Deezer announced on Thursday that it will offer its AI detection tool, launched last year, to other streaming platforms to combat AI-generated music, fraudulent streams, and promote industry transparency.
The tool automatically identifies and tags fully AI-generated music tracks, preventing their appearance in algorithmic and editorial recommendations on Deezer’s platform. This measure ensures listeners encounter content created by human artists rather than machine-generated equivalents. Deezer processes uploads continuously, applying the detection process to incoming tracks before they enter the catalog available to users worldwide.
Deezer provided data underscoring the scale of the issue. Among streams from fully AI-generated tracks, 85 percent qualify as fraudulent. Currently, the platform receives 60,000 such tracks each day, accumulating a total of 13.4 million AI-detected songs in its database. This volume reflects the rapid proliferation of AI-generated content across music services.
For context, in June of the previous year, fully AI-generated music constituted 18 percent of Deezer’s daily uploads. That figure exceeded 20,000 tracks per day, marking a substantial portion of new content at the time and highlighting early growth in AI music production.
The detection tool identifies every AI-generated track produced by major generative models, including Suno and Udio. Beyond tagging and excluding tracks from recommendations, Deezer demonetizes them entirely. This action removes AI tracks from the royalty pool, directing payments solely to human musicians and songwriters based on verified streams.
A company spokesperson stated to TechCrunch that the tool achieves 99.8 percent accuracy in distinguishing AI-generated music from human-created works. Deezer CEO Alexis Lanternier noted there has been “great interest” in the tool, adding that several companies have already performed successful tests.
One tester is Sacem, the French management company representing over 300,000 music creators and publishers. Sacem manages rights for prominent artists such as David Guetta and DJ Snake, processing licensing and royalties on their behalf across various uses. Deezer confirmed Sacem’s successful evaluation but did not disclose other interested parties.
Pricing for access to the tool varies depending on the type of deal negotiated with prospective clients. Deezer provided no specific figures, indicating customized agreements based on platform scale, volume, or integration requirements.
Industry concerns encompass AI companies training models on copyrighted material without permission. Additional worries involve manipulation of streaming systems through artificial inflation of play counts. These practices divert revenue from legitimate artists.
In 2024, the U.S. Department of Justice charged a musician from North Carolina with fraud. Authorities accused the individual of producing AI-generated songs and deploying bots to stream them billions of times. This scheme resulted in more than $10 million in stolen streaming royalties.
AI-created groups like The Velvet Sundown have accumulated millions of streams on platforms, blending into regular catalogs without disclosure. Such instances contribute to distorted charts and payout distributions.
Bandcamp implemented a total ban on AI-generated music uploads. Spotify revised its policies to specify permitted AI involvement in production, measures to curb spam, and an explicit prohibition on unauthorized voice clones mimicking artists without consent.
Major record labels reached settlements in lawsuits against Suno and Udio. Last fall, Universal Music Group and Warner Music Group signed licensing agreements with these AI startups. The deals permit use of their music catalogs for model training, with compensation flowing to artists and songwriters whose works contribute to the datasets.
In 2024, Deezer became the first music streaming platform to sign the global statement on AI training. Signatories include actors Kate McKinnon, Kevin Bacon, Kit Harington, Rosie O’Donnell, and other creatives advocating for protections in AI development.
