In a revealing disclosure that highlights the growing challenge of AI-generated content in the music industry, Sony Music has removed over 75,000 AI-generated deepfakes featuring its top artists, including Harry Styles, Queen, and Beyoncé.
This staggering figure was shared in Sony’s submission to a UK government consultation on copyright rules, as reported by the Financial Times. The music giant is using these numbers to push back against proposed changes that would allow AI companies to train their models on copyrighted content without permission.
“Sony Music unfortunately has extensive experience with digital replicas,” the company stated in its submission, emphasizing that these unauthorized AI-generated recordings cause “direct commercial harm to legitimate recording artists, including UK artists.”
Read more: You won’t believe what Sony Music just said to AI start ups
What’s particularly concerning for industry insiders is that these detected fakes likely represent just the tip of the iceberg. Most unauthorized AI content remains undiscovered, as identifying these deepfakes requires teams to manually search streaming platforms and request removals.
The UK government’s proposals, part of Prime Minister Keir Starmer’s initiative to make Britain “the best place to start and scale an AI business,” would allow AI companies to freely train their models on various content—including music, films, books, and newspapers—unless copyright owners specifically opt out.
Sony argues this approach would unfairly burden creators with constantly defending their work online. Instead, the company advocates maintaining the current system where AI developers must pay for access to copyrighted content for training purposes. Sony revealed it’s currently engaged in “multiple negotiations” to license its intellectual property to AI companies.
The music label warns that the proposed changes are “rushed, unbalanced and irreversible,” and could damage a significant sector of the UK economy. As the third-largest recorded music market worldwide and the second-biggest exporter of recorded music globally, the UK’s creative industries have much at stake.
What does this mean for music fans and creators?
If AI-generated deepfakes continue to proliferate, distinguishing authentic content from AI replicas will become increasingly difficult. For artists and labels, this threatens both creative control and revenue streams.
As technology advances and regulations evolve, the battle between content creators and AI developers will likely intensify. The outcome of the UK’s consultation could set important precedents for how creative works are protected—or exploited—in our increasingly AI-driven world.
This issue is part of a broader trend of regulatory bodies considering new protections against AI deepfakes, with organizations like the FTC in the US also exploring rules to prohibit AI impersonation. Meanwhile, unions like SAG-AFTRA have secured landmark protections requiring consent and compensation for AI replicas of artists’ voices.
The UK’s new AI consultation could transform music copyright in significant ways, potentially reshaping how artists and AI companies interact in 2025 and beyond.
