Deezer curbs fraud in AI-generated music and opens its detection technology to the entire industry

The streaming platform Deezer has demonetized up to 85% of plays of songs created entirely by artificial intelligence after detecting a high level of fraud. The company is now launching its detection tool for the entire music industry, aiming to protect artists' revenue and ensure greater transparency in the streaming ecosystem.

 The streaming platform  Deezer  has announced it will demonetize up to  85% of AI-generated music streams after discovering that most of these streams are fraudulent. The company, which launched its AI-generated music detection tool in early 2025, now plans to market it to the entire music industry, aiming to  promote transparency and protect artists’ revenue .

According to Deezer data, over  13.4 million AI-generated tracks were identified  on the platform in 2025. Currently, around  60,000 AI-generated songs are uploaded daily , representing approximately  39% of daily content . Although AI-generated music accounts for only 1% to 3% of total plays, up to  85% of these streams were considered fraudulent , compared to an average fraud rate of 8% across the rest of the catalog.

Deezer explains that  creating fake plays is the primary goal  of AI-powered mass music uploads, designed to manipulate royalty distribution systems. All fraudulent plays are  automatically excluded from revenue sharing .

Transparency and rights protection
“Content generated entirely by AI is almost indistinguishable from human music. Our focus remains clear: transparency for fans and protection of the rights of artists and composers,” said  Alexis Lanternier, CEO of Deezer .

The platform explicitly labels AI-generated music and excludes it from algorithmic recommendations and editorial playlists, so users can make informed choices and fraudulent actors cannot exploit the system. This policy aims to  protect creators’ revenue from massive amounts of synthetic content .

Detection technology now available to the industry.
Until now, Deezer was the only platform that publicly identified AI-generated music and removed it from its recommendation systems. With this announcement, the company  is licensing its technology for use by other companies and entities in the music industry .

According to Lanternier, the tool has already been tested with key players, such as the French collecting society  Sacem . The company states that expanding this technology can  reduce the economic incentives for fraud  and create a  common standard of transparency in music streaming .

Deezer’s technology can  identify music generated entirely by AI models like Suno or Udio , and is designed to adapt to new systems using reference data. Furthermore, more general methods are being developed that can  recognize synthetic content even without specific training . In December 2024, Deezer filed  two patent applications  based on the detection of unique signatures that distinguish synthetic content from authentic recordings.

A structural challenge for music:
The growth of AI-generated music coincides with concerns about the  use of copyrighted works to train models  and the potential erosion of copyright. Deezer is the only streaming platform that has signed the  global declaration on AI training and copyright .

According to a study by  CISAC and PMP Strategy  , with Deezer’s participation, nearly  25% of creators’ revenue could be at risk by 2028  due to AI, equivalent to approximately  €4 billion . This scenario underscores the importance of  combating fraud and protecting artists’ rights , positioning it as one of the major challenges facing the music industry in the coming years.

Scroll to Top