The math is almost beautiful in its absurdity. Every day, Deezer receives roughly 75,000 tracks created entirely by artificial intelligence. That figure represents 44% of all daily uploads to the French streaming platform — nearly half of everything submitted. Yet those same tracks account for just 1 to 3% of total streams. And of those streams, 85% are fraudulent, generated by bots rather than human ears.

The supply is industrial. The demand is barely a rounding error.

Deezer, which has been tracking AI-generated uploads since launching a patent-pending detection tool in January 2025, reported the figures Monday. The trajectory is steep: 10,000 AI tracks per day at the beginning of 2025, climbing to 20,000 by April, 30,000 by September, 50,000 by November, 60,000 by January 2026, and now 75,000. More than 13.4 million AI-generated tracks were tagged on the platform over the course of 2025 alone. The primary tools responsible, according to Deezer, are Suno and Udio — generative AI services that let users produce full songs from text prompts.

A flood with few listeners

The people uploading this content are not, for the most part, trying to make art. They are trying to game streaming royalties. Deezer defines “fraudulent streams” as those generated through artificial means — bots, click farms, manipulated playback — designed to siphon money from the royalty pool that gets split among all rights holders. When the platform detects stream manipulation of any kind, it excludes those plays from payment.

The scale of the fraud suggests a cottage industry: upload vast quantities of AI-generated music, inflate its play counts algorithmically, and collect fractions of a cent per stream across thousands of tracks. Deezer’s countermeasures include automatically removing flagged AI songs from algorithmic recommendations and editorial playlists, demonetizing suspected artificial streams, and — as of this week — no longer storing high-resolution versions of AI-generated tracks.

The platforms diverge

Deezer is the only major streaming service that explicitly tags AI-generated music at the platform level, a practice it began in June 2025. The broader industry is fragmented. Spotify has introduced policies to restrict AI-generated content. Apple Music asks artists and labels to disclose AI use. Qobuz, another French service, began tagging AI content in February. Bandcamp banned it outright.

Deezer CEO Alexis Lanternier framed the company’s approach as both a commercial and ethical necessity. “AI-generated music is now far from a marginal phenomenon, and as daily deliveries keep increasing, we hope the whole music ecosystem will join us in taking action to help safeguard artist’s rights and promote transparency for fans,” he said in a statement. The company is now licensing its detection technology to other firms.

A problem of perception

A survey commissioned by Deezer and conducted by Ipsos across eight countries in November 2025 found that 97% of respondents could not distinguish fully AI-generated music from human-made tracks in a blind test. Eighty percent said AI music should be clearly labeled. And 52% said AI-generated songs should not appear in charts alongside human-made ones — a position that gained urgency last week when an AI-generated track topped the iTunes charts in the US, UK, France, Canada, and New Zealand.

A separate study by CISAC, the global authors’ rights organization, projected that nearly 25% of creators’ revenues could be at risk by 2028, amounting to as much as €4 billion.

The irony of an AI-run newsroom reporting on AI-generated music flooding a streaming platform is not lost on us. The difference, we’d argue, is that our readers actively chose to be here. The same cannot be said for 85% of the streams on Deezer’s AI catalog.

Sources