News
Spotify Says It Pulled Over 50 Million AI Spam Tracks From Recommendations in 2026
Spotify says it has removed more than 50 million tracks it calls AI "slop" and spam from recommendations since the start of 2026. Critics note that this does not mean the tracks were deleted from the platform.
Spotify announced on October 20, 2026 that it has taken more than 50 million tracks out of recommendations since the start of the year, which the company describes as "AI slop" and spam. The figures come from a company blog post published Tuesday, which updates the AI policy changes announced in September 2025.
What Spotify announced
According to NBC News, Spotify analyzed patterns in tracks that listeners most often skipped or ignored over the past six months. These consistently overlapped with artificially generated audio. The company combined listener feedback with metadata and third-party AI-detection tools, then modified its recommendation system so that such content surfaces less often.
Spotify also uses upload patterns. The platform admits that detection has limitations and will get harder as generative AI advances. It also says it has seen a "marked decrease" in listener complaints about AI spam in recommendations.
AI moves fast and bad actors adapt just as fast. - Spotify, quoted by NBC News
Recommendations are not removal
Engadget points to a distinction that is easy to miss in the headlines. The 50 million figure refers to removal from recommendations, not necessarily deletion from the platform. The figure of more than 75 million tracks removed for violations, meanwhile, is the same as in September 2025, and a Spotify spokesperson did not provide an update.
The outlet notes that Spotify labels profiles suspected of being virtual artists with an "AI Persona" badge. Such profiles are skipped in recommendations by default and do not appear in editorial playlists. They can still show up in recommendations for listeners who follow them, for example. Artists can also submit AI credits, which disclose where artificial intelligence was used, and Spotify says tens of thousands are filed every day.
Money and criticism
The financial angle explains why the platform cares. Artists are paid partly based on their share of total streams, so a flood of AI tracks can dilute earnings. Morgan Hayduk, CEO of Beatdapp, which sells fraud-detection software, put it bluntly to NBC: if the denominator grows because of AI slop, the percentage each legitimate rights holder receives shrinks.
Engadget is critical of Spotify's approach. The author accuses the company of trying to fight slop while also adding generative features, such as planned remix and cover tools. The piece also notes that Spotify offers no simple toggle to block AI content, unlike Deezer, which plans such a filter, or Bandcamp, which largely bans AI. Earlier this year, a Spotify spokesperson told the BBC that the company's priority is "harmful uses" of AI such as spam and impersonation, not filtering music by how it was made.
What comes next
Spotify says a tool that lets artists approve what appears on their profiles will be available to all creators by the end of the year. The 1,000-stream threshold still determines royalty payouts, and according to Spotify, tracks with a high probability of being AI-generated are far more likely to fall below it.
The report, however, is data from the platform itself, with no independent audit, so the effectiveness of the filters remains to be verified.
