Deezer has launched a free tool that scans Spotify and Apple Music playlists for AI-generated tracks, which now make up nearly half of daily uploads. Music is a preview of what AI does to every content channel.
Nearly half of everything uploaded to a major streaming service each day is now made by a machine. The flood has already arrived.
Deezer has launched a free tool that scans playlists on Spotify, Apple Music, SoundCloud, YouTube Music and other services to flag AI-generated tracks. It works across 20 of the most popular platforms and 27 languages, and it identifies fully AI-generated songs from the major generative models Suno and Udio with what Deezer says is over 99% accuracy.
The scale of the problem behind it is the real story. Deezer says it receives nearly 75,000 AI-generated tracks every day, which is around 44% of all new uploads. Those tracks pull just 1% to 3% of total streams, so the supply massively outstrips the demand. Deezer removes AI tracks from recommendations and editorial playlists, while Spotify and Apple Music have chosen to tag rather than remove.
Why it matters
This is a preview of what happens to every content channel as AI generation gets cheap. The volume explodes, the average quality drops and platforms are forced to build detection and labelling just to keep the experience usable. Music is simply further down the road than text and images.
For marketers, the signal is about trust and labelling. As AI content floods every feed, audiences and platforms start sorting hard for what is real. Authenticity stops being a nice-to-have and becomes a filter people apply. If your content is indistinguishable from machine-made filler, it gets treated like filler.
The share of new tracks uploaded to Deezer each day that are AI-generated, around 75,000 songs. Source: Music Ally
What to do about it
Decide where you stand on AI-made content. Using AI to produce is fine. Passing off generic machine output as something considered is what audiences punish.
Label honestly. As detection tools spread, getting caught hiding AI content costs more trust than disclosing it ever would.
Compete on what machines cannot fake. Real experience, a genuine point of view and proprietary data are the things that stand out in a flood of sameness.
Watch the platform rules. Some remove AI content, some tag it. Those choices will shape what gets distributed. Know the policy where you publish.
Raise your bar, not your volume. The answer to a flood of cheap content is not more cheap content. It is being clearly better.
Every channel is heading where music already is. A wave of machine-made content, then a scramble to sort signal from noise. The brands that win the next few years are the ones audiences can tell apart from the slop.