AIThis post was created with the assistance of artificial intelligence (AI).

🔍 Read the full analysis: Claude And Pirated Books: A Guide To The Claims Against Anthropic on ThorstenMeyerAI.com

Before you orderOffer from Amazon

Get the little things that make your day delivered free with Prime

  • Fast, free delivery on millions of items
  • Prime Video, Amazon Music and more included
  • Member-only deals all year
Start your free Prime trial Free trial for eligible customers · Cancel anytime
As an affiliate, we earn on qualifying purchases.

TL;DR

A federal judge ruled in June 2025 that Anthropic’s use of books to train Claude was fair use, while finding that the company’s downloading and retention of pirated book copies was not. Anthropic later agreed to a court-approved $1.5 billion settlement over roughly 500,000 registered books; separate lawsuits filed by authors in 2026 have not been decided.

A federal judge ruled in June 2025 that Anthropic’s use of books to train Claude was fair use, but that downloading and keeping millions of pirated book copies was not. The company later received final approval for a $1.5 billion settlement covering about 500,000 registered books, while separate lawsuits filed by authors in 2026 remain unresolved.

The ruling came in Bartz v. Anthropic, a copyright case brought by three authors in 2024. On June 23, 2025, U.S. District Judge William Alsup distinguished between how Anthropic obtained books and how it used them. He found that training Claude on books was fair use and that buying print books, cutting off their bindings and scanning them was also fair use. But he ruled against Anthropic’s downloading and retention of pirated copies.

The court order described 196,640 books from Books3, at least five million copies from Library Genesis, or LibGen, and at least two million from the Pirate Library Mirror, or PiLiMi. It said the total exceeded seven million copies. The order was a pretrial ruling, not a jury verdict; Alsup accepted the authors’ account of the piracy for purposes of the decision, and a jury had been expected to address remaining issues, including damages. The case settled before that trial.

The settlement, finally approved in July 2026, provides $1.5 billion for owners of roughly 500,000 registered books downloaded from LibGen or PiLiMi. That works out to about $3,000 per book before fees. The agreement requires Anthropic to destroy the downloaded collections and their copies, subject to legal-hold requirements. Its release applies to past conduct through August 25, 2025, and does not release claims about Claude’s outputs.

Anthropic disputes wrongdoing. In the settlement, the company represented that the LibGen and PiLiMi datasets, or portions of them, were not in the training corpus of any commercially released large language model. That statement sits alongside the court record describing pirated copies and training; the distinction between datasets downloaded, materials used in training and commercially released models is central to understanding the dispute.

At a glance
reportWhen: As of October 28, 2026; a hearing on An…
The developmentNew 2026 author lawsuits have revived scrutiny of Anthropic’s use of pirated books, while the record distinguishes a 2025 fair-use ruling from claims that remain unresolved.
Thorsten Meyer AI
Court record

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.

7M+pirated book copies downloaded, per the 2025 court order
$1.5Bclass settlement, final approval July 2026
~$3,000per book, for about 500,000 books
5related opt-out cases now before one judge
  • 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.

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 Difference Between Training and Piracy

The case matters because it draws a legal distinction that is often lost in shorthand descriptions of AI copyright disputes. The judge found that training on books can qualify as fair use, while ruling that obtaining and keeping pirated copies was unlawful on the facts presented. The decision does not mean every form of AI training on copyrighted work is lawful, nor does it grant Anthropic a general licence to use books.

The settlement also sets a substantial financial consequence for the downloading at issue: $1.5 billion for registered works, alongside a requirement to destroy the specified pirated collections. For authors and publishers, the case offers one path to compensation while leaving questions about training practices and AI-generated outputs open. For AI companies and users, it shows why the source of training material and the way it is acquired can carry separate legal consequences.

Products Worth Considering

Bose TV Speaker (2nd Gen) Compact Soundbar with Bluetooth and HDMI-ARC, Black

Bose TV Speaker (2nd Gen) Compact Soundbar with Bluetooth and HDMI-ARC, Black

As an affiliate, we earn on qualifying purchases.

