Music companies and distributors have agreed new rules to clamp down on the use of AI-generated music to defraud streaming services and divert royalties away from legitimate artists. Industry executives estimate that almost 10 per cent of streams are now deemed to be fraudulent, with AI generating much of the music. Streaming fraud involves the creation of huge numbers of tracks from simple prompts before using networks of bot accounts to play them repeatedly, generating large sums in royalty fees for the owner of the tracks and diminishing the pot of cash available for genuine artists.

French streaming service Deezer said that AI-generated songs made up more than half of new music uploaded in July, averaging 90,000 tracks a month. It also found that up to 85 per cent of streams of fully AI-generated tracks last year were fraudulent. The International Federation of the Phonographic Industry (IFPI), which represents recorded music companies worldwide, will say on Monday that major distributors have agreed to new guidelines to help stop fraudulent AI-generated tracks reaching streaming platforms. “Streaming fraud misleads fans and deprives legitimate artists and songwriters of income that should rightfully flow to them,” said Victoria Oakley, IFPI chief executive. “This is theft, plain and simple.” Distributors will commit to do more to prevent fraudulent tracks being uploaded, including verifying rights ownership and customer identities.

They have also agreed to vet content for infringement, fraudulent activity and AI-related risks, and will act against repeat offenders and share information. The Streaming Integrity Initiative is supported by more than a dozen labels and distributors such as Sony Music, Universal Music, Virgin Music, Absolute Label Services, Revelator, Warner Music, Absolute, FUGA and The Orchard. The organisations are calling on remaining distributors and digital service providers to back the initiative, which establishes a set of baseline practices for the industry to help prevent, detect and respond to streaming fraud.

The new initiative builds on the End Streaming Fraud campaign, launched by the IFPI and leading music companies to raise awareness of the threat posed to the music ecosystem. Some streamers have already committed to tackling AI-generated music fraud. Deezer said it excludes fraudulent AI streams from royalty payments when they are discovered.

Some countries have also started to crack down on music fraudsters. A man pleaded guilty this year in the first US criminal prosecution for AI-assisted streaming fraud, while Brazilian authorities have targeted businesses selling fake streams, followers and engagement.