Editorial illustration for AI tracks surge on streaming services, says Deezer researcher Manuel Moussa
AI tracks surge on streaming services, says Deezer...
You can find a new song on a streaming service every 20 seconds. Roughly half the time now, it wasn’t written by a person. It was generated by a machine. Deezer researcher Manuel Moussa says his platform now sees 75,000 fully AI-made tracks uploaded daily.
That’s more than one per second. These aren’t tools for artists. They are wholesale replacements.
The volume has gone from worrying to absurd. In late 2025, Deezer reported 34 percent of its new uploads were synthetic, about 50,000 a day. That number now looks quaint.
The curve isn’t flattening. It’s a straight shot up. Human musicians are about to be outnumbered on the very platforms that promised them a living.
In September of 2025, Deezer said that 28 percent of music uploaded was fully AI-generated. By the end of the year, that had grown to over 50,000 tracks per day, accounting for 34 percent of uploads. Both users and artists have expressed frustration, demanding streaming platforms do something to combat the growing problem that is watering down playlists and siphoning millions in royalties away from legitimate artists. Things have only gotten worse at Deezer, where daily uploads of AI-generated content have grown to 75,000, and are threatening to overtake actual human-made music.
The math is brutal and simple. Each of those 75,000 daily files gets thrown into the royalty pool. Real artists watch their already microscopic per-stream payouts get diluted by a factor of thousands.
Listeners get lost in a swamp of generic, algorithmically pleasant noise. Platforms face a crisis they engineered. Their business was infinite shelf space.
They got infinite, functionally free content. Now they have to explain why anyone should care. Deezer’s numbers are the clearest proof yet that the floodgates are broken.
The water is rising. The industry built on music now has to decide if it actually wants any.
Common Questions Answered
How many AI-generated tracks does Deezer report being uploaded daily according to Manuel Moussa?
According to Deezer researcher Manuel Moussa, the platform now sees 75,000 fully AI-made tracks uploaded daily, which amounts to more than one per second. This represents a dramatic increase from late 2025 when Deezer reported 34 percent of its new uploads were synthetic, approximately 50,000 tracks daily.
What percentage of new uploads on Deezer were synthetic as of late 2025?
As of late 2025, Deezer reported that 34 percent of its new uploads were synthetic AI-generated tracks, totaling about 50,000 uploads daily. This significant proportion demonstrates how rapidly AI-generated music has infiltrated streaming platforms.
How does the influx of AI-generated tracks affect real artists' earnings on streaming services?
Real artists watch their already microscopic per-stream payouts get diluted by a factor of thousands as each of the 75,000 daily AI-generated files gets thrown into the royalty pool. This dilution effect means artists earn even less per stream as the pool of available content expands exponentially with machine-generated music.
What negative impact do AI-generated tracks have on listener experience according to the article?
Listeners get lost in a swamp of generic, algorithmically pleasant noise as AI-generated music floods streaming platforms. The overwhelming volume of synthetic content makes it increasingly difficult for users to discover meaningful music among the vast quantity of machine-generated tracks.
What business model contradiction does Deezer face with the rise of AI-generated music?
Deezer's business was built on providing infinite shelf space with infinite, functionally free content from AI generators, but now the platform faces a crisis it engineered. The company must now explain why listeners should care about their service when it is overwhelmed with generic, low-quality synthetic music rather than curated human-created content.
Further Reading
- Papers with Code - Latest NLP Research — Papers with Code
- Hugging Face Daily Papers — Hugging Face
- ArXiv CS.CL (Computation and Language) — ArXiv