Deezer Will Let You Erase AI Music From Search. The Industry Is Finally Picking Sides.

Streaming services spent the last two years learning how to label AI-generated music. Deezer just decided labeling is not enough. On October 13, the French streaming platform announced a user-facing toggle that removes fully AI-generated tracks, albums, and their related artists from search results entirely, to roll out globally before the end of 2026.

The announcement makes Deezer the most combative major platform in the fight over synthetic music. It already excludes AI-generated tracks from algorithmic recommendations and editorial playlists. Now it is handing listeners a master switch: flip it, and the synthetic catalog disappears from your search.

The scale of the flood

The numbers behind the move are difficult to overstate. Deezer says it is receiving nearly 90,000 fully AI-generated tracks every day. On peak days, more than half of all new music delivered to the platform is machine-made. That compares with just 10,000 AI tracks a day when the company launched its detection tool in January 2025, a ninefold increase in under two years.

Here is the part that reframes the whole debate: those tracks generate only 1 to 3 percent of streams. Half the uploads, a sliver of the listening. Deezer says the conclusion is obvious, that the primary purpose of uploading these tracks is fraudulent. Its internal figures show up to 85 percent of streams on fully AI-generated tracks in 2025 were manipulated, excluded from royalty payments by the company.

Between January 2025 and August 2026, Deezer's detection systems flagged more than 42 million AI-generated tracks, and since it began explicitly labeling synthetic music in June 2025, over 14 million albums have been tagged.

"The explosion of AI-generated content continues to impact the whole music ecosystem, and we believe in shaping a future based on transparency and choice, with respect for artists, song writers and music fans," said Alexis Lanternier, CEO of Deezer.

Why a toggle, and why now

Deezer points to listener demand. A November 2025 survey it commissioned from Ipsos, covering 9,000 people across eight countries, found that 45 percent of streaming users would like to filter out 100 percent AI-generated music. The same study found that 97 percent of respondents could not tell the difference between AI-generated and human-made music in a blind test, which is exactly why the company argues the choice should be explicit rather than algorithmic.

The timing is not accidental. On October 7, the US Copyright Office opened a formal inquiry into music streaming fraud, naming dilution of the royalty pool as the primary harm and giving commenters until November 23 to weigh in. The inquiry landed one day after Michael Smith, the North Carolina man who used bot accounts and AI-generated songs to siphon off more than $8 million in fraudulent royalties, was sentenced to 18 months in prison. Deezer's data, up to 85 percent of AI-track streams being fraudulent, is the evidentiary backbone of that policy conversation.

Detection as the actual moat

The toggle is only as good as the detection behind it, and Deezer has spent the last two years turning detection into a business. Its proprietary, patent-pending tool, with two patents published by the EU and US patent offices in June 2026, identifies output from the dominant generative platforms: Suno, Udio, and Riffusion, the predecessor to Google-owned ProducerAI. In January 2026, the company began licensing the technology to the wider industry, with French collecting society Sacem as its first partner. In June, it released a free public scanner that lets anyone check playlists on 20 competing services for synthetic tracks.

The industry split is now visible. TIDAL announced in June it would tag fully AI-generated music and stop paying royalties on it. Qobuz launched its own detection system in February. Apple Music's tagging system, launched in March, relies on labels and distributors to declare AI use, with "Made With AI" labels appearing later this year. Spotify began testing AI tags in song credits in April but only where the artist discloses use, and in August said it would label AI-generated artist profiles as "AI Personas" from mid-September. Deezer is the only platform that has moved from tagging to erasing.

What this means for listeners

There is an important boundary in the new toggle: it targets fully AI-generated music, not AI-assisted music. Tracks where AI plays a limited role inside a broader human creative process are treated differently, which matters as AI-assisted production becomes standard practice across every genre. The toggle is a filter for synthetic catalogs, not a purity test for modern recording.

The deeper question is economic. Research Deezer participated in, conducted by CISAC and PMP Strategy, projects that 24 percent of music creators' revenues are at risk by 2028, a potential loss of 4 billion euros. Deezer described that as a colossal, even critical, challenge for the music creation sector as a whole. A toggle does not fix royalty dilution on its own, but it tests a hypothesis the whole industry is watching: whether listeners, given an explicit choice, will pay to keep the human in music. The answer will shape every platform's policy for the next decade.