Spotify to distinguish AI artists from real people – and stop recommending them

Spotify, the world's leading music streaming service, is set to introduce new measures to distinguish artificial intelligence-generated artists from human performers. Beginning in September, the platform will label "AI personas" and subsequently remove them from its personalized recommendation algorithms.

This initiative aims to foster trust and authenticity between artists and fans, ensuring listeners are aware of the identity behind the music they consume. The labels will appear on artist profiles and alongside tracks within playlists. Spotify stated that a true artist-fan connection relies on trust, which becomes increasingly vital in the era of generative AI.

The company will not solely depend on self-disclosure from creators; human review and AI investigative tools will be employed to identify fake artists. This move comes after Spotify removed 75 million spam tracks from its service over a 12-month period last year, a consequence of AI tools flooding the platform with artificial content. Spotify's legitimate catalog currently features 100 million tracks.

Listeners have expressed a clear preference for transparency, disliking profiles that appear human but are revealed to be AI-generated. Last year, the origins of the 70s-style US rock band The Velvet Sundown sparked debate before their social media confirmed them as an AI creation. The group, which attracts 117,000 monthly listeners, currently features a "synthetic music project" description on its Spotify page but lacks explicit AI labeling. Similarly, soul singer Sienna Rose, with over half a million monthly listeners and a rapid release schedule, has generated discussion due to her apparent lack of real-world presence, though her profile does not specify her AI status.