🔍 Read the full analysis: The Claude Copyright Debate: Allegations, Proof, And IPO Implications on ThorstenMeyerAI.com
Get business pricing on monitors, keyboards and dev gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
A federal judge ruled in June 2025 that Anthropic’s use of books to train Claude was fair use, but that downloading pirated copies to build a permanent library was not. Anthropic later agreed to a $1.5 billion settlement over roughly 500,000 registered books, while new opt-out authors’ claims remain unresolved. The source says Anthropic’s IPO filing describes AI risks, but does not claim Claude is sentient or will blackmail people.
Anthropic faces new copyright lawsuits from authors who opted out of a $1.5 billion settlement over books downloaded from pirate libraries, as the company prepares for an initial public offering. The claims have not been decided: the key earlier ruling, issued in June 2025, found that training Claude on books was fair use but that downloading pirated copies for a permanent library was not.
The earlier decision came in Bartz v. Anthropic, filed by three authors in August 2024. On June 23, 2025, U.S. District Judge William Alsup distinguished between using books to train a model and obtaining books through piracy. He found the training at issue to be fair use, while ruling against Anthropic’s downloading and retention of pirated copies. The ruling was a pretrial decision; the damages question was not decided by a jury because the parties later settled.
The $1.5 billion settlement, approved in July 2026, covers roughly 500,000 registered books downloaded from Library Genesis and Pirate Library Mirror. The source describes payments in four instalments, with the last scheduled for September 2027, and says the agreement requires destruction of the downloaded collections and copies, subject to legal-hold obligations. It releases claims concerning past conduct only; claims about Claude’s outputs are not included.
Authors who opted out have brought separate actions. The source reports that a hearing on Anthropic’s motion to dismiss is set for December 17, 2026. Anthropic has denied wrongdoing and told the settlement court that the LibGen and PiLiMi books were not in the training data of any commercially released model. That statement sits alongside the 2025 court record concerning copies and training; the precise relationship between the collections, particular models and the settlement representation remains a point requiring careful distinction.
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
Copyright Risk as Anthropic Goes Public
The lawsuits matter to authors and publishers because the unresolved claims could clarify how copyright law applies to training data, not just to the acquisition of source material. The 2025 ruling drew a consequential line: training was treated as fair use in that case, while maintaining a library built from pirated downloads was not. That is a case-specific decision, not a blanket ruling that all AI training practices are lawful.
The dispute also matters to investors evaluating Anthropic’s planned public offering. Copyright claims can create financial, legal and operational uncertainty, including potential litigation costs and constraints on data practices. The source does not establish that the new cases will succeed, what damages might be sought or whether they will materially affect the IPO. Those outcomes depend on proceedings that are still underway.
Public discussion of the company’s S-1 has also blurred distinct statements about AI risk. According to the source’s review, the filing describes models exhibiting behavior “resembling blackmail” in a staged safety test, references debate about possible AI sentience and notes existential concerns raised by others. It does not state that Claude is sentient or will blackmail users. The difference matters when readers and investors assess what the filing actually discloses.
Products Worth Considering
Anker SOLIX S2000 Portable Power Station, 2,010Wh, 1,500W
Samsung 27 Odyssey OLED G5 (G50SF) QHD QD-OLED Gaming Monitor
Jackery HomePower 3000 Portable Power Station with Two 200W Solar Panels
As an affiliate, we earn on qualifying purchases.
The Ruling Behind the Headlines
Judge Alsup’s June 2025 order addressed three practices. He found that Anthropic’s use of books to train Claude was fair use, and that buying print books and scanning them was also fair use. He ruled that downloading pirated copies to retain as a permanent library was not. The order described more than seven million copies, including 196,640 books from Books3, at least five million from LibGen and at least two million from PiLiMi.
That ruling did not determine every issue after a full trial. On the pirated-copy question, the judge accepted the authors’ version of the facts at the summary-judgment stage; a jury was expected to address remaining matters, including damages. The settlement avoided that trial. It set compensation at about $3,000 per registered book before fees, according to the source, but it does not resolve the separate claims brought by authors who opted out.
The Stanford Daily reported that books by Stanford-affiliated writers, including Tobias Wolff, Abraham Verghese and Chang-rae Lee, were among the affected works. The source says it did not independently verify the paper’s book list against the settlement database. It also notes that Chang-rae Lee is named in one new lawsuit, while it could not locate Wolff among plaintiffs in five related cases listed as of August 2026. These details should be treated as reported, not as findings made by the court.
““neither the LibGen or PiLiMi datasets, nor any portions of those datasets, were in the training corpus of any of its commercially released large language models.””
— Anthropic, as described in the settlement record
Open Questions in the New Cases
No ruling has been issued on the opt-out authors’ claims, and the source reports that Anthropic’s motion to dismiss is set for a December 17, 2026 hearing. The claims, the company’s responses and any evidence developed in those cases remain subject to court proceedings.
It is also not clear from the supplied material how each newly named author’s books relate to the settlement’s registered-book list, or how the parties will address the apparent tension between the court record and Anthropic’s statement about commercially released models. The source itself says it did not check the Stanford Daily’s book list against the settlement database. No conclusion about an individual title or model should be inferred from the reported list alone.
The IPO’s timing and final terms, the potential financial effect of the litigation, and whether any settlement talks will occur are not established in the source. Viral claims about the S-1 likewise should not be treated as filing language unless confirmed against the document: the source says it found descriptions of risk scenarios, not a prediction that Claude is sentient or will blackmail people.
Court Hearing and IPO Disclosures
The next identified legal milestone is the December 17, 2026 hearing on Anthropic’s motion to dismiss the opt-out authors’ claims. The court’s decision could determine which claims proceed, but it would not by itself decide the full merits or damages. The source also lists a final settlement instalment of $450 million as due by September 27, 2027.
For investors and readers, the next points to watch are the court filings, any ruling on the dismissal motion and the company’s formal IPO documents and updates. Those records may clarify how Anthropic describes its copyright exposure and AI-related risks. Until then, the established record is narrower than some headlines suggest: a court found training fair use in the case before it, ruled that pirated-library downloads were not fair use, and the new authors’ claims remain unresolved.
Key Questions
Did a judge rule in May 2026 that Anthropic illegally trained Claude?
No. The source identifies the relevant order as a June 23, 2025 ruling in Bartz v. Anthropic. It found training to be fair use in that case but ruled that downloading pirated books for a permanent library was not. The new opt-out lawsuits have not yet been decided.
What does the $1.5 billion settlement cover?
It covers claims concerning past conduct involving roughly 500,000 registered books downloaded from LibGen and PiLiMi. The source says the agreement requires destruction of those downloaded books and copies, subject to legal-hold obligations, and does not release claims about Claude’s outputs.
Does the S-1 say Claude is sentient or will blackmail people?
According to the source’s review, no. The filing discusses behavior resembling blackmail in a staged safety test, debate about possible AI sentience and existential concerns raised by others. Those risk descriptions are not a statement that Claude is sentient or will blackmail users.
What happens next in the opt-out authors’ cases?
The source reports a hearing on Anthropic’s motion to dismiss scheduled for December 17, 2026. The hearing may address whether claims proceed; it is not a final ruling on liability or damages.
Source: ThorstenMeyerAI.com
