The Pulse
Stanford Authors’ Books Surface in Anthropic Copyright Dispute
The Stanford Daily identified six Stanford professors whose books it says were used in Claude training, while court records distinguish fair-use training copies from pirated books kept in Anthropic’s library. The records also show that Chan
AI.info Team · Written with AI · fact-checked ·
Stanford authors’ books enter the Claude dispute
The Stanford Daily reported on Oct. 27 that books by Stanford professors Tobias Wolff, Abraham Verghese, Chang-Rae Lee, Daniel Mason, Hayagreeva Rao and Myra Strober were among works linked to Anthropic’s Claude training data. The report says 10 of Wolff’s publications had been used for AI training without permission. Wolff and 27 other authors say the $1.5 billion settlement in Bartz v. Anthropic does not resolve their objections to the company’s use of their work.
“To put me to work that way, without remittance, is a form of slave labor,” said Tobias Wolff, professor emeritus of English, in comments to The Stanford Daily.
The settlement covers 482,460 works on its Works List. The court granted final approval on July 20, 2026. The settlement’s estimated payment was about $3,000 per work, according to the final approval order.
Alsup’s ruling drew a line between two kinds of copying
The key decision in Bartz came on June 23, 2025—not in a May 2026 lawsuit. Judge William Alsup ruled that Anthropic’s use of copies to train specific large language models was fair use. Separately, he found that keeping pirated copies in a permanent central library was not fair use. The ruling did not find that Anthropic illegally trained Claude on the Stanford professors’ books.
The distinction matters: the court did not treat pirated library copies as training copies simply because some books might later be used for training. The order says Anthropic downloaded more than seven million pirated book copies and retained them in its library. It left questions about other copying unresolved and said a trial would address the pirated copies used to create the central library and resulting damages.
The May lawsuits are not the same case
The Stanford Daily’s report says Wolff and Lee opted out of the Bartz settlement and filed a separate non-class lawsuit, Cruz v. Anthropic. But the May 13, 2026, Cruz complaint does not list either author among its plaintiffs. Lee appears instead in a separate case, Kwon v. Anthropic, whose complaint was filed May 15. The court filing in that case names him as a plaintiff. The Cruz complaint seeks a jury trial and statutory damages, including up to $150,000 per infringed work.
The Stanford Daily compares that requested amount with the settlement’s estimated per-work payment of about $3,000. Those figures describe a request and a settlement estimate, not a damages award in the separate cases. The allegations in those suits remain claims, not findings that Anthropic infringed each plaintiff’s copyright through model training.
Authors want compensation and control
Wolff told The Daily that the settlement still gives Anthropic broad freedom to use literature it considers valuable. Stanford Law professor Paul Goldstein, whom The Daily describes as a plaintiff in the class action, pointed to the difficulty of licensing books one by one in an email to the publication:
“Millions of individual copyrighted works may be copied in the course of any training incident, and it’s simply impracticable for the platform to obtain licenses from the copyright owner of every — or, indeed, any — single work.”
The dispute involves more than the presence of pirated books in Anthropic’s library. The 2025 ruling separated that library from fair-use training copies; the authors’ separate lawsuits raise claims about particular works, damages and possible restrictions on future use. The settlement resolves covered claims for works on its Works List, while authors who opted out may pursue their own cases.