also leaves detection accuracy, a complete rollout timetable and any music-specific effects on recommendations or royalty payments unresolved. The company did not announce a ban on AI-generated songs.
Radio News Now question: What process will artists and distributors have to challenge an incorrect label, and will the disclosure distinguish AI-generated audio from AI-generated visuals?