Paris, January 30, 2026, 16:12 CET
- Deezer licensed its AI music detection technology to France’s royalty agency Sacem in a commercial deal.
- The streamer says it flagged more than 13.4 million AI tracks in 2025 and stripped fraudulent plays from payouts.
- Deezer is pushing wider adoption as AI-made uploads rise to about 39% of daily submissions.
Deezer said it has licensed its AI-generated music detection technology to France’s royalty agency Sacem, marking a commercial deal for a tool the streamer wants to see used more widely across the industry. Reuters
The move lands as generative AI makes it easier to churn out songs at scale, and as streaming platforms face a newer kind of fraud: mass uploads built to trip recommendation systems and skim royalties meant for artists and songwriters.
Deezer said it identified and removed up to 85% of fraudulent AI-generated music streams from its royalty pool in 2025, while flagging more than 13.4 million AI tracks. The royalty pool is the money set aside for rights holders; CEO Alexis Lanternier said it represents 70% of subscriber revenue.
Uploads have jumped fast. Deezer said it now receives about 60,000 fully AI-created tracks a day, roughly 39% of its daily uploads, up from 10% in January last year.
The company said its detection tool scans audio signals for patterns linked to AI music generators such as Suno and Udio, picking up anomalies that are hard to hear. Deezer said it trained the system on 94 million songs and filed two patents for the technology in 2024.
Lanternier said AI-made music has become “nearly indistinguishable” from human work, and said Deezer is focused on labelling AI tracks and protecting rights holders. He said the platform removes AI-generated music from algorithmic recommendations — automated suggestions that can drive plays at scale.
Deezer said it “demonetizes” fraudulent streams, meaning it excludes them from royalty payments, and is now making the detection technology available for others to license after testing it with Sacem. Newsroom Deezer
Deezer said fully AI-generated tracks still account for only a small share of listening on its service — up to 3% of total streams — but it found that fraud dominates within that slice. It said streaming fraud across its full catalogue accounted for 8% of all streams in 2025.
Sacem did not respond to a request for comment from Reuters. Deezer said it is in talks with other European collective societies about licensing and plans to engage with organisations in Los Angeles during Grammy Week.
Other platforms are also tightening rules around synthetic audio and impersonation, and weighing how to disclose AI use to listeners. Spotify has been rolling out policies and working on metadata standards, while Bandcamp has taken a harder line with a ban on AI-generated content, The Verge reported. Theverge
But detection tools may not settle the bigger fight over authorship and copyright. Swedish royalty society Stim told Reuters that detection alone cannot answer questions around musical composition and rights, adding that “copyright and technology can go hand in hand.” It has pushed for mandatory licensing and full transparency on training data to curb abuse at the source.
There is also a practical risk: as AI generators improve, detectors can end up in a cat-and-mouse game, and mistakes carry costs — false positives can hit legitimate releases, while misses leave money leaking out of royalty pools. Adoption is another unknown. A tool is only as useful as the number of major players willing to use it consistently.
