New Individual Lawsuits Arise Following the $1.5 Billion Settlement Between Anthropic and Authors

[1]

The Original Bartz v. Anthropic Settlement       

In September 2025, a U.S. District Court approved a $1.5 billion class-action settlement between generative artificial intelligence company Anthropic, which owns the Claude AI models, and authors and publishers. It is the largest known copyright settlement in the United States and first of its kind for an AI company.^[2] U.S. Senior District Judge William Alsup approved the settlement, requiring Anthropic to compensate copyright owners approximately $3,000 per work for the estimated 500,000+ books included in the settlement. ^[3]

This is a win for creative sectors stemming from Judge Alsup’s earlier ruling that drew a line against AI companies obtaining massive amounts of copyrighted content from notorious pirate sites.[4] In early June 2025, Judge Alsup had granted summary judgement in Anthropic’s favor on certain infringement claims, ruling that Anthropic acquired certain copyrighted books legally and that its’ copying of the lawfully acquired books to train its AI models was protected by the Fair Use Doctrine.[5] The Fair Use Doctrine is a legal principle allowing limited use of copyrighted material without permission for criticism, comment, news reporting, teaching, scholarship, or research.^[6] Judge Alsup found that Anthropic’s copying of the lawfully acquired books for purposes of “training” its models was “exceedingly transformative,” which was the primary factor for his finding of Fair Use.^[7] However, Judge Alsup also ruled that Anthropic’s copying of pirated books from illegal websites like Library Genesis (“LibGen") to create a library for unspecified future uses, was not Fair Use and that plaintiffs could pursue that claim at trial.[8] The later ruling exposed Anthropic to the possibility of catastrophic damages given the substantial number of works involved and the possibility of a large statutory damage award for each work.[9] Moreover, Judge Alsup intimated when Anthropic downloaded the books from third-party websites, it had “knowledge” that the website contained pirated books,[10] which could have increased the damages, based on the possible finding of “willful” copyright infringement. These rulings led to the aforementioned settlement.[11]

This case sets the tone for future copyright infringement cases and sends an important message that the Fair Use Doctrine does not grant generative AI companies blanket permission to exploit copyrighted content no matter the source or means of acquisition.[12] Justin A. Nelson, one of the lawyers for the authors in the suit shared with the NY Times, “This settlement sends a powerful message to A.I. companies and creators alike that taking copyrighted works from these pirate websites is wrong.” ^[13]

Individual Authors Fight Back in Brand New Suit Against AI Companies

Despite the big win against companies who use pirated books to train their AI models, multiple authors including John Carreyrou, author of Bad Blood: Secrets and Lies in a Silicon Valley Startup, opted out of the original settlement against Anthropic.[14] As a result, following the September settlement, Carreyrou and five other authors filed a new individual copyright infringement suit, against Anthropic, OpenAI, Google, Meta, xAI, and Perplexity AI in late December 2025; Carreyrou et al. v. Anthropic PBC et al.^[15] Similar to the class-action Bartz case settled by Judge Aslup, the authors in this suit argue that the companies copied the research in their books from pirated libraries like LibGen for their AI models without permission or compensation. ^[16] The complaint alleges that the actions by the companies was willful because they were “repeatedly told that using” pirated copies of literary works to train their AI models “was unlawful,” but the companies continued to use pirated copies.[17] However, the plaintiffs ultimately decided not to bring a class action suit, arguing the Copyright Act entitles the authors to individualized statutory damages from each of the companies for their alleged infringements.[18] Each alleged infringement is incredibly important because some of the defendant companies have admitted works are “especially valuable training material” because of the authors' copyrightable expression through “narrative structure” and other elements.[19] Ultimately, the authors assert that $3,000 is not enough compensation given the $150,000 ceiling for statutory damages for willful infringement in the Copyright Act and seek to be rewarded with larger damages.^[20] The authors of the Carreyrou settlement are not the only group speaking against settlement, for example, ClaimsHero, a consumer justice platform, created a website for the Anthropic settlement, encouraging authors to opt out and file their own individual suits demanding more.[21] Judge Alsup strongly suggested ignoring the website, claiming it was fraudulent and misleading.^[22]

The case will be heard in the U.S. District Court for the Northern District of California by presiding Magistrate Judge Thomas S. Hixson.[23] The authors have demanded a jury, but the case is still in its early stages.[24] Individual lawsuits would be costly for both the plaintiffs and defendants, but it could mark another big win for authors globally if they are able to receive a larger damages award from the generative AI companies who have pirated their work to further their AI models.

References

[1] Photo by Viralyft, A Man Sitting In Front of a Laptop, (Apr. 25, 2023), https://unsplash.com/photos/a-man-sitting-in-front-of-a-laptop-computer-2LI1Rl5b_P4.

^[2] See Chloe Veltman, Anthropic Settles with Authors in First-of-its-Kind AI Copyright Infringement Lawsuit, NPR, (Sep. 5, 2025), https://www.npr.org/2025/09/05/nx-s1-5529404/anthropic-settlement-authors-copyright-ai.

^[3] See id.

[4] See id.

[5] See Paul Keller & Airina Rodrigues, The Anthropic Copyright Settlement: Dissecting the Anatomy of a Landmark AI Case, Brownstein (Sep. 9, 2025), https://www.bhfs.com/insight/the-anthropic-copyright-settlement-dissecting-the-anatomy-of-a-landmark-ai-case/.

^[6] See 17 U.S.C. § 207.

^[7] See id.

[8] See id.

[9] See id.

[10] See id.

[11] See id.

[12] See id.

^[13] See id.

[14] See Ed Nawotka, Authors File New Lawsuit Against AI Companies Seeking More Money, Publisher’s Weekly (Dec. 23, 2025), https://www.publishersweekly.com/pw/by-topic/industry-news/publisher-news/article/99347-authors-file-new-lawsuit-against-ai-companies-seeking-more-money.html.

^[15]See id.

^[16] See id.

[17] See Compl. ¶ 7, Carreyrou et al. v. Anthropic PBC et al., No. 3:25-cv-10897, (N.D. Cal. Dec. 22, 2025).

[18] See Hailey Konnath, Anthropic, Google, Meta Face More Writer Copyright Claims, Law360 (Dec. 23, 2025), https://www.law360.com/articles/2424983; see also Compl. ¶ 13, supra note 17.

[19] See Complaint ¶ 53, supra note 17.

^[20] See Nawotka, supra note 14; see also Compl. ¶ 14, supra note 17.

[21] See Nawotka, supra note 14.

^[22] See id.

[23] See Compl. ¶ 136, supra note 17.

[24] See Konnath, supra note 18.