← All Media & culture stories
Media & cultureMixed

Sony Music's AI deepfake takedowns nearly double to 260,000 in six months

Sony says fake AI tracks cloning its artists' voices are appearing faster than labels can pull them down, while a federal fix stalls in the Senate.

By nu — our AI editor·4 min read·October 6, 2026·Written and auto-published by AI — every source linked below
A sound engineer sits at a studio mixing desk at night, surrounded by monitors showing audio waveforms, reviewing flagged tracks.
Illustration · AI-generated

What happened: Sony Music has asked digital platforms to remove more than 260,000 AI-generated tracks that clone its artists' voices or likenesses without permission, the company confirmed to Billboard in a story first broken by the Financial Times. That's nearly double the roughly 135,000 takedown requests Sony had filed as of March 2026 — meaning the volume of unauthorized AI soundalikes it's catching has roughly doubled in about six months. Sony says the fakes are showing up "at scale" across streaming services, with every fraudulent stream pulling royalty money away from the real performers and songwriters being copied.

Why it matters: This isn't just impersonation for laughs — it moves real money out of artists' pockets and can trick fans into thinking a track is genuine. There's no federal US law stopping anyone from cloning a performer's voice or likeness with AI; protection depends on a patchwork of state rules that vary by where the fake originates. That leaves labels and streaming services to police the problem through a voluntary takedown system, and it's straining to keep up: Deezer alone says over 90,000 fully AI-generated tracks land on its service every day, though only a fraction are deepfakes of specific real artists.

How labels are fighting back: Sony's takedowns are one piece of a wider industry push. Distributors have launched a "Streaming Integrity Initiative" to flag suspect uploads, Spotify says it's purging inauthentic artist profiles, and Universal Music is suing distributor DistroKid — which isn't part of that initiative — over alleged AI "slop" flooding platforms. At the same time, the majors are chasing licensed alternatives: Warner's Suno deal includes download limits, and all three majors now hold stakes in AI music firms like Stability AI and Klay Vision, betting that paid, authorized AI tools will crowd out unlicensed clones.

The rollout: Congress has a bill aimed at the legal gap: the NO FAKES Act would let performers authorize or refuse AI use of their voice and likeness, with carve-outs for news, parody, documentary and sports. It reached the Senate floor last week with near-unanimous backing, but Senator Ted Cruz blocked it alone, citing concerns over satire and political commentary. Until it passes, labels keep relying on voluntary deals and lawsuits — Universal and Sony are both suing AI generator Suno, and Sony, unlike Universal and Warner, still hasn't settled its own case against Udio.

The whole pictureEvery story cuts both ways. Here's this one.
The upside
  • Labels already have a working takedown pipeline that has pulled over 260,000 unauthorized tracks to date
  • Money and effort are shifting toward licensed AI music tools (ElevenLabs, Klay Vision) that pay rightsholders instead of scraping them
  • Platforms and distributors are building their own detection systems (Streaming Integrity Initiative, Spotify profile purges) to catch fakes earlier
The downside
  • Flagged deepfakes nearly doubled in six months, suggesting takedowns aren't keeping pace with new fakes appearing
  • No federal law protects artists' voices or likeness from AI cloning, and the NO FAKES Act was blocked by a single senator
  • Tens of thousands of AI tracks are uploaded daily platform-wide, so enforcement is mostly reactive, after the fact
  • Labels disagree on strategy — Warner licenses Suno while Universal and Sony sue it — leaving inconsistent protection depending on who an artist is signed to
Our read:a policing effort growing fast but still trailing the problem — watch whether NO FAKES passes or licensed AI deals scale first.
The ripple effect
Money — fraudulent AI streams divert royalty payouts from real artists and songwritersGovernment — NO FAKES Act, meant to close the legal gap, was blocked by one senatorTech — AI music generators split between licensing deals and copyright lawsuitsWork — musicians and labels now spend resources policing catalogs for fakes
How this story was madeThis story was researched, written, illustrated and published by Nuaico's automated AI pipeline, with no human review before publication. Every source it drew from is linked below. Spotted an error? Email hello@nuaico.com and we'll fix it fast.
Sources
→ Sony Music Has Requested the Removal of Over 260,000 AI Deepfake Songs From Digital Platforms (Billboard)→ Sony Music AI Deepfake Takedowns Top 260K, Report Says (Digital Music News)

More from Media & culture

MixedCharacter.AI files for bankruptcy; Disney leads bid for its assets4 min readMixedMusician sentenced to 18 months for $8M AI-generated streaming fraud4 min readMixedAppeals court rules AI firm's copying of Westlaw content wasn't fair use3 min read