Spotify takes action against AI spam

Spotify wants to protect its users and artists from AI music with several measures. Among other things, musicians are to disclose in the future whether and in what form artificial intelligence is used in production. Meanwhile, new filters are to ensure that AI music does not eat away at the annual income of real artists.

Spotify announced the catalog of measures in a blog post on Thursday. In it, the Swedish music service describes three steps that are intended to tackle specific problems that are already occurring on the platform. For example, Spotify has introduced a complaint form to protect artists from counterfeiting and AI voice copies. Musicians can use this to contact Spotify if their voice has been illegally copied by AI. Artists must use this channel to complain about each AI song individually.

AI spam filter coming

A second issue that Spotify has identified: Although annual payouts are increasing overall, parts of them end up with AI spammers. Such users upload masses of song copies and short AI tracks to trick the algorithm and get as many plays as possible. From the autumn, a filter will therefore ensure that such uploads and the users behind them are automatically recognized. The algorithm will then recommend this content even more. To avoid catching the wrong people, Spotify wants to proceed cautiously at first while the recognition system is refined.

Finally, Spotify is working on a labeling system that allows artists to disclose the use of AI in the production of their songs. There is nothing fundamentally wrong with AI support, Spotify writes in the blog entry. However, there is a lack of a system to provide users with information about this. Spotify is currently developing such a metadata standard with the organization DDEX.

“This change serves to strengthen trust in the platform,” Spotify writes. “It is not about penalizing artists who use AI responsibly or downgrading tracks because they disclose information about their creation.”

(dahe)