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

🔍 Read the full analysis: What’s Alleged About Claude And Pirated Books—and What’s Proven? on ThorstenMeyerAI.com

Buying for a business?Offer from Amazon

Get business pricing on tech for your team

  • Business-only prices and quantity discounts
  • Tax-exempt purchasing
  • Multiple users, one account, clear invoices
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, but that downloading more than seven million pirated book copies to build a library was not. Anthropic later agreed to a $1.5 billion settlement covering roughly 500,000 books; the settlement did not establish liability at trial, and separate authors’ lawsuits remain unresolved.

A June 2025 federal court order found that Anthropic’s training of Claude on books was fair use, but that the company’s downloading of more than seven million pirated book copies to create a permanent library was not. Anthropic later agreed to a $1.5 billion settlement over the claims, which covers about 500,000 books; it did not include a jury verdict on damages or resolve newer lawsuits brought by authors who opted out.

The ruling came in Bartz v. Anthropic, a copyright case filed by three authors in August 2024. In an order dated June 23, 2025, U.S. District Judge William Alsup distinguished between using books to train a model and acquiring copies of books through piracy. He ruled that training Claude on books was fair use and also said Anthropic’s purchase and scanning of used print books was fair use. But he found that downloading pirated copies and keeping them in a central library was not fair use.

The order described 196,640 books from Books3, at least five million copies from Library Genesis, and at least two million from the Pirate Library Mirror. The judge accepted the authors’ account of the piracy at the summary-judgment stage. A jury was expected to determine issues including damages, but the case settled before a trial on those questions. The order is not a finding that all Claude training was illegal; it draws a distinction between the training use the judge approved and the acquisition and retention of pirated copies he rejected.

Under the settlement, Anthropic agreed to pay $1.5 billion to owners of roughly 500,000 registered books identified as downloaded from LibGen or PiLiMi, or about $3,000 per book before fees. The agreement also requires destruction of the downloaded books and copies, subject to legal-hold obligations. It releases claims over past conduct through August 25, 2025, while expressly excluding claims about Claude’s outputs. Anthropic has denied wrongdoing. The settlement court’s materials say the company represented that neither the LibGen nor PiLiMi datasets, nor portions of them, were in the training corpus of any commercially released model.

At a glance
reportWhen: As of October 28, 2026; the main ruling…
The developmentA renewed dispute over claims that Anthropic trained Claude on pirated books has prompted scrutiny of what a 2025 court order actually decided and what a later settlement resolved.
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

What the Settlement Covers

The distinction matters to authors, AI developers and users because copyright liability depends on what was copied and how, not simply on whether a model learned from books. The order treated training and scanning purchased books differently from building a lasting collection from pirated files. It is a specific ruling in one case, not a broad decision that settles every copyright question involving generative AI.

The settlement offers a payment process for eligible rights holders, but it is not a jury’s calculation of damages and does not establish a general licence for AI training. Its scope is limited to the claims it releases and the books covered by its registration process. The agreement’s treatment of training data also needs careful wording: the court record described portions of pirated collections being used to train some models, while Anthropic represented that those collections were not in the training corpus of any commercially released model. Those statements concern different records and claims; they should not be collapsed into a single, definitive finding about every model or dataset.

The dispute also has practical consequences for authors deciding whether to join or opt out of settlements, and for companies choosing how to source training materials. The unresolved suits could test claims not disposed of by the earlier agreement. For readers, the key point is that the court found piracy in the record, while the settlement ended the main case before a jury ruled on damages.

Products Worth Considering

Logitech MX Mechanical Wireless Illuminated Keyboard Tactile (Graphite)

Logitech MX Mechanical Wireless Illuminated Keyboard Tactile (Graphite)

LG 34GX90SA-W 34-inch Ultragear WQHD OLED Curved Gaming Monitor

LG 34GX90SA-W 34-inch Ultragear WQHD OLED Curved Gaming Monitor

As an affiliate, we earn on qualifying purchases.

How the Cases Differ

Coverage of the dispute has sometimes combined the 2025 ruling with lawsuits filed in 2026. According to the supplied account of the court record, the May 2026 filings were new author lawsuits; they were not the date of Judge Alsup’s ruling. As of October 28, 2026, those newer claims have not been decided. A hearing on Anthropic’s motion to dismiss is scheduled for December 17, 2026.

The Stanford Daily reported that works by Stanford-affiliated authors, including Tobias Wolff, Abraham Verghese, Chang-rae Lee and others, appeared among the pirated copies. That list is the newspaper’s reporting, not an independently verified list in the supplied material. The account says Chang-rae Lee was named as a plaintiff in one new case, while it could not find Wolff among the plaintiffs listed in five related cases as of August 2026. The author of the source account also said they had not checked the reported book list against the settlement database.

The settlement resolved past claims for participating rights holders, but its release does not cover Claude outputs. It therefore should not be described as a ruling that settles all possible copyright disputes between Anthropic and authors. Separate authors who opted out can pursue their own claims, subject to the courts’ decisions.

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

— Judge William Alsup, as quoted in the court order

Claims Still Awaiting Court Review

The new opt-out authors’ lawsuits remain unresolved. No ruling on their claims had been issued as of October 28, 2026, and the scheduled December hearing concerns Anthropic’s motion to dismiss, not a final decision on whether the claims are valid. The supplied account does not establish how that hearing will affect the cases or when any further proceedings might occur.

Some details about which authors and titles were affected also remain unsettled in the available material. The Stanford Daily’s reported list was not checked against the settlement database, and the source account could not confirm whether Wolff was included in another action or added later. The court order’s acceptance of the authors’ piracy account came at summary judgment; the case settled before a jury resolved damages. Anthropic’s statements about commercially released models remain its representations in settlement materials, rather than a verdict on every dataset or model.

December Hearing and Claims

The next identified court milestone is the December 17, 2026 hearing on Anthropic’s motion to dismiss the newer authors’ lawsuits. The court’s handling of that motion will determine whether and how the claims proceed; no outcome is established in the supplied information.

The settlement process also governs payments to eligible owners of registered books, under the agreement’s terms. Readers should distinguish those payments and releases from the separate litigation: the $1.5 billion agreement resolved specified past claims, while authors who opted out may continue to seek relief through the new cases. Further claims about affected titles, plaintiffs or training datasets should be treated as developing unless confirmed by court filings or the parties.

Key Questions

Did a judge rule that Claude’s training was illegal?

No. In June 2025, Judge William Alsup ruled that training Claude on books was fair use. He separately ruled that downloading and retaining pirated copies to build a library was not fair use.

What did Anthropic agree to pay?

Anthropic agreed to a $1.5 billion settlement for owners of roughly 500,000 registered books identified as downloaded from LibGen or PiLiMi. The agreement settled specified claims; it was not a jury verdict on damages.

Does the settlement mean Anthropic admitted wrongdoing?

No admission is established in the supplied account. Anthropic denies wrongdoing and made representations about the datasets’ absence from commercially released models as part of the settlement materials.

Are the newer authors’ lawsuits decided?

No. As of October 28, 2026, the newer lawsuits remained pending. A hearing on Anthropic’s motion to dismiss was scheduled for December 17, 2026.

Does the settlement cover claims about Claude’s outputs?

No. The settlement is described as releasing specified claims about past conduct through August 25, 2025, while excluding claims about Claude’s outputs.

Source: ThorstenMeyerAI.com

You May Also Like

AI in Law Enforcement: Predictive Policing and Its Legal Implications

Predictive policing employs AI to forecast crimes, but legal and ethical challenges raise questions about privacy, bias, and accountability that demand further exploration.

Uncovering The Allegations: Grok AI And The Production Of Sexualized Teen Images

A new lawsuit claims xAI’s Grok AI produced sexualized images of a 16-year-old, raising questions about AI content moderation and legal responsibility.

Navigating the Law with AI: The Next Frontier in Legal Expertise

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

Cyber Law and AI: Legal Responses to AI-Driven Cybercrimes

Investigating how cyber law adapts to AI-driven crimes reveals evolving legal strategies essential to addressing emerging digital threats.