Skip to main content
Live
Main content

Sony Music and Warner Chappell sue Anthropic over Claude training data

Publishers accuse the AI lab of pirating millions of books containing lyrics and sheet music, building on the $1.5B Bartz precedent.

Jaeden Schafer
Editor in Chief · · 5 min read
Anthropic logo

Sony Music Publishing, Warner Chappell and a group of other music publishers sued Anthropic and co-founders Dario Amodei and Benjamin Mann on August 29, 2026, alleging the AI lab ran a "brazen campaign of illegally torrenting, scraping, and downloading copyrighted works" to train Claude. The complaint, filed late Friday in the U.S. District Court for the Northern District of California, accuses Anthropic of "blatant theft" involving thousands of copyrighted works. It arrives roughly two months after Anthropic was ordered to pay $1.5 billion in the Bartz v. Anthropic case, the largest copyright judgment yet against a frontier AI lab.

The publishers' theory is broader than earlier music-industry suits. They allege Anthropic obtained millions of copies of books through illegal torrenting, and that many of those books contained embedded lyrics and sheet music owned by the plaintiffs. In other words, the case argues that Anthropic's pirated book corpus doubled as a pirated music-publishing corpus.

brazen campaign of illegally torrenting, scraping, and downloading copyrighted works
Sony Music Publishing and Warner Chappell, Plaintiffs in the complaint

Anthropic could not be reached for comment prior to publication of the source report. The company has not yet filed a response in the docket.

Key facts

  • 01Sony Music Publishing, Warner Chappell and other publishers sued Anthropic and co-founders Dario Amodei and Benjamin Mann on August 29, 2026.
  • 02The complaint accuses Anthropic of pirating millions of copies of books containing lyrics and sheet music to train Claude.
  • 03The suit was filed in the U.S. District Court for the Northern District of California.
  • 04Anthropic was ordered to pay $1.5 billion in the Bartz v. Anthropic case, which turned on how training data was acquired.
  • 05Some of the same lawyers also represent Concord Music Group and Universal Music Group in a related January case.

The legal team behind the complaint has been building toward this filing for a year. Some of the same attorneys represent Concord Music Group and Universal Music Group in a case filed in January against Anthropic, and they also led Bartz v. Anthropic, in which a group of authors accused the company of using copyrighted works to train Claude and related products.

The Bartz ruling is the load-bearing precedent here. The judge in that case drew a sharp line: training an AI model on copyrighted works can be lawful under fair use, but acquiring those works through piracy is not. Anthropic was ordered to pay $1.5 billion on the acquisition question, not the training question. The Sony and Warner Chappell complaint leans directly on that distinction, alleging "flagrant piracy" through illegal torrenting as the underlying conduct.

For the music publishers, the strategic upside is that they do not need to relitigate whether AI training on lyrics is fair use. They can point to the Bartz record, argue Anthropic obtained their catalog through the same torrenting pipeline, and seek statutory damages on a per-work basis. Thousands of copyrighted compositions, multiplied by statutory damages ceilings, is the arithmetic that made Bartz a $1.5 billion judgment.

The suit also names Amodei and Mann personally. Naming founders as individual defendants is a pressure tactic — it raises the stakes for settlement discussions and complicates any future fundraising or M&A conversation, because personal liability exposure is harder to indemnify away than corporate liability. Whether the personal claims survive early motion practice is a separate question.

Anthropic's defense in Bartz did not dispute that training on copyrighted material was central to Claude's development; it disputed how the material was acquired. That is the same battleground here. If the publishers can show Anthropic pulled catalog music from torrent sources rather than licensing it, the fair-use umbrella from Bartz does not cover them. If Anthropic can show it later replaced pirated sources with licensed ones before commercial deployment, that narrows the damages window.

Related · from this week
Anthropic accuses Alibaba of 28.8M-query distillation attack on Claude
Jaeden Schafer · 5 min read →

There are still open questions. The complaint has not been tested against Anthropic's actual data-provenance records, and the company has not filed its response. Discovery in Bartz was extensive and painful; discovery in this case will likely reach further into Anthropic's training pipeline, including how books containing lyrics and sheet music were ingested and whether any filtering was applied. A settlement is possible before any of that becomes public.

The commercial pressure on Anthropic is now cumulative. A $1.5 billion Bartz judgment, an ongoing Concord and Universal case from January, and now a Sony Music Publishing and Warner Chappell filing that explicitly leverages the Bartz theory — the company is being asked to price legal liability into every future training run. That reshapes the economics of open-web data acquisition for the entire frontier-model industry, not just Anthropic. Labs that were quietly relying on shadow libraries and torrent archives now have a $1.5 billion data point telling them what that shortcut costs, and a second data point coming that could tell them what the music catalog premium looks like on top.

ShareXLinkedInEmail
AI Box

Every AI model. One chat.

The latest models from ChatGPT, Claude, Gemini, Sora, ElevenLabs — 80+ models in a single chat. Compare answers side by side. Pick the best one every time.

  • ChatGPT, Claude, Gemini, Grok, DeepSeek — in one chat
  • Generate images & video with Sora, Veo, Ideogram
  • Compare any two models side by side
  • From $8.99/mo · 80+ models, all included
Try AI Boxaibox.ai
Trusted by 3,000+ teams
Got a tip?

Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.

Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.

AI Box Daily briefingFree · Daily · No fluff

Stay ahead of everyone in AI.

The tightly edited AI news email engineers, founders, and investors actually open. One email. Every weekday. Five minutes to finish.

Loved by 10,000+ AI professionals
Free forever. Unsubscribe with one click.

The briefing read inside teams at

Keep reading

More from Security

Anthropic logo
Security

Anthropic accuses Alibaba of 28.8M-query distillation attack on Claude

A letter to the Senate Banking Committee calls it the largest known distillation attack on Anthropic to date.

Jaeden Schafer5 min read
Anthropic logo
Security

Anthropic pulls Claude Fable 5 after Trump administration 90-minute ultimatum

A jailbreak warning from Amazon researchers triggered emergency export controls, taking Anthropic's newest public model offline within hours.

Jaeden Schafer5 min read
Anthropic logo
Business

Anthropic hits near-$1T valuation while warning AI could destroy the world

Founded in 2021 by OpenAI defectors, Anthropic now sells Claude to the Pentagon and argues that dominating AI is the only way to make it safe.

Jaeden Schafer5 min read