How the Cases Differ

The headline “Anthropic illegally trained Claude with Stanford professors’ books” compresses separate events and legal issues. The June 2025 decision was in Bartz, not in a May 2026 lawsuit. The 2026 filings are new cases, not a ruling that Anthropic illegally trained Claude on the named authors’ books.

The Stanford Daily reported that books by Tobias Wolff, Abraham Verghese, Chang-rae Lee, Daniel Mason, Hayagreeva Rao and Myra Strober appeared among pirated copies, and that ten Wolff works were affected. That identification is the newspaper’s reporting; the source material says the book list has not been checked against the settlement database. Chang-rae Lee is listed as a plaintiff in one of the newer cases, while Wolff was not found among the plaintiffs listed in five related cases as of August 2026. The reported authors’ list and the court’s findings should not be treated as interchangeable.

Authors filed the newer suits in May 2026, alleging copyright violations. Anthropic has sought dismissal, and the record described here contains no court ruling on those claims. The settlement of Bartz covers specified past conduct; it does not decide the merits of the new cases.

““Buying a copy later will not absolve it of liability for the theft.””

— Judge William Alsup, on Anthropic’s pirated-book downloads

What the New Claims Establish

No judge has ruled on the claims filed by authors in May 2026. The December 17, 2026 hearing is scheduled to address Anthropic’s motion to dismiss, not to decide whether the new allegations are true. The lawsuits’ factual and legal outcomes remain open.

It is also not clear from the material available whether every book identified in newspaper reporting is included in the settlement’s registered-book database. The settlement’s coverage, Anthropic’s account of what entered commercially released models, and the court’s earlier findings concern related but distinct questions. The court did not hold a jury trial on damages in Bartz, because the parties settled.

December Hearing and Settlement Payments

The next scheduled court milestone for the newer cases is the December 17, 2026 hearing on Anthropic’s motion to dismiss. The court could rule after the hearing, but no outcome is established in the current record.

The approved settlement also sets payment deadlines. According to its terms, the final $450 million instalment is due by September 27, 2027; another $450 million instalment was due by September 25, 2026. The settlement requires destruction of the specified LibGen and PiLiMi copies, subject to legal-hold obligations. Further reporting on payments, compliance and the new lawsuits may clarify how the separate proceedings develop.

Key Questions

Did a court rule that all training of Claude on copyrighted books was illegal?

No. In June 2025, Judge William Alsup ruled that training Claude on books in the Bartz case was fair use. He separately ruled that Anthropic’s downloading and retention of pirated copies was not fair use.

What does the $1.5 billion settlement cover?

It covers claims over roughly 500,000 registered books downloaded from LibGen or PiLiMi, with payments averaging about $3,000 per book before fees. The agreement also requires destruction of the downloaded collections and copies, subject to legal-hold requirements.

Have the authors’ 2026 lawsuits been decided?

No. The newer lawsuits remain unresolved. A hearing on Anthropic’s motion to dismiss is scheduled for December 17, 2026.

Did Anthropic say pirated books were used to train released Claude models?

Anthropic represented in the settlement that the LibGen and PiLiMi datasets, or portions of them, were not in the training corpus of any commercially released model. That company statement is distinct from the court’s findings about pirated downloads and training in the earlier case.

Source: ThorstenMeyerAI.com

You May Also Like

ByteDance Adds Model Safety And Alignment To Seed AI Lab’s Structure

TechNode reports Seed has created a department for model safety and alignment, but its leadership, authority and responsibilities are not yet known.

How To Digitize Documents In A Paper-Heavy Small Organization

A proposed scanner workflow would test automatic document naming and filing at five small offices, using real backlogs to measure accuracy and demand.

The referral. How AI search severs the content-for-traffic contract that funded the open web.

AI search now answers queries directly, ending the traditional referral traffic to publishers—impacting revenue models and small publishers most.

The 4.8 Staircase: What the Market Actually Believes About Claude’s Next Release

Market signals suggest a likely release of Claude 4.8 by mid-2026, but official confirmation from Anthropic is still pending. Here’s what is known and what remains uncertain.