‘Made with AI’... YouTube to label AI music [Global AI Briefing]

[Financial News] YouTube is introducing a label to distinguish music made with artificial intelligence (AI). It will provide information reading “Made with AI” for music in which AI played a primary role in composition or vocal generation, allowing users to learn how the music was made.
As the use of AI has grown recently, major music platforms have increasingly begun to identify AI-generated tracks separately.
On the 14th (local time), YouTube announced in an official blog post that it would gradually roll out AI music labels on YouTube Music. The labels will apply when AI played an important role in making the music, such as when a generative AI model was used to create a song or generate vocals.
Not every use of AI tools in the music-making process will qualify for a label. The aim is to distinguish ordinary editing tools, such as standard pitch correction, from cases where generative AI played a primary role in composition or vocal production.

Users can tap the information displayed on the music playback screen to see an explanation that AI was involved in its creation. YouTube plans to extend the same labeling to the regular YouTube app in the future.
YouTube explained, “This measure is not about blocking all AI music, but focuses on providing information about the music-making process.”
YouTube said it would use its own detection technology to identify AI-generated music, even when its music distribution partners do not disclose that the music was made with AI.
Recently, music streaming platforms have increasingly begun distinguishing AI-generated tracks and attaching separate labels to them.
Spotify is beta-testing a feature that indicates in music credits how AI was used in vocals, lyrics, production, and other areas. It is also introducing an “AI Persona” label to identify fictional artist identities created with AI, and is pursuing a policy of keeping such artists out of standard recommendation lists.
The French-based music platform Deezer has announced that it will use its own detection technology to find and label AI-generated tracks, and exclude them from editorially curated playlists and algorithmic recommendations.
[email protected] Lee Gu-sun Reporter