Spotify has removed approximately 75 million AI-generated spam tracks from its platform over the past 12 months and plans to launch a dedicated detection filter this fall.
The purge targets what the company describes as AI-generated slop, a flood of fully synthetic songs that impersonate real artists, divert streaming royalties, and crowd out legitimate human-created music. When a track by a human artist gains traction, AI impersonators often release copycat versions that siphon away streams and royalty payments.
To curb the practice, Spotify will introduce an AI music spam filter in the autumn. The system is designed to identify uploaders and tracks that use AI to mimic existing artists, tag them, and halt their algorithmic recommendations. The company stressed that the filter will not penalize music that employs AI as a production assistant, such as for noise reduction or equalization, and will focus exclusively on fully AI-generated impersonator content.
Spotify indicated it is taking a measured approach to avoid mistakenly flagging legitimate artists. The detection system will be refined to distinguish between human-made recordings that use AI tools and entirely synthetic tracks designed to deceive listeners and divert revenue.
