Stanford Daily Says Claude Trained on Books by 10-Plus Stanford Professors; Two Trackers Disagree on Who Sued in Cruz v. Anthropic
A.I. / news
Stanford Daily Says Claude Trained on Books by 10-Plus Stanford Professors; Two Trackers Disagree on Who Sued in Cruz v. Anthropic
The Daily names Tobias Wolff and Chang-Rae Lee as Cruz plaintiffs. A litigation tracker lists Lee in a separate 34-author suit and Wolff outside the 28-name complaint list.
The Stanford Daily reported on Oct. 27 that Anthropic trained its Claude models on pirated books by more than 10 Stanford professors. The article, by April Yu, says the books came from the shadow libraries LibGen and PiLiMi.
The Daily's piece carries no statement from Anthropic. No other source read for this report carries one either.
What the Stanford Daily reported
The Stanford Daily article names six Stanford-affiliated authors whose books it says were used: Tobias Wolff, professor emeritus of English, with 10 of his publications; Abraham Verghese; Chang-Rae Lee; Daniel Mason; Hayagreeva Rao; and Myra Strober. It also names Paul Goldstein of Stanford Law as a plaintiff in the Bartz class action.
It says Wolff and Lee filed Cruz v. Anthropic in May 2026 after opting out of the Bartz settlement. Wolff is quoted: "To put me to work that way, without remittance, is a form of slave labor."
The Daily puts the Bartz settlement at $1.5 billion, covering nearly 500,000 registered works at roughly $3,000 a work. It says Cruz seeks a jury trial and statutory damages of up to $150,000 per infringed work, plus a permanent injunction against Claude training on the plaintiffs' works.
Where the plaintiff lists do not match
The litigation tracker ChatGPT Is Eating the World covered the Cruz filing on May 14. It says the complaint is dated May 13, names 28 plaintiffs led by Angie Cruz, and says the group is not filing a class action.
Its introduction names Dave Eggers, Andrew Sean Greer and Tobias Wolff. The 28-name list on the same page includes Eggers but neither Greer nor Wolff.
Chang-Rae Lee appears elsewhere. A May 16 post on the same tracker places Lee among 34 authors, with R.O. Kwon, Sonia Nazario and Erika Sanchez, in a separate suit. The tracker's Cruz category page says Anthropic faced nine lawsuits at that point.
| Author | Stanford Daily | ChatGPT Is Eating the World |
|---|---|---|
| Tobias Wolff | Cruz plaintiff | Named in intro, absent from 28-name list |
| Chang-Rae Lee | Cruz plaintiff | In separate 34-author suit |
| Paul Goldstein | Bartz class plaintiff | Not mentioned |
The two accounts can both be right if the complaint was amended or the post's introduction was loose. The court docket is the only document that settles it, and it was not among the pages read.
Where the cases stand
The tracker's latest entry, dated July 29, says Anthropic opposed a motion to consolidate briefing on motions to dismiss in five suits, Cruz among them. Anthropic argued that "the respective complaints have their own deficiencies." No ruling on those motions appears in either source.
The Daily summarizes Judge William Alsup's Bartz ruling this way: legally acquired books used for training were fair use, while building a permanent central library from pirated copies was not. The tracker adds that Alsup rejected the "dilution" theory, which Judge Chhabria endorsed only in dicta in Kadrey v. Meta. The Cruz complaint rests on dilution.
The Daily counts more than 70 infringement suits against AI companies. For Anthropic's other exposure, see the S-1 coverage, and for a separate argument over machine-written text, Wikipedia's cleanup of AI content.
The next dated item is a ruling on Anthropic's motions to dismiss, which neither source schedules.
Sources
More in A.I.
- 01Lean Creator Leo de Moura Says AI Makes Proofs Cheap. Here Is What the Checkers ShowOn a podcast released October 29, the AWS scientist who built Lean claimed AI-written proofs now run to 50 million lines. The claims check out only in part, and he names the two gaps himself.
- 02Claude's Watermark Quality Claim Rests on Single Responses, Blogger SaysAnthropic says watermarking does not affect output quality. A post citing Google's own SynthID-Text paper says the tests never measured what happens when you ask the same question twice.
- 03Nanbeige4.3-33B Ships Under Apache-2.0 With a 524,288-Token Window and Self-Reported WinsNanbeige Team's 33B model is built as a teacher for smaller agents. Its own table has Qwen3.8-27B ahead on SWE-Bench Pro and GDPval.
- 04GPT-5.6 Sol Launched a Factorio Rocket on Peaceful Mode, With Hauling Done by HandBlogger Ryan Madden's single run took 13.5 in-game hours. The agent carried supplies between areas because it never linked some production lines.