Apple Music to Label AI-Made Tracks Later This Year

Apple Music will begin labeling songs on the platform derived from AI platforms, adding a “Made With AI” label to such songs later this year.

In a note to its industry partners and one viewed by Variety, the service said that record labels and distributors will be required to tag songs in which “a material portion of the content” was created using AI. The labels follow Apple Music’s introduction in March of AI Transparency Tagging, which dictates in the service’s systems whether a track used AI.

“As we implement Made With AI Labels, content providers will be important partners,” Apple Music said. “They are best positioned to know how their content was created, and the AI Transparency Tags that should be used.”

It was unclear when the new labels would begin to appear on the streamer. Record labels and the streaming services they feed have spent the last three years navigating the rise of AI music generators such as Suno and Udio, opting between litigation to licensing deals to developing competing tools, such as remixing products.

The announcement follows Spotify’s decision to label the profiles of artists who appear to be AI-generated as “AI Personas” and Tidal’s decision to label AI music and ban those tracks from accruing royalties. A collection of groups representing musicians and the music industry at large — including the RIAA, IFPI, the Grammys and SAG-AFTRA — also banded together last month to advocate for a new labeling program to indicate whether a song has been produced with AI.

Apple Music vice president Oliver Schusser told Billboard earlier this year that “more than a third” of the tracks uploaded to the platform were “100% AI.” He added that, while the company had in-house detection software to screen songs and determine how AI was used and what models created them, it was up to those who upload tracks to “take responsibility“ for disclosing the use of AI.

“You should know that every label in the world is delivering AI, [even if] they might not know about it,” he said. “They themselves also need to develop tools to understand [the scope of AI use].”