44 Percent of New Music Is Now AI-Generated and Human Artists Are Paying for It
Deezer reports 75,000 AI-generated tracks uploaded daily. Most streams are fraudulent. The royalty pool that pays real artists is being drained by bots.
C-Tribe Society
Seventy-five thousand AI-generated tracks are uploaded to Deezer every single day.[1] That is not a typo.
It is 44 percent of all new music delivered to the platform.[1] Deezer published the numbers on April 20, and they land like a brick through the window of every conversation about AI and creativity. In January 2024, the platform was receiving 10,000 synthetic tracks per day. By September 2024, it was 30,000. By January 2025, 60,000. Now 75,000. More than two million machine-generated songs per month,[2] flooding the same royalty pool that pays human artists. The scale is no longer theoretical. It is operational, and it is accelerating.
Image via TechCrunch
Almost Nobody Is Listening to This Music
AI-generated tracks account for only 1 to 3 percent of total streams on Deezer,[3] and 85 percent of those streams are flagged as fraudulent and demonetized.[4] The music is not being made for listeners. It is being made for algorithms.
The business model is straightforward and parasitic. Generate thousands of tracks using AI tools. Upload them to streaming platforms. Use automated play farms to generate streams. Collect royalty payments from a pool that is funded by the subscription fees of real listeners who came to hear real artists. Deezer has tagged 13.4 million AI tracks since June 2025,[5] but the volume keeps climbing because the economics still work for the people running the fraud.
Streaming royalties are paid from a shared pool. Every fraudulent stream that pays out to a bot operator is money subtracted from an independent artist's quarterly check. CISAC, the international confederation of authors' societies, projects that AI-generated content puts 25 percent of music creators' revenues at risk by 2028, potentially costing artists up to 4 billion euros.[6]
Deezer CEO Alexis Lanternier was direct about the stakes: "AI-generated music is now far from a marginal phenomenon...we hope the whole music ecosystem will join us in taking action to help safeguard artists' rights and promote transparency for fans."[7]
Listeners Cannot Tell the Difference
Deezer commissioned research on that question, and the answer is uncomfortable: 97 percent of listeners cannot distinguish AI-generated music from human-made music.[8] At the same time, 80 percent of those listeners said they support mandatory labeling of AI content.[9]
That gap between detection ability and desire for transparency is the crux of the problem. People cannot tell, but they want to know. The music sounds real enough. Whether "real enough" should be the standard the industry accepts is another question entirely.
Image via Music Business Worldwide
Platforms Are Responding With Buckets to a Flood
Deezer has been the most aggressive, building detection systems that tag and demonetize AI tracks.[10] Qobuz launched its own detection tool in February. Apple Music introduced "Transparency Tags" in March that rely on label declarations.[11] Spotify launched DDEX standard support with a "Song Credits" beta that surfaces production metadata.[12]
Sony Music has taken a harder line, requesting the removal of over 135,000 AI-generated songs that impersonated its artists.[13] Universal and Warner have pursued similar takedowns, though the whack-a-mole dynamics are obvious: removing 135,000 tracks matters less when 75,000 new ones arrive tomorrow.
Detection tools help. Labeling helps. But none of it addresses the fundamental incentive: generating and uploading synthetic music to streaming platforms costs almost nothing, and even a small fraction of fraudulent streams that slip through detection generates revenue.
Berklee Students Are Not Buying It
Over 400 students at Berklee College of Music signed a petition in April demanding the school discontinue generative AI tools and cancel an AI songwriting elective.[14] Students paying six figures for a music education are watching AI produce technically competent tracks in seconds and asking a reasonable question: what exactly are they training for?
The Berklee petition sits alongside a broader creative backlash. Over 1,000 musicians, including Paul McCartney, contributed to Is This What We Want?, a protest album released this spring that highlights the threat of AI-generated music to human artistry.[15] The title is not rhetorical. It is a genuine question about whether the music industry's infrastructure, built around streaming economics and platform scale, can protect the people who actually make the music.
The Streaming Economy Was Already Broken
The uncomfortable truth underneath the AI flood is that the streaming royalty model was never designed to support working musicians at scale. The average per-stream payout on major platforms hovers between $0.003 and $0.005.[16] An independent artist needs roughly 300,000 streams per month to earn minimum wage.[17]
Adding two million synthetic tracks per month to the same ecosystem does not break a healthy system. It accelerates the collapse of one that was already failing.
AI-generated music is a symptom, not the disease. The disease is a royalty structure that treats all streams as equal, pays from a shared pool, and has no mechanism to distinguish between a song someone chose to hear and a song a bot played 10,000 times on a phone in a server closet. Forty-four percent of new music is synthetic.[18] Eighty-five percent of its streams are fraud.[19] And the other 56 percent of music—the kind made by humans who spent years learning instruments and writing lyrics and touring in vans—is subsidizing the whole operation through a shared royalty pool that does not care who is real and who is not.
References
Deezer Newsroom, "AI-generated tracks now represent 44% of all new uploaded music", 2026. Link
Music Business Worldwide, "75,000 AI-generated tracks now flood Deezer daily, representing 44% of all new music uploaded to the platform", 2026. Link
Ars Technica, "Deezer says 44% of new music uploads are AI-generated, most streams are fraudulent", 2026. Link
Dataconomy, "Deezer Reports AI Music Reaches 44% Of New Uploads", 2026. Link
The Decoder, "The flood of AI music is reshaping how streaming platforms handle new uploads", 2026. Link
Forbes, "AI's Impact On Music In 2025: Licensing, Creativity And Industry Survival", 2024. Link
Digital Music News, "Streaming Royalty Rates and Artist Earnings Analysis", 2025. Link
