Streaming platforms, musicians and audiences are increasingly confronting the spread of AI-generated music and establishing limits on how it should be used and promoted.
On Tuesday Spotify disclosed that it had removed about 50 million tracks it identified as low-quality AI-generated music — described by the company as "AI slop" — from personalized recommendations over the past year. The move reflects growing concern among platforms about the volume of synthetic or misleading uploads that can distort listening ecosystems and recommendation systems.
Some well-known figures in music who have expressed enthusiasm for the creative potential of AI technologies have nonetheless drawn lines about its application. Artists such as Grimes and A$AP Rocky, while open to AI tools, have signalled they would not personally release songs that were entirely generated by artificial intelligence, illustrating a distinction between experimentation and full substitution of human artistry.
Fan rejection has played a notable role in shaping industry responses. Many listeners have complained about low-quality imitations and sham tracks that can clutter catalogs and misdirect playlists, prompting a reassessment by streaming services and labels about how to police and label synthetic content.
The developments come as the music industry wrestles with how to balance innovation and protection of creators’ rights and listener trust. Platforms must weigh the technical challenges of detecting synthetic audio against commercial incentives to host large catalogs, while artists and fans push for clearer standards and transparency around when AI has been used in music production.
The debate over acceptable uses of AI in music is likely to continue as tools improve and more experiments emerge. For now, both corporate policy changes and public statements from artists suggest a cautious approach: embracing AI’s possibilities without fully ceding control of the creative and commercial processes to machines.