Spotify says it has removed more than 50 million tracks classified as low-quality AI content or spam from its recommendations this year. These tracks remain available on the service, but its systems and editors no longer suggest them to listeners. This differs from deleting tracks: last year, Spotify reported removing more than 75 million spam tracks for serious or repeated rule-breaking. It says those deletions continue.
During the past six months, Spotify has adjusted its recommendation system to promote more music by human artists. It examined tracks that listeners skipped, ignored or said they did not want to hear, and found a clear connection with content identified as spam or low-quality AI material. The company says complaints about such recommendations have since fallen. Its checks combine listener behaviour, upload patterns, track information and outside AI-detection tools, although those tools have limitations.
Spotify has also verified more than 300,000 human artists, giving them a badge that confirms their identity. Thousands of virtual artists carry an AI Persona label, which identifies an AI identity rather than a real person. Their tracks are excluded from recommendations by default, although exceptions can apply when a listener follows the artist.
