Spotify has removed over 50 million AI slop and spam tracks from its recommendations in a year
Music streaming service Spotify says it has removed more than 50 million tracks considered AI-generated slop or spam from user recommendations over the past year. It has also changed its recommendation system to give more prominence to music made by human artists.
Spotify, the Sweden-founded music streaming service, announced on Tuesday that over the past year it has taken more than 50 million songs it considers AI-generated low-quality content, known as "slop", or spam out of user recommendations. The company has also reworked its recommendation system to highlight more music made by human artists. The move matters because streaming platforms are grappling with a rapid influx of AI-generated tracks.
Spotify says its editors follow the same standard and do not program content identified as slop or spam. According to the company, listener complaints about such content appearing in recommendations have dropped noticeably since the changes were introduced.
How content is identified
To judge whether a track may be AI-generated slop, Spotify uses several signals: listener engagement and feedback, upload patterns, metadata and third-party AI detection tools. Last year, the company reported removing more than 75 million tracks identified as spam.
AI personas and other platforms
Spotify has not required labels on all AI tracks, but it has introduced an "AI Persona" badge for artists who use an AI identity and do not represent a real person. The platform does not recommend music from such personas. Other streaming services have taken a tougher line: YouTube Music labels AI songs, Deezer lets users filter them out, and Bandcamp has banned AI music entirely.
Spotify is also partnering with Universal Music Group and the independent music publisher Believe to launch a tool for creating AI music remixes.
