Sony Music Says AI Deepfake Song Takedowns Have Surpassed 260,000
Sony Music says its campaign against AI-generated deepfake songs has reached a new milestone: more than 260,000 removal requests tied to tracks that allegedly imitate or misuse its artists' voices, identities, or catalog associations.
The figure, reported by the Financial Times and echoed by music-industry trade outlets, marks a sharp escalation in the fight over unauthorized AI music uploads. Industry observers increasingly see the number not just as a headline statistic, but as evidence that voice cloning and fake artist releases are becoming a growing enforcement burden for labels and rights holders.
Sony says the pace is accelerating
What stands out is how quickly the number has grown. Earlier coverage in 2026 said Sony had filed more than 135,000 takedown notices related to AI deepfakes. Going back further, reporting in 2025 cited a substantially lower total.
That does not by itself prove how much AI-generated impersonation exists across every platform, but it does suggest Sony's enforcement activity is rising quickly. In practical terms, the company appears to be signaling that the problem is expanding faster than traditional notice-and-removal systems were designed to handle.
The reported problem is scale, not just discovery
A recurring theme in the coverage is that once one deepfake track is removed, similar uploads can reappear elsewhere or be reposted again. Sony and several trade publications describe that pattern as a whack-a-mole problem, where enforcement is continuous rather than one-and-done.
Because the available reporting relies heavily on Sony's account, it is worth being cautious about broader conclusions. Still, the central concern is easy to understand: if AI tools make imitation faster and cheaper, rights holders may have to spend more time identifying copies, filing notices, and checking whether the same material has resurfaced.
What kinds of misuse are at issue
The dispute is not limited to songs that merely sound AI-generated. Recent coverage focuses on tracks that allegedly impersonate real artists, mimic recognizable voices, or present fake releases in ways that could confuse listeners.
Some reports also connect AI deepfakes to wider risks such as fan deception, fake artist uploads, and scam-adjacent behavior. Those claims should be understood as stakeholder concerns rather than settled findings in every case. But they help explain why labels are framing the issue as one of identity, consent, and artist control, not only copyright.
Why the policy debate is intensifying
The music industry has increasingly pointed to examples like Sony's takedown totals to argue for stronger federal protections around digital replicas of a person's voice or likeness. In the United States, the main bill cited in that debate is the NO FAKES Act, which has been reintroduced in Congress.
Supporters say the proposed legislation could create a clearer nationwide framework than the current patchwork of laws and claims. Even so, many observers note that legislation alone may not solve the practical enforcement challenge if platforms continue to face high volumes of uploads and reposts.
What is known, and what remains uncertain
The clearest reported fact is the number Sony now attaches to its takedown effort: more than 260,000 requests. Also well supported is the broader timeline showing that Sony had previously cited lower totals, including more than 135,000 earlier in 2026.
Less certain are the harder-to-measure effects, such as how much listener confusion these uploads cause, how much revenue may be diverted, or how often removed tracks reappear in altered form. Available coverage points to a fast-growing enforcement challenge, but much of the framing still comes from Sony and allied industry reporting.
For now, the bigger story may be the direction of travel. If takedown numbers keep climbing at this pace, the conflict over AI-generated music impersonation is likely to become an even more visible test of how platforms, lawmakers, and rights holders respond to synthetic media.