🔍 Read the full analysis: Pirated Books And Claude: What Anthropic Is Accused Of And Why It Matters on ThorstenMeyerAI.com
Get business pricing on tech for your team
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
A 2025 federal court ruling found that training Claude on books was fair use but that Anthropic’s downloading of millions of pirated books was not. A $1.5 billion settlement covering about 500,000 books was approved in 2026; separate lawsuits by authors who opted out are still pending, and Anthropic denies wrongdoing.
Anthropic’s use of books to train Claude and its downloading of pirated copies are distinct legal issues, according to a 2025 federal court ruling and later settlement records. The court found that training on books was fair use, but that downloading more than seven million copies from pirate collections was not; a $1.5 billion settlement covers claims tied to roughly 500,000 books, while separate lawsuits from authors who opted out remain unresolved.
In Bartz v. Anthropic, U.S. District Judge William Alsup ruled on June 23, 2025, that Anthropic’s training of Claude on books was fair use. He also said the company’s purchase and scanning of print books was fair use. The ruling drew a different conclusion about copies Anthropic downloaded from online collections: it described about 196,640 books from Books3, at least five million from Library Genesis and at least two million from the Pirate Library Mirror. Alsup wrote that downloading and keeping the pirated copies was not excused by fair use.
The ruling was a pretrial decision, not a jury verdict on damages. It accepted the authors’ account of the facts about pirated copies, but the case settled before a jury decided the remaining issues. The settlement, finally approved in July 2026, totals $1.5 billion and provides about $3,000 per book for owners of roughly 500,000 registered titles, before fees. Its terms require Anthropic to destroy the downloaded LibGen and PiLiMi books and copies, subject to legal-hold requirements.
Anthropic disputes wrongdoing. In settlement materials, the company represented that neither the LibGen nor PiLiMi datasets, nor portions of them, were in the training corpus of any commercially released large language model. The settlement resolves specified past claims through August 25, 2025; it does not grant a general license to train on books, and claims about Claude’s outputs were not released.
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 distinction matters because copyright liability does not turn on a single question of whether AI training is lawful. The judge treated the use of books to train a model differently from the acquisition and retention of pirated copies. The ruling does not establish that every form of AI training on copyrighted material is fair use, nor does it grant companies permission to obtain books from pirate libraries.
For authors, the settlement puts a substantial dollar value on one defined set of claims, but it does not resolve every dispute over how AI companies acquire or use copyrighted works. For AI developers, it highlights that the source and handling of training materials can carry separate legal risk from the training itself. The unresolved cases may test claims outside the settlement’s covered period and scope.
Products Worth Considering
LG 34GX90SA-W 34-inch Ultragear WQHD OLED Curved Gaming Monitor
LG UltraGear 27-Inch QHD OLED Gaming Monitor, 240Hz, 0.03ms, G-SYNC Compatible, FreeSync Premium Pro, HDMI 2.1
Amazon Kindle Paperwhite (12th Generation, 2024), 16 GB, 7-inch Display, Jade
As an affiliate, we earn on qualifying purchases.
The Bartz case was filed by three authors in August 2024. Alsup’s June 2025 order separated the authors’ claims into training on books, scanning books Anthropic had bought, and downloading books from unauthorized collections. The parties settled before a trial could determine damages over the pirated copies. The court-approved settlement applies to registered books identified under its terms, rather than establishing a blanket arrangement with all authors.
Authors who did not join the settlement have filed separate cases. A report by The Stanford Daily named writers including Tobias Wolff, Abraham Verghese, Chang-rae Lee and Daniel Mason as affected by pirated copies. That list is the newspaper’s reporting; the book entries have not been independently verified here. Chang-rae Lee appears as a named plaintiff in one of the newer suits, but the available filings cited in the source do not list Wolff among plaintiffs in five related cases as of August 2026.
The source article also discusses Anthropic’s public filing, or S-1. It says the filing describes model behavior “resembling blackmail” in a staged safety test, refers to debate over possible AI sentience as a potential regulatory issue, and notes existential concerns raised by others. Those passages are not findings that Claude is sentient or that it has blackmailed people.
““spectacularly so””
— Judge William Alsup, in the June 23, 2025 order
Claims Still Awaiting a Court Ruling
The newer authors’ lawsuits have not been decided. A hearing on Anthropic’s motion to dismiss is scheduled for December 17, 2026, but the outcome and any later trial schedule are unknown. The source material says there was no ruling in May 2026 on those claims; the relevant ruling was issued in June 2025 in the separate Bartz case.
The settlement resolves defined claims involving past conduct, not every dispute about Anthropic’s systems or their outputs. It is also not clear from the available material how every author or title identified in press coverage maps to the settlement database or the newer cases. Anthropic’s statements about the training corpus are its representations in settlement materials, not a separate judicial finding about every model or dataset.
December Hearing and Case Scope
The next scheduled milestone is the December 17, 2026 hearing on Anthropic’s motion to dismiss the newer suits. The court’s decision could determine which claims proceed, though no outcome has been announced. The settlement also sets out remaining payment instalments, including $450 million due by September 27, 2027, according to the terms summarized in the source material.
Readers should watch for court orders clarifying which authors and books are included in the pending cases and how those claims differ from the settled Bartz action. Until then, the confirmed record supports a specific distinction: a court found the training use fair in that case, rejected fair use as a defense for the pirated downloads, and did not issue a final trial judgment on damages.
Key Questions
Did a judge rule that Claude was illegally trained on books?
No. In June 2025, Judge William Alsup found that training Claude on books was fair use in the Bartz case. He separately ruled that downloading and retaining pirated copies was not fair use.
What does the $1.5 billion settlement cover?
It covers specified claims involving about 500,000 registered books that Anthropic downloaded from LibGen or PiLiMi. The settlement provides about $3,000 per book before fees and requires destruction of the downloaded books and copies, subject to legal-hold obligations.
Does the settlement mean Anthropic admitted wrongdoing?
No. Anthropic disputes wrongdoing. The company represented in settlement materials that the LibGen and PiLiMi collections, including portions of them, were not in the training corpus of commercially released models.
What happens to authors who opted out?
Some have filed separate lawsuits that remain unresolved. A hearing on Anthropic’s motion to dismiss is scheduled for December 17, 2026.
Does Anthropic’s S-1 say Claude is sentient or could blackmail people?
Not as a confirmed fact. The source describes the filing as referring to model behavior “resembling blackmail” in a staged safety test, debate about possible sentience and existential concerns raised by others. Those references do not establish that Claude is sentient or has blackmailed anyone.
Source: ThorstenMeyerAI.com
