AI-Generated Music Now Tops Half of Deezer's Daily Uploads

Something crossed a line at Deezer this year, and it's worth sitting with for a second. More than half of everything uploaded to the streaming service each day is now made by AI. Not a niche corner of the catalog. Not a fringe experiment. The majority of new tracks landing on the platform daily.

Deezer has been counting these tracks since last year, and the trend has only pointed one direction — up, month after month, without pause. What started as a small slice of daily uploads has grown into the bigger half of the pile.

The Numbers Behind Deezer's AI Music Surge

The climb is easier to feel when you line the figures up. When Deezer first started publishing these stats, AI made up a modest fraction of what came in each day. Now it's the dominant share. Here's how that progression unfolded:

The volume hit its high point in June 2026, when AI-generated uploads averaged 90,000 tracks a day across the month. From 10,000 to 90,000 in roughly a year and a half — that's the shape of the thing. And now the share has pushed past the halfway mark.

What Deezer Is Doing About It

Reaching that threshold pushed Deezer to act rather than just watch the counter tick. The company plans to remove AI-generated tracks that fall into two buckets: those that haven't been streamed at all in the past six months, and those tangled up in fraudulent streams designed to inflate revenue. In other words, tracks nobody's listening to and tracks that exist mainly to game the payouts.

Deezer CEO Alexis Lanternier framed it as a continuation of work the company's been at for a while. He noted that Deezer has spent nearly two years fighting fraud and pushing back against the way AI music dilutes payments to real creators. With half of all daily uploads now AI-generated, he said the company is adding new steps to protect what artists and songwriters are owed — while keeping the focus on music that fans actually choose to listen to.

That last part matters. The concern here isn't only about who made a track. It's about money. When a flood of low-effort AI uploads siphons off streaming revenue through fake plays, the pool that pays working musicians shrinks. Deezer's cleanup targets that leak directly.

How Streaming Platforms Are Splitting on AI Music

There's no shared rulebook across the industry yet, and the differences are stark. Each platform is drawing its own line, and those lines are all over the map.

• Bandcamp took the hardest stance, banning AI-generated tracks from the platform outright.
• Tidal went after the incentive instead, cutting off monetization for AI music.
• Apple Music leans on a voluntary AI-tagging system, letting the labeling happen by choice rather than mandate.
• Spotify built its own policy centered on how much AI was actually used in making a given track.

Deezer's approach sits somewhere in the middle of all that — not an outright ban, but active removal of the tracks that are either ignored by listeners or clearly gaming the system. The variety here tells you the industry is still figuring out what "acceptable" even means when the machines can generate songs faster than humans can listen to them.

Deezer's Detection Tools and the Suno and Udio Question

None of this works without a way to spot AI tracks in the first place, and that's where Deezer has been building. The company started labeling AI music on its platform last year, flagging it for what it is. Its detection technology can specifically identify tracks made with models from Suno and Udio — two AI-music startups currently caught up in copyright lawsuits.

Deezer hasn't kept the tech to itself, either. Earlier this year it opened up its detection tool to other platforms, though it's still unclear whether any of the major players have picked it up. More recently, it released a separate tool that can scan Apple Music and Spotify playlists for AI-generated tracks — a way to see how far the AI tide has crept beyond Deezer's own walls.

Taken together, the labeling, the detection tech, and the new takedown rules point to a company trying to stay ahead of a problem that keeps doubling on itself. The uploads aren't slowing down. If anything, the June peak suggests the opposite. What Deezer's betting on is that better detection and steady cleanup can keep the platform tilted toward music people genuinely want — even as the raw volume of AI tracks keeps climbing.