The volume of music reaching streaming platforms continues to grow every day, but that does not necessarily mean more artists, projects, or releases are connecting with real audiences.

‍

Deezer reported that fully AI-generated tracks made up more than 50% of all new music delivered to the platform during its busiest days in June 2026.

‍

The service is now receiving around 90,000 AI-generated tracks every day.

‍
Despite that volume, they account for just 1% to 3% of all streams on the platform.

‍

‍

More music, very little real listening

‍

The gap between the number of tracks being uploaded and the number of people actually listening suggests that much of this content is not being created in response to genuine audience demand.

‍

Deezer found that as much as 85% of the streams generated by fully AI-made tracks in 2025 were fraudulent.
Those plays were demonetized and removed from the royalty pool to prevent them from reducing the payments owed to legitimate artists.

‍

The platform already keeps fully AI-generated music out of its editorial playlists and algorithmic recommendations.
It has also begun labeling this content, giving listeners more clarity about when a track was created entirely through artificial intelligence.

‍

‍

Deezer is tightening its rules

‍

Under its new policy, Deezer will remove AI-generated tracks linked to streaming manipulation.

It will also take down tracks that have not received a single play for six months or longer.

‍
The goal is to prevent inactive content from continuing to accumulate and to protect the overall quality of the platform’s catalog.

‍

Deezer’s detection system analyzes certain characteristics within the audio to identify tracks that may have been created entirely with specific generative AI models.

‍

The company says the technology is 99.8% accurate. Still, as with any detection system, it may produce false results or fail to identify every case.

‍

‍

The issue goes beyond AI

‍

The rise in automated uploads raises a broader question: can streaming platforms continue to manage catalogs growing at this pace?

‍

The discussion is not only about whether a song was created using artificial intelligence.

‍
It is also about consent, ownership, proper rights management, the unauthorized use of an artist’s identity, and whether the streams behind that music are legitimate.

‍

AI can be part of a human creative process.
The problem begins when it is used to mass-produce content, flood distribution systems, or manipulate how royalties are shared.

‍

As millions of new songs continue to enter the market, transparency, accurate metadata, and catalog integrity will only become more important.