The Pulse
Spotify Took 50 Million Slop and Spam Tracks Out of Recommendations
Spotify says it removed more than 50 million slop and spam tracks from recommendations in 2026, while separately deleting more than 75 million spammy tracks for rule violations. The company says its filters combine listener feedback, upload
AI.info Team · Written with AI · fact-checked ·
Spotify says it took more than 50 million tracks it classified as “slop and spam” out of recommendations this year, a figure that does not mean the service removed 50 million AI-generated songs. The company’s October 20 update groups AI-generated material with spam, and does not break down how many of the tracks involved were made with AI. Spotify describes the measure as a change to what listeners are recommended, not a catalog-wide deletion.
50 million out of recommendations, not necessarily off Spotify
The recommendation figure is separate from another total in Spotify’s update: more than 75 million spammy tracks removed from the platform altogether for serious or repeated rule violations. The company does not explain whether those categories overlap, so the figures should not be added together as a combined count of unique tracks. A track excluded from recommendations may still remain available on the service.
Spotify says it changed its recommendation system repeatedly over the past six months, after reviewing tracks listeners skipped, ignored or said they did not want to hear. The company says those tracks consistently matched its identification of slop or spam. Its editors apply the same standard and do not program content identified that way.
Spotify combines listener signals with detection tools
The company says no single signal can reliably identify slop or spam. Its system combines listener engagement and feedback with upload patterns, metadata and information from multiple third-party AI detection tools. Spotify also says AI detection has limitations and will become harder as generative systems advance.
Spotify is watching for other behaviors it associates with spam, including mass uploads, duplicate tracks, search-engine manipulation, spam-focused playlist networks and artificially short songs. The company says it will keep updating its systems as uploaders change tactics. It reports a marked decrease in listener concerns about this content appearing in recommendations, but provides no number or time series for that change.
AI labels and verification address a different problem
The update also describes tools aimed at helping listeners and artists distinguish identities and credits. Spotify says artists submit tens of thousands of AI Credits each day, which identify roles or instruments made with AI; listeners can see those disclosures in Song Credits in the mobile Now Playing view. That is a disclosure mechanism, distinct from the systems Spotify uses to decide what to recommend.
Spotify says more than 300,000 artists are verified, representing more than 99% of what listeners search for on the service; more than 70% of the verified artists are independent. It also says thousands of AI Personas carry labels and are excluded from editorial or algorithmic recommendations by default, with an exception when a listener follows one.
The 50 million figure leaves key questions open
Spotify’s report does not define a public threshold for labeling a track slop or spam, provide a breakdown by detection signal, or say how many tracks in the recommendation count were confirmed as AI-generated. That distinction matters: the figure measures material Spotify says it kept out of recommendations, not a verified count of AI songs. Its own account also acknowledges that AI detection tools have limits.
The company cites its 1,000-stream royalty threshold, in place since 2024, as another barrier to large-scale low-performing uploads. Spotify says tracks with indicators of a high probability of being AI-generated are 10 times more likely than other tracks to fall below that threshold. The update does not specify the number of tracks behind that comparison.