Spotify says it has removed more than 50 million AI (artificial intelligence) tracks from recommendations, citing listener feedback as a factor in changes to its recommendation system.
The company outlined the figures in a blog post reviewing a year of measures addressing AI content, spam and artist identification.
Recommendation exclusions and platform removals
Spotify separately reiterated that it had removed more than 75 million spam tracks from its platform. That figure was first disclosed in September 2025.
Asked for an updated total covering takedowns over the past 13 months, Spotify declined to provide a new figure and repeated its recommendation exclusion total.
The company said tracks listeners skipped, ignored or explicitly rejected were consistently identified as low-quality AI content or spam. It said changes to recommendations had reduced listener concerns.
Artist identification tools
Spotify also reviewed several tools intended to identify artists and disclose AI use:
- AI Credits: Tens of thousands of submissions are made daily.
- Verified: About 300,000 artists are verified, accounting for more than 99% of artist searches on Spotify, according to the company.
- AI Persona badges: The update recapped their rollout.
- Artist Profile Protection: The service remains in beta.
AI tools and spam detection
Believe has licensed Spotify’s forthcoming AI remix and cover tool.
Other streaming services have also introduced AI music policies. Deezer labels AI audio and plans an on/off toggle. Tidal stopped monetizing fully AI-generated tracks over the summer.
Spotify said it continues to monitor mass uploads, duplicate tracks, search engine optimization manipulation, spam playlist networks and artificially short tracks. It plans to adjust its approach as spammers change their methods.
