AI Music Surges Past Half of Deezer’s Daily Uploads as Streaming Fraud Concerns Grow

Artificial intelligence is rapidly reshaping the streaming landscape, with Deezer revealing that more than half of all new tracks uploaded to its platform are now entirely AI-generated.

The Paris-based streaming service reported that nearly 90,000 AI-created songs were uploaded each day throughout June, marking a sharp increase from the roughly 75,000 daily uploads it disclosed just two months earlier. On some days, synthetic recordings represented over 50% of all music arriving on the platform.

In response, Deezer is tightening its policies around AI content. The company says it will remove generative AI tracks used to manipulate streaming numbers, as well as AI uploads that remain unplayed for six months, in an effort to reduce fraud and protect royalty pools for human creators.

While fully AI-generated songs now dominate new uploads, Deezer says they still account for only a small share of listening activity. However, the platform estimates that the vast majority of streams generated by those tracks in 2025 were tied to fraudulent behavior, prompting the service to exclude manipulated plays from royalty calculations.

Chief Executive Alexis Lanternier described the current moment as a turning point for the industry, arguing that streaming platforms, rights holders and technology companies must work together to establish common standards for AI-generated music while ensuring artists and songwriters remain protected.

Deezer has positioned itself at the forefront of AI detection, introducing proprietary technology capable of identifying music created with popular generative platforms such as Suno and Udio. Since last year, the service has also labeled fully AI-generated recordings and kept them out of editorial playlists and recommendation algorithms. The company says it has now identified millions of AI-created tracks and plans to license its detection technology to other businesses across the music industry.

Streaming platforms including TIDAL, Apple Music and Qobuz have introduced their own AI-labeling initiatives, while organizations such as the RIAA and IFPI continue to advocate for standardized tags distinguishing fully AI-generated music from recordings that simply incorporate AI tools.