Sony Music ramps up takedowns of AI deepfake songs on streaming apps

In short: Sony Music says it is filing far more takedown requests as AI-made fake songs and other streaming scams spread across music apps.

What's going on

Sony Music told the Financial Times it asked streaming platforms to remove more than 260,000 tracks by the end of September. That is almost double the 135,000 tracks it had asked to be taken down by the end of March.

Sony says many of the tracks are “generative AI deepfakes,” meaning songs made with AI that copy an artist’s voice or image without permission (like a convincing impersonator). Sony said this is happening “at scale,” and that it harms artists, misleads fans, and makes it harder to use AI in fair ways.

The issue goes beyond celebrity impersonations. The industry also says streaming services are being flooded with large batches of AI-made songs that are played over and over by bots (automated accounts), mainly to earn royalty payments. Sony executive Dennis Kooker has said fraudulent streams could account for as much as 10 percent of tracks across streaming sites, and label executives estimate streaming fraud could cost the industry up to $2.2 billion a year.

Other companies report similar pressure. Deezer said this summer that about 90,000 AI-generated songs were being uploaded to its platform every day, more than half of new music. Deezer also said up to 85 percent of streams of fully AI-generated tracks last year were fraudulent.

What to watch

Sony said it expects platforms to act quickly and stop removed content from coming back, but called the situation an “uphill struggle.” Sony and Universal Music are also pursuing legal action against AI music company Suno, which they accuse of copying recordings without a licence. Suno has said the labels’ claims are flawed.

Source: Financial Times