Music Labels will do better with music that doesn’t use that much AI, Apple has warned, as it tries to keep the humanity in songs.

Like other areas of the creative world, artificial intelligence is becoming part of the music we listen to. However, Apple has issued a warning to the industry at large that it is working to combat the influx of AI audio content.

In an open letter to music labels and music industry partners seen by 9to5Mac titled “What we’re doing to keep music fair,” Apple Music offers a balanced take. There are benefits to AI, even in music, but it is working to make things fair for those who don’t use it.

Technology should amplify and not replace artists, and so far AI isn’t winning in the music scene. Apple claims that less than 1% of plays on Apple Music stem from AI music.

The low plays of AI music surfaced during an interview in April with Apple Music VP Oliver Schusser. He said that, while AI-generated music makes up a lot of submissions to the service, it’s not massively listened to by users.

More than a third of tracks were “100% AI,” he claimed at the time, but the listening level for AI-based songs was below 0.5%.

Labeling labels

It concedes that AI is a potential benefit to musicians finding new ways to tell stories and to innovate. But at the same time, Apple is putting in rules and limits to prevent its abuse.

Rather than outright banning AI music on Apple Music, Apple is instead aiming for a more transparent marketplace. One with clear labels of where AI is used.

The labeling of music is something that Apple asked of record labels and distributors in March. Transparency Tags are meant to indicate if select artwork, tracks, compositions, or music videos are AI-generated.

Distributors are adding the labels, Apple insists, and it will be required for future submissions. However, while the tags aren’t being actively used in the Music app itself, it does let Apple see what’s happening for the moment.

Bad AI

At the same time, Apple said it had developed internal tools for identifying AI tracks. The same tools will also be used to fight spam tracks and impersonation.

This includes monitoring streams of music. If an AI-generated track is getting its plays from manipulated streams, Apple Music can automatically pull that song from view.

Apple also updated the penalty for stream manipulation in February. In 2025, it excluded approximately 2 billion manipulated streams, making associated royalties be redistributed to its payout pool for artists and labels.

As it stands, stream manipulation accounts for about 0.5% of Apple Music streams. This is apparently one of the lowest rates across music streaming services.

Apple’s use of AI to police its storefronts isn’t a new thing for the company. It has already been using machine learning to help the App Store Review team check increasingly more apps over time.

Apple concludes the letter by reaffirming its pro-human stance on music. It still relies on human curation with its editorial team to discover and recommend the best music to users.