Deezer Reveals 28% of Songs Uploaded Are Now AI-Generated

Share this article
Share this article
Prioritise Us on Google
Some in the music industry are becoming especially concerned by the amount of AI-generated music flooding streaming platforms every day
The French music streaming platform has reported that around 30,000 AI-generated tracks go live on its site each day, but what does that mean for music?

Deezer, one of the world's premier music streaming platforms, has made a shocking admission. According to the French streaming giant, 28% of all songs uploaded to its library are now generated wholly by AI.

That is equivalent to around 30,000 AI-generated tracks each and every day.

The French firm reported in January that 10% of its uploads were AI-generated, but that number has only continued to swell in recent months as AI models that specialise in music generation have become more sophisticated and prolific.

While some observers may view this as an irritating but harmless development, there are those in the music industry that are deeply concerned about what this means for the future of the sector.

Deezer is one of the world's most widely used music streaming platforms | Credit: Deezer

'Slop of the Pops'

Deezer's AI detection technology, implemented at the beginning of 2025, has tracked a relentless climb in synthetic music submissions.

The platform now receives more than 30,000 fully AI-generated tracks daily, tripling from approximately 10,000 submissions it recorded each day in February. The growth shows no signs of slowing, either, with the April figure of 18% jumping to 28% by September.

It's something that Deezer's leadership team are looking to address head on, in a way that sets an example for other streaming sites like Spotify, Apple Music and YouTube, all of which have been experiencing similar things in recent months.

"Following a massive increase during the year, AI music now makes up a significant part of the daily track delivery to music streaming and we want to lead the way in minimising any negative impact for artists and fans alike," says Alexis Lanternier, Deezer's CEO.

The streaming service has positioned itself as the industry's sole guardian against synthetic content infiltration, becoming the first platform to explicitly label AI-generated music in June.

Alexis Lanternier, CEO of Deezer | Credit: Deezer

A pattern of fraud

Whether the issue of AI-generated music is regarded as serious or not, there is a troubling reality behind these statistics: the fraudulent manipulation of streaming royalties.

According to Deezer, fraudsters are manipulating the listening figures of the AI-generated music they upload, with a view to receiving pay-per-play royalties.

The French tech firm says that around 70% of the plays that AI-generated content receives can be identified as stemming from fraudulent activity.

This affects not only the streaming services that are being duped by AI, but the real-life musicians whose work is elbowed out of the way by artificially-generated songs.

Deezer plans to respond by removing fully AI-generated content from algorithmic recommendations and excluding it from editorial playlists.

Youtube Placeholder

"Our approach is simple: we remove fully AI-generated content from algorithmic recommendations and we don't include it in editorial playlists," Alexis explains. "This way, we ensure the impact on the royalty pool remains minimal while providing a transparent user experience."

While Deezer remains steadfast in its stance towards these things, it has been suggested that Spotify, the world's most popular music streaming service, is taking a more insidious approach.

In her bestselling book 'Mood Machine', Liz Pelly alleges that the Swedish firm has been purposely flooding its site with AI-generated music for a long time now.

In the book, Liz accuses Spotify of pushing these AI-generated songs onto some of its high traffic playlists at the expense of real artists. The reason? To avoid paying them royalties.

Liz Pelly, author of Mood Machine | Credit: Liz Pelly

The implications

This trend poses some pressing existential questions for musicians, record labels and for the economics of streaming altogether.

With traditional artists facing potential displacement as synthetic alternatives flood recommendation algorithms and compete for listener attention, it is incumbent on the sector's major players to protect the integrity of their product.

Streaming platforms confront the challenge of maintaining content quality whilst managing unprecedented upload volumes.

The revelation follows broader industry incidents, including AI-generated songs appearing on deceased musicians' profiles without permission.

One instance involved country artist Blaze Foley, whose Spotify page recently started to be populated by synthetic content, despite the fact that the singer had died in 1989.

Deceased country singer Blaze Foley's Spotify page recently became overrun with AI-generated music | Credit: Blaze Foley

How technology can help solve this problem

Deezer's detection capabilities represent just the beginning of a technological arms race between synthetic content creators and platform gatekeepers.

The company aims to expand its AI-detection technology to identify deepfake voices, suggesting even more sophisticated synthetic content lies ahead.

Other platforms have yet to implement similar detection or labelling systems, potentially making Deezer an outlier when it comes to transparency and governance.

For some campaigners, Deezer is doing just the right thing. "Streaming services should take measures to monitor fraud and report on fraudulent usage to their rightsholder partners," says Elizabeth Moody, Partner & Chair of the New Media Practice at Granderson Des Rochers.

"AI companies can fingerprint and/or watermark content created on their platforms for distribution in order for the DSPs to determine the source of the AI content.

"Record labels and music publishers can help by creating industry standard solutions for tracking and identifying ownership and royalty splits across industry and AI company stakeholders, to ensure artists are properly attributed."

Elizabeth Moody, Partner & Chair of the New Media Practice at Granderson Des Rochers | Credit: Liz Moody

Is AI-generated music necessarily a bad thing?

While the idea of AI-generated music need not necessarily be regarded with caution and disdain, it is prudent to consider the effects the practice might have on the music industry going forwards, especially since the proportion of synthetic content will likely continue to rise among streaming platforms as AI technologies become more accessible and sophisticated.

I encourage music industry creatives and leaders to be cautiously optimistic about the changing ecosystem," says Elizabeth.

"I don’t believe that AI will replace creativity; if used appropriately with proper controls, it will enhance creativity. They can be used to create deeper, more personalised experiences and increase fan engagement."

This summer, a band known as The Velvet Sundown were listened to by millions of people, before it was revealed that the project was entirely created by AI, from their image to their music itself.

The Velvet Sundown are an AI-generated band, which, until recently, listeners believed were real | Credit: The Velvet Sundown

While the music was embraced by a great many people, experts in digital content licensing like Elizabeth are calling for regulation and action to prevent the worst consequences of AI's growing influence over music creation and listening habits.

"Of course, when used incorrectly, these tools can present some threats to existing artists and to the industry," she says.

"Fraud has been an issue in the music industry for quite some time. Fraudsters could use AI music tools to generate music more quickly and then attempt to flood the DSPs with AI music.

"With proper controls and enforcement, fraud can be suppressed and we can enjoy the benefits of AI music."

Company portals