Spotify said on Tuesday it removed 50 million tracks it considers AI “slop and spam” from its recommendations, while changing discovery to give more prominence to music from human artists. The company made the recommendation changes over the past six months after reviewing tracks that listeners were skipping, ignoring or rejecting. The figure concerns recommendation removal; it does not mean Spotify confirmed that all 50 million recordings were deleted from its catalogue.
Key facts
- Spotify says 50 million tracks classified as slop and spam have been removed from recommendations.
- Its editors do not program content identified as slop or spam, and Spotify says listener concerns about such recommendations have fallen.
- An AI Persona badge marks artists using an AI identity that does not represent a real person; their music is excluded from default recommendations and editor-curated playlists.
- Detection combines listener engagement and feedback, upload patterns, metadata, and third-party AI detection tools.
- Artists are submitting tens of thousands of AI credits each day to describe how AI was used in uploads.
Why Spotify’s recommendation change matters
Spotify’s recommendation system shapes what listeners encounter, so excluding 50 million tracks can change which songs and artists gain exposure. The company says its changes over the last six months were designed to promote more music from “authentic, human artists.” Its editors apply the same standard by refusing to program content identified as slop or spam. Spotify says it has seen a marked decrease in listener concerns since making the changes.
Spotify also says its royalty threshold is intended to hinder scaled AI spam. It keeps a 1,000-stream threshold before paying royalties on a track, and tracks with indicators showing a high probability of being AI-generated are 10 times more likely to fall below that threshold. More than 300,000 human artists are now verified on the service, covering more than 99 percent of the artists users actually search for.
What Spotify counts as slop and spam
Spotify’s 50 million figure is a recommendation-cleanup figure, not a claim that every AI-made track was removed from the service. The company uses several signals to flag possible slop or spam: engagement and feedback from listeners, upload patterns, metadata, and third-party detection tools. It acknowledges that detection technology has limitations and may struggle as generative AI systems improve, so its systems will need continual updates as spammers alter their tactics.
The broader removal total needs separate context. Spotify says it has removed more than 75 million spammy tracks from the platform for serious or repeated violations of its rules. However, the company offered the same figure in September 2025 and did not provide an updated catalogue-removal number, so the 50 million recommendation exclusions should not be added to that total as though both were new deletions.
Labels, credits and Spotify’s wider AI policy
Spotify has not required labels on every AI track. Instead, artists can submit AI credits describing the technology’s role in a song’s creative process. Spotify says tens of thousands of these credits arrive each day; mobile listeners can see them in the Now Playing view, and the service plans to add more ways to surface disclosures. Thousands of virtual artists now carry the AI Persona label, according to the company. Music from these personas will not appear in editor-curated playlists or default recommendations, although users who follow them can still hear their tracks.
Spotify also does not provide a general listener switch for blocking all AI-generated music. A spokesperson told the BBC that its priority is addressing harmful uses such as spam and impersonation rather than filtering music according to how it was made. Other services have taken different routes: YouTube Music labels AI songs, Deezer has announced a user filter for generated audio, Bandcamp largely forbids AI music, and Tidal does not pay royalties for fully AI-generated tracks.
Spotify is also allowing AI remixing and covers and is partnering with Universal Music Group and Believe to launch a tool for AI remixes. For broader context on the technology, see GetKG’s AI section.
What’s next
Spotify says it must keep updating its detection and recommendation systems because spammers can change tactics. The company plans to make AI credits easier to find and to make its artist-profile approval tool available to all artists by the end of the year. It has not provided a refreshed platform-removal total, so its figure of more than 75 million spammy tracks remains the number cited in September 2025. For broader updates on platform policy, see GetKG’s news section. Spotify says listener concerns have already fallen following its recommendation changes, making further reductions in complaints and increased visibility for verified human artists useful indicators to watch.
Sources: The Verge, Engadget — Technology News & Expert Reviews.
