Deezer launched its AI music detection tool last year as part of efforts to “prevent fraudulent actors from stealing royalties from real artists through mass produced AI-generated music.” The company says it has used the tool to identify and tag more than 13.4 million AI songs in 2025, even as the flood of AI-generated tracks continues to grow. Deezer claims its tool can detect AI songs with a 99.8 percent accuracy rate.
In the press release, Deezer says it receives more than 60,000 AI tracks uploaded every day, making up 39 percent of total uploads. That’s double the 30,000 daily AI track uploads that Deezer reported receiving in September 2025. Deezer also found that up to 85 percent of the AI-generated music streams it identified in 2025 are “fraudulent,” compared to 8 percent of all streams in 2025.
“We know that the majority of AI-music is uploaded to Deezer with the purpose of committing fraud, and we continue to take action,” Deezer CEO Alexis Lanternier says in the press release. “Every fraudulent stream that we detect is demonetized so that the royalties of human artists, songwriters and other rights owners are not affected.”
As songs generated with platforms like Suno and Udio become harder to identify, music streaming services are taking steps to ensure listeners know when the music they’re listening to wasn’t created by a human. Spotify began rolling out new policies to address AI music and impersonation last year, and it’s also working on a new metadata standard for disclosing the use of AI. Meanwhile, Bandcamp is taking a tougher stance and is banning AI-generated content completely.