Spotify removes 50 million 'slop and spam' AI tracks from recommendations

Spotify has removed more than 50 million tracks identified as “slop and spam” from its recommendations this year as the streaming platform grapples with the growing use of artificial intelligence to create music. 

In an Oct. 20 statement, the company said the move was part of its efforts to keep spam and misleading content from appearing in listeners' recommendations.

"We looked closely at the tracks listeners were skipping, ignoring, or telling us they didn’t want to hear. A clear pattern emerged: These tracks were consistently ones identified as slop or spam," Spotify said. "We’ve made numerous updates to our recommendation system [...] Since we made these changes, we’ve seen a marked decrease in listener concerns about this content showing up in recommendations."

To identify potentially problematic content, the platform considers several factors, including listener engagement and feedback, upload patterns, metadata, and information from third-party AI detection tools.

The company also monitors tactics such as mass uploads, duplicate tracks, search engine optimization manipulation, spammy playlist networks, and artificially short songs.

"AI detection technology in particular has limitations, and it will only get harder as generative AI evolves," Spotify added.

Spotify has also introduced AI Credits, which allows artists to disclose how artificial intelligence was used in a song, and AI Persona badges for artist profiles believed to represent AI-generated identities.

Under its current policy, tracks from AI Personas are excluded from editorial and algorithmic recommendations by default, unless listeners choose to follow those artists.