Both sides seek a ruling without trial in AI book case

The Authors Guild, 13 lead author co-plaintiffs, OpenAI and Microsoft have all filed motions for summary judgment in the consolidated book-training copyright case, asking the court to decide it without a full trial. Publishers Weekly reported the filings, which are before U.S. District Judge Sidney Stein in the Southern District of New York.

At a glance

• Summary judgment lets a judge weigh undisputed evidence and rule without a full trial, and here both the author plaintiffs and the OpenAI defendants have asked for that outcome on the same record.
• The plaintiffs' September 5 brief argues OpenAI built its business on mass piracy, citing book content reproduced from the pirate site LibGen and then deleted, with much of the supporting detail redacted.
• OpenAI's September 4 filing calls the use highly transformative and puts ChatGPT's alleged regurgitation rate at 0.00007%, with the longest excerpt the plaintiffs' expert could generate running 1,899 words from A Game of Thrones.

The two motions put the same record to opposite uses, and a ruling on either would likely set the terms for the other book and news claims folded into the same docket. The authors' theory turns less on what ChatGPT outputs than on how the training corpus was assembled, which appears to be the ground where the Google Books and Kadrey analogies fit least comfortably.

Seven of the 13 lead co-plaintiffs have been in the case since September 2023

The Authors Guild and 17 co-plaintiff authors filed the class action against OpenAI and Microsoft in September 2023. As of October 2025 the suit was consolidated with other class actions, and its book-related components are now considered alongside a similar case brought by news organizations and led by the New York Times.

The combined docket, In Re: Open AI Inc. Copyright Infringement Litigation (1:25-md-03143), sits in the Southern District of New York. The lead co-plaintiffs now number 13, among them David Baldacci, Taylor Branch, Michael Connelly, Sylvia Day, Jonathan Franzen, Christopher Golden, Andrew Sean Greer, John Grisham, David Henry Hwang, George R.R. Martin, Jodi Picoult, Stacy Schiff and James Shapiro.

Seven of those authors, Baldacci, Connelly, Day, Franzen, Grisham, Martin and Picoult, have been parties since the original complaint. The court has sealed many letters and exhibits from public view, and the Authors Guild's own motion is heavily redacted, including the number of books at issue.

The plaintiffs' brief calls OpenAI's GPT models an existential threat to book publishing

The September 5 request opens with the assertion that "OpenAI's GPT models pose an existential threat to those who write and publish books" and that this threatens the incentives to create, the livelihoods of authors and the future of book publishing. The plaintiffs write that OpenAI "built the foundations of its business on mass piracