Editorial illustration for Music Publishers Cite Anthropic Staff's Piracy Chat in Lawsuit
Music Publishers Sue Anthropic Over Piracy Claims
Music Publishers Cite Anthropic Staff's Piracy Chat in Lawsuit
Anthropic's $1.5 billion settlement with book authors, announced earlier this year, was supposed to close the chapter on claims that the company pirated more than 7 million books to train Claude. Music publishers don't see it that way. Sony, EMI, and Warner Chappell filed a lawsuit Friday arguing that the settlement did nothing to stop Anthropic from doing the same thing with song lyrics and compositions, and that the payout was pocket change for a company now valued at $2 trillion.
The publishers say Anthropic's torrenting habit swept up "thousands upon thousands" of copyrighted musical works, including songbooks covering the Beatles' full catalog and Taylor Swift's biggest hits. Court filings say Anthropic intends to hold onto this pirated material "forever" for AI training purposes. That detail, along with internal staff communications about where to find pirated content, features prominently in the new complaint. Publishers argue Claude doesn't just learn from these works but reproduces them, spitting out verbatim lyrics or generating songs that borrow the "heart" of popular tracks, sometimes at a user's request to rewrite one artist's lyrics in another's style.
“$1.5 billion is obviously not a large enough settlement to deter infringing conduct by a company that has parlayed such mass infringement into a staggering $2-trillion-dollar valuation,” music publishers said in a lawsuit filed Friday.
Why this matters
The $1.5 billion book settlement was supposed to be the deterrent that made everyone else in AI training data cleanup their act. Sony, EMI, and Warner Chappell's filing suggests it's doing the opposite: they're arguing the number is small change against a $2 trillion valuation, and internal chats celebrating Z-Library make the "we didn't know better" defense harder to sell next time. For founders building on scraped corpora, the math here is instructive.
A nine-figure settlement barely dents a company's valuation if the underlying model still ships and still earns. That's a signal to boards, not just lawyers: provenance audits and licensing deals are cheaper than litigation exposure, but only if regulators or courts eventually make the penalties scale with company size rather than dataset size. Researchers should also note the discovery angle.
Slack chats and internal messaging are now evidence in copyright suits, which means training pipeline decisions made casually months ago can resurface in front of a judge. Data hygiene just became a legal function, not just an engineering one.
Related Reading
Browse all Anthropic news →
Business & Startups Nordic pilot adds Gemini for Education, NotebookLM to boost AI literacy
Business & Startups Kling launches Video O1, all-in-one model with MVL bridge using transformer
Business & Startups DeepSeek Seeks More Capital Weeks After USD 7B Funding Round
Related Topic Claude gains shared context in Excel, PowerPoint; Microsoft adds Copilot Cowork
AI Tools & Apps
Related Topic Trump cracks down on Anthropic after Amazon tip; staff largely foreign
Open Source
From Archives OpenAI's ChatGPT Ad Business Hits USD 1 Billion Annual Run Rate
From Archives AI labs buy tens of thousands of Mac minis to train agents amid chip shortage
From Archives Anthropic's Claude Code limit change: a raise on paper, a cut in practice
Research & Benchmarks
From Archives Anthropic Proposes Standard for AI to Control Physical Hardware
AI Tools & Apps
Common Questions Answered
Why are music publishers filing a lawsuit against Anthropic despite the $1.5 billion book settlement?
Music publishers Sony, EMI, and Warner Chappell argue that Anthropic's settlement with book authors does nothing to address the company's alleged use of song lyrics and compositions to train Claude without permission. They contend that the $1.5 billion payout is insufficient to deter infringement by a company valued at $2 trillion, and that internal staff communications celebrating piracy sources like Z-Library demonstrate intentional misconduct.
What evidence of piracy did Anthropic staff discuss in internal chats cited in the lawsuit?
According to the lawsuit filing, Anthropic staff members exchanged messages extolling Z-Library, a known piracy platform, with one message stating 'Zlibrary my beloved.' These internal communications are being used as evidence that the company knowingly used illegally obtained materials to train its AI models, undermining any defense that Anthropic was unaware of the infringement.
How much did Anthropic settle with book authors, and why do publishers say it's insufficient?
Anthropic agreed to pay $1.5 billion to settle claims that it used more than 7 million pirated books to train Claude. Music publishers argue this settlement amount represents only a small fraction of the company's $2 trillion valuation, making it an inadequate deterrent against continued mass infringement of copyrighted musical works.
What broader implications does this lawsuit have for other AI companies training on scraped data?
The lawsuit suggests that nine-figure settlements may not be sufficient financial deterrents for AI companies to stop using copyrighted material without permission, especially when their valuations far exceed settlement amounts. This creates a concerning precedent where the financial incentive structure makes large-scale data scraping a potentially profitable strategy rather than a risky one for well-funded AI firms.
Further Reading
- Papers with Code - Latest NLP Research - Papers with Code
- Hugging Face Daily Papers - Hugging Face
- ArXiv CS.CL (Computation and Language) - ArXiv
