Sony asks platforms to remove 260,000 AI tracks

Sony Music is stepping up its fight against streaming fraud as it confronts a surge in artificial intelligence tracks imitating stars like Harry Styles and Adele. The company told the Financial Times that it asked streaming platforms to take down more than 260,000 songs by the end of September. That figure almost doubles the 135,000 tracks it tried to remove by late March.

The music group noted that deepfakes copying voices and likenesses without permission harm artists, mislead fans, and undermine the ethical use of artificial intelligence. Target artists include Britney Spears, Queen, and Michael Jackson. In addition to unauthorized impersonations, services face mass-generated tracks played repeatedly by bots to capture royalties meant for legitimate creators.

Dennis Kooker, president of Global Digital Business at Sony Music, stated earlier this year that fraudulent streams could represent up to 10% of songs on music platforms. Record label executives estimate that streaming fraud could be costing the industry up to $2.2 billion per year.

French streaming platform Deezer reported that an average of 90,000 AI-generated songs were uploaded to its service daily during the summer, making up more than half of all new music. Deezer found that up to 85% of plays on fully AI-generated tracks last year were fraudulent. A report from Sonic Intelligence Academy found that 92.8% of AI tracks submitted to its registry were created using Suno, while DistroKid distributed over three quarters of them.

Sony and Universal Music are pursuing legal action against AI platforms such as Suno, accusing them of using unlicensed recordings to train generative models. A spokesperson for Suno said the record labels' claims were fundamentally flawed in both fact and law.