Anthropic faces $1.5B lawsuit over AI training on pirated books

A federal judge in San Francisco has certified a class action lawsuit against Anthropic on behalf of nearly every US book author whose works were used to train the company’s AI models, marking the first time a US court has allowed such a case to proceed in the generative AI context. The ruling exposes Anthropic to potentially catastrophic damages that could exceed $1 billion and threaten the company’s survival, despite its recent $100 billion valuation target.

The big picture: Judge William Alsup made a crucial distinction between training AI models on legally acquired books (which he deemed fair use) and Anthropic’s wholesale downloading and storage of millions of pirated books from sites like LibGen and PiLiMi (which he found to be clear copyright infringement).

What’s at stake: The financial exposure is staggering, with damages potentially ranging from $1.5 billion to $750 billion depending on how many works are covered and what a jury awards.

• Assuming only two-fifths of five million books qualify for damages at the statutory minimum of $750 per work, Anthropic faces $1.5 billion in liability.
• At the statutory maximum of $150,000 per work across five million books, damages could reach $750 billion—what Anthropic’s lawyers called “ruinous.”
• The company acknowledged in court filings that the books covered likely number “in the millions.”

Why Anthropic is particularly vulnerable: The judge described this as the “classic” candidate for a class action due to Anthropic’s systematic approach to acquiring pirated content.

• Internal messages show engineers celebrating when they found new sources of pirated books, with one co-founder writing “[j]ust in time” after discovering a mirror of the FBI-shuttered Z-Library.
• The company kept pirated books accessible to engineers in multiple copies and used them for various internal tasks beyond training.
• When pirate sites were taken down, Anthropic scrambled to torrent fresh copies.

The legal precedent: Santa Clara Law professor Ed Lee warned the ruling means “Anthropic faces at least the potential for business-ending liability.”

• Judge Alsup “has all but ruled that Anthropic’s downloading of pirated books is [copyright] infringement,” leaving only damages calculation for the jury.
• If Anthropic loses at trial, it must post a bond for the full damages amount during any appeal, typically requiring 100% collateral plus annual premiums.
• The previous record was Cox Communications’ $1 billion verdict in 2019 for music piracy, which was later overturned.

Industry implications: Other AI companies face similar risks if Anthropic’s case sets precedent, though their legal strategies differ.

• OpenAI and Microsoft face 12 consolidated copyright suits in New York, where a similar ruling could create even greater liability.
• Meta’s internal messages show engineers calling LibGen “obviously pirated” data, with the approach approved by Mark Zuckerberg.
• The outcome could determine whether AI training on copyrighted material requires permission from rights-holders across the entire industry.

What they’re saying: President Trump recently dismissed strict copyright enforcement as “not doable,” arguing that “China’s not doing it,” though his AI Action Plan remains silent on copyright issues.

Anthropic’s response: The company told Obsolete it “respectfully disagrees” with the decision and is “exploring all avenues for review,” arguing the court “failed to properly account for the significant challenges and inefficiencies of having to establish valid ownership millions of times over in a single lawsuit.”

The funding pressure: The legal threat comes as Anthropic seeks investment from Gulf states including the UAE and Qatar to compete with better-funded rivals.

• CEO Dario Amodei acknowledged in a staff memo that this would likely enrich “dictators,” calling it a “real downside.”
• The company has raised $17.2 billion total, though much came as Amazon and Google cloud credits rather than cash.
• Meta has been offering nine-figure pay packages to poach AI researchers, with Zuckerberg committing “hundreds of billions of dollars” to AI development.

What’s next: The trial is scheduled for December 1st, with Anthropic facing a choice between a potentially multibillion-dollar settlement or risking catastrophic damages at trial that could bankrupt the four-year-old company.