🔍 Read the full analysis: Claude’s Book Data Debate: What’s Proven And What Remains A Claim on ThorstenMeyerAI.com
Get smart everyday buys 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
A 2025 federal court ruling found that training Claude on books was fair use, while ruling against Anthropic’s downloading and retention of more than seven million pirated book copies. Anthropic later agreed to a court-approved $1.5 billion settlement covering about 500,000 books, while new opt-out authors’ claims remain unresolved.
The record in Anthropic’s book-copyright dispute is more specific than claims that a court found Claude’s training illegal: in June 2025, a federal judge ruled that training the model on books was fair use, but found that Anthropic’s downloading and permanent retention of more than seven million pirated book copies was not. Anthropic later agreed to a $1.5 billion settlement, while separate lawsuits filed by authors who opted out have not yet been decided.
In Bartz v. Anthropic, Judge William Alsup’s June 23, 2025 summary-judgment order treated different kinds of copying differently. It found that using books to train Claude qualified as fair use and that Anthropic’s purchase and scanning of used print books was also fair use. But the judge ruled against downloading and keeping pirated digital copies. The order described 196,640 books from Books3, at least five million from Library Genesis and at least two million from the Pirate Library Mirror.
The decision was not a final jury verdict on damages. On the piracy issue, Alsup accepted the authors’ account of the facts at the summary-judgment stage; a jury was expected to consider remaining issues. The case settled before that trial. The court-approved settlement provides $1.5 billion for owners of roughly 500,000 registered books, equivalent to about $3,000 per book before fees. It also requires destruction of the downloaded LibGen and PiLiMi books and copies, subject to legal-hold obligations.
Anthropic has denied wrongdoing. In the settlement, it represented that neither the LibGen nor PiLiMi datasets, or portions of them, were in the training corpus of any commercially released large language model. The settlement covers specified past conduct through August 25, 2025, and does not release claims about Claude’s outputs. New lawsuits by authors who opted out are separate; no judge has ruled on those claims, and a hearing on Anthropic’s motion to dismiss is scheduled for December 17, 2026.
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
What the Ruling Means for AI Training
The dispute matters because it draws a legal distinction between learning from books through model training and acquiring copies from sources the court treated as pirated. The 2025 decision did not establish that every use of copyrighted books to train an AI system is unlawful; it found the training in this case to be fair use while ruling against the separate copying and retention of pirated files.
The settlement gives eligible rights holders a substantial payment for the covered books, but it is not a general licence for future training and does not resolve every author’s claim. Authors who opted out are pursuing separate cases, leaving courts to address claims outside the settlement. For readers, authors and AI developers, the outcome may shape how future cases distinguish training practices, source material and the handling of copied works.
Popular right now
Vitamix 5200 Professional-Grade Blender, 64 oz. Container, Black
Ninja Luxe Café Premier 3-in-1 Espresso Machine, Drip Coffee & Cold Brew
Vitamix Propel Series 510 Professional-Grade Blender, 3 Programs, Black
As an affiliate, we earn on qualifying purchases.
From Bartz to the Opt-Out Cases
The original case, Bartz v. Anthropic, was filed by three authors in August 2024. In June 2025, Alsup issued his order on the competing copying practices. Anthropic and the authors later settled the claims covered by the agreement, which received final approval in July 2026. The settlement’s release is limited to past conduct within its stated scope; it does not settle claims about the outputs of Claude.
Authors who did not join the settlement have brought new lawsuits. A report by The Stanford Daily named Stanford-affiliated writers, including Tobias Wolff, Abraham Verghese and Chang-rae Lee, among people whose books it said were in the pirated collections. That book list has not been independently verified here against the settlement database. Court filings reviewed for the source report list Chang-rae Lee as a plaintiff in one of the new cases, but did not list Wolff among plaintiffs in five related cases as of August 2026. Those details do not establish whether Wolff is involved in another action or a later filing.
“spectacularly so”
— Judge William Alsup, in the June 23, 2025 order
What the New Claims Have Not Settled
The opt-out authors’ cases remain unresolved. No ruling has established that the claims in those suits are true, and the December hearing will address Anthropic’s request to dismiss rather than decide every factual issue. The precise scope of the authors and works involved may change as cases proceed or filings are amended.
There is also a distinction between the court record and Anthropic’s settlement representation. The 2025 order records the downloading of pirated collections and discusses their use in training, while Anthropic represented in the settlement that the LibGen and PiLiMi datasets were not in the training corpora of its commercially released models. The full extent of how particular books related to particular models is not resolved by the available material. Nor does the settlement determine the new authors’ claims or create a broad rule for all AI training.
December Hearing and Remaining Payments
The next scheduled court milestone for the opt-out cases is the December 17, 2026 hearing on Anthropic’s motion to dismiss. The court’s decision on that motion will determine whether some or all claims can proceed, but it will not by itself establish that the allegations are proven.
The settlement also includes staged payments. The source report says a $450 million instalment was due by September 25, 2026, and a further $450 million is due by September 27, 2027; it does not confirm whether the September 2026 payment has been made. The parties’ filings and later court orders may clarify both the payment status and the course of the pending lawsuits.
Key Questions
Did a judge rule that Claude’s book training was illegal?
No. In June 2025, Judge Alsup ruled that training Claude on books was fair use. He separately ruled against Anthropic’s downloading and retention of pirated copies.
What does the $1.5 billion settlement cover?
The court-approved agreement covers specified past claims involving roughly 500,000 registered books downloaded from LibGen or PiLiMi. It does not grant a general licence for future training or release claims about Claude’s outputs.
Have the authors who opted out won their lawsuits?
No. Their lawsuits are pending, and no ruling has decided their claims. A hearing on Anthropic’s motion to dismiss is scheduled for December 17, 2026.
What did Anthropic say about the pirated collections and released models?
In the settlement, Anthropic represented that the LibGen and PiLiMi datasets, including portions of them, were not in the training corpus of any of its commercially released models. That statement is Anthropic’s representation, not a separate court finding resolving every question about historical training.
Source: ThorstenMeyerAI.com
