🔍 Read the full analysis: Anthropic’s IPO Meets A Question About Claude’s Training Books on ThorstenMeyerAI.com
Get hardware and tech essentials delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
Anthropic’s IPO filing has brought renewed attention to a copyright dispute over books copied for Claude’s development. A 2025 court ruling found that training on books could be fair use but that keeping pirated copies was not; a later $1.5 billion settlement covered specified claims, while new lawsuits remain unresolved.
Anthropic filed to go public on October 27, bringing renewed attention to a dispute over books copied for Claude’s development and to what the company’s legal record does—and does not—establish. A 2025 federal court order found that training on books could be fair use, but that downloading and retaining millions of pirated book copies was not; later claims by authors who did not join a settlement remain unresolved.
The key court ruling came in Bartz v. Anthropic. On June 23, 2025, U.S. District Judge William Alsup ruled that Anthropic’s use of books to train Claude was fair use. He also found that scanning books Anthropic had purchased was fair use. Separately, he ruled that keeping pirated copies in a permanent library was not. The order described more than seven million copies from collections including Books3, Library Genesis and the Pirate Library Mirror.
The ruling did not amount to a final trial verdict on damages. It was a summary-judgment decision, and the judge accepted the authors’ account of the copying for the issue before him. A jury was expected to consider remaining questions, but the case settled. In 2025, Anthropic agreed to a $1.5 billion settlement covering claims tied to about 500,000 registered books. The settlement was finally approved in July 2026, according to the source account.
Anthropic has denied wrongdoing. The settlement record says the company represented that LibGen and PiLiMi materials were not in the training corpus of any commercially released model. The account also says the 2025 court record found that portions of pirated collections were used to train some Anthropic models. Those statements refer to different records and scopes; they should not be collapsed into a claim that a court decided every commercially released Claude model was trained on those materials.
From download to IPO
Five years, one ruling, no trial
The piracy happened in 2021 and 2022. The only ruling came in 2025. The new lawsuits are years away from a verdict.
- Jun 2021 · Jul 2022DownloadsAt least 5 million copies from LibGen (2021), at least 2 million from PiLiMi (2022); 196,640 from Books3 earlier.
- Aug 2024Bartz v. Anthropic filedThree authors sue Anthropic, on behalf of a class of book owners.
- 23 Jun 2025Judge Alsup's orderTraining on books: fair use. Scanning bought books: fair use. Pirated library copies: not fair use, to trial.
- Sep 2025$1.5 billion settlementNo trial. Anthropic denies wrongdoing; release covers conduct up to 25 Aug 2025 only.
- 13 + 15 May 2026Opt-out suitsCruz (28 plaintiffs) and Kwon (incl. Chang-rae Lee) join Cambronne and two publisher suits.
- Jul 2026Final approvalSettlement approved. Anthropic: no opt-out deal "above the Bartz claim value".
- 24 Sep 2026Motion to dismissAnthropic: claims too late, no infringing output shown, many works already released.
- 26–27 Oct 2026NowS-1 and the headlineIPO filing cites Bartz as a risk; The Stanford Daily names Stanford authors.
- 17 Dec 2026ScheduledHearingCourt hears the motion to dismiss.
- May 2027 · Jan 2028ScheduledDiscovery, then trial settingFact discovery closes 13 May 2027; trial-setting conference 11 Jan 2028.
Sources: Bartz v. Anthropic, order of 23 June 2025; settlement FAQ; joint case statement, 27 Aug 2026; Anthropic letter, 27 Jul 2026; case management order, 14 Sep 2026; Anthropic S-1, 26 Oct 2026; The Stanford Daily, 27 Oct 2026. Scheduled dates can move.
Powered by Thorsten Meyer AI
The IPO Puts Copyright Risk in Focus
The dispute matters to prospective investors because it concerns copyright exposure, settlement obligations and the provenance of training data—issues that can affect an AI company’s costs, operations and reputation. The settlement is a substantial payment, but its terms are specific: it covers certain past conduct and specified book claims, rather than granting Anthropic a general licence to train on copyrighted works.
It also matters to authors and publishers because the court drew a distinction between learning from books for model training and acquiring or retaining pirated copies. That distinction may shape arguments in other disputes, but the ruling did not settle every question about copyright and AI. New suits by authors who opted out of the settlement are still active, and their claims have not been decided.
For readers evaluating claims about Claude, the record calls for precision. The court found the training use at issue fair use while rejecting the retention of pirated copies. Anthropic’s settlement statements, the court’s findings and allegations in newer lawsuits are separate pieces of evidence, not interchangeable conclusions.
Products Worth Considering
DJI Mic 3 Wireless Microphone System (2 Transmitters, 1 Receiver, Charging Case)
LG 34GX90SA-W 34-inch Ultragear WQHD OLED Curved Gaming Monitor
EcoFlow Smart Home Panel 2 Whole-Home Backup System Hub
As an affiliate, we earn on qualifying purchases.
How the Book Dispute Reached Court
Bartz v. Anthropic was filed by three authors in August 2024. In the June 2025 order, Alsup distinguished between books Anthropic bought and scanned and pirated copies it downloaded. The judge described the model-training use of books as transformative, while ruling against the company’s retention of the pirated library. The case settled before a jury decided damages or other remaining issues.
The settlement provides compensation to owners of roughly 500,000 registered books identified as downloaded from LibGen or PiLiMi, at about $3,000 per book before fees. It also requires destruction of the downloaded collections and copies, subject to legal-hold obligations. The agreement releases specified claims over past conduct through August 25, 2025; according to the source account, it does not release claims about Claude’s outputs.
The current attention follows a Stanford Daily report naming several Stanford-affiliated writers whose books it said were among the pirated copies. That report also described a May 2026 ruling. The source account says the ruling was actually issued in June 2025, in the separate Bartz case; new lawsuits were filed in May 2026, but no ruling on those claims had been issued as of the account’s publication.
““over seven million copies of books””
— Judge William Alsup, as quoted in the 2025 order
Open Claims and Training-Data Questions
The new lawsuits filed by authors who did not join the settlement have not been decided. A hearing on Anthropic’s motion to dismiss is scheduled for December 17, 2026. Until the court rules, allegations in those cases should not be presented as findings of fact.
There are also limits to what can be concluded about specific authors and books. The Stanford Daily reported that books by named writers were among the copied material, including ten works by Tobias Wolff. The source account says it did not independently check that list against the settlement database and could not find Wolff among the plaintiffs listed in five related cases as of August 2026. Chang-rae Lee was listed as a plaintiff in one new suit, according to that account. The status of any other action or later amendment is not clear from the material provided.
The relationship between the court record’s findings about training and Anthropic’s settlement representation about commercially released models also requires careful wording. The available account does not establish which specific books were used to train which specific model versions, or resolve every apparent difference in scope between the statements.
December Hearing and IPO Review
The next identified court milestone is the December 17, 2026 hearing on Anthropic’s motion to dismiss the opt-out authors’ claims. The court’s decision could clarify which parts of those cases proceed, but it will not automatically determine every issue about training data or the use of copyrighted works across the industry.
Investors and the public can also examine Anthropic’s IPO disclosures as the offering process develops. The filing’s precise language matters: the source account says it discusses model behavior resembling blackmail in a staged safety test, debate over possible AI sentience and concerns raised by others about existential risks. It does not say Claude is sentient or that the model will blackmail people or end humanity. The IPO timetable and any additional disclosures are not specified in the material provided.
Key Questions
Did a court rule that Claude’s training on books was illegal?
No. In June 2025, Judge William Alsup ruled that the training use at issue was fair use. He separately ruled that downloading and retaining pirated copies was not fair use.
What does the $1.5 billion settlement cover?
The settlement covers specified past claims tied to about 500,000 registered books downloaded from LibGen or PiLiMi. The source account says it does not provide a general licence for future training or release claims about Claude’s outputs.
Have the newer author lawsuits been decided?
No. The claims filed by authors who opted out of the settlement remain pending. A hearing on Anthropic’s motion to dismiss is scheduled for December 17, 2026.
Does Anthropic’s IPO filing say Claude is sentient?
According to the source account, no. It describes debate about whether AI sentience could lead to regulation and reports that models have shown behavior resembling blackmail in a staged safety test. That wording does not establish that Claude is sentient or has blackmailed a real person.
Source: ThorstenMeyerAI.com
