Spotify to introduce AI label and spam filter to stop AI music slop

When AI Sols began to turn on Spotify – bands like sunset in velvet for example – users have urged Spotify to do something. They wanted a label showing that music on their weekly discovery and that the recommendations were created by AI. Some users have even gone so far as to say that they should “boycott spotify” until a label is made.

On Thursday, Spotify said that he would start doing exactly that, saying in a press release that “the aggressive protection against the worst parts of General AI is essential to allow its potential for artists and producers”. The platform incorporates a new spam filtering system, IA disclosure and “improving the application of identity violations” like Deepfakes.

Spotify has worked with DDEX, or the exchange of digital data, which is an organization of standards in the music industry, to require a “new standard of industry for IA disclosure in musical credits”. Indeed, as Spotify says, many artists use AI tools in a responsible manner when creating music, so the addition of a simple “AI” or “Not ia” label does not really solve the problem of listeners who want to know if they listen to the music of AI.

“This standard gives artists and rights holders a way to clearly indicate where and how AI played a role in creating a piece – whether it is voice, instruments or post -production generated by AI,” said Spotify in its press release. “This change consists in strengthening confidence through the platform. It is not a question of punishing the artists who use the AI ​​of the tracks in a responsible or downstream manner to disclose information on the way in which they were manufactured.”

“To its best, the AI ​​unlocks new incredible ways for artists to create music and so that the listeners discover it. At worst, I can be used by bad actors and content farms to confuse or deceive listeners, push` Sold ” in the ecosystem and interfere with authentic artists working to build their career, “said the press release from Spotify. “This kind of harmful AI content degrades the user experience for listeners and often tries to divert the royalties towards bad actors.”

The new spotify impulse policy published specifically diverts the way it plans to give artists stronger protections against the vocal clones of AI. Spotify plans to attack spam music – such as “mass downloads, duplicates, referencing hacks, artificially short track abuses and other forms of sloat” – by unfolding a new system which “will identify the downloaders and the tracks engaged in these tactics, label them and stop recommending them”. They will start in a conservative way so that they do not accidentally punish bad people, then add more signals as the system accelerates.

“These updates are the latest of a series of changes that we make to support a more reliable musical ecosystem for artists, for right -handers and listeners. We keep them to come as the technology evolves, so stay attentive,” wrote Spotify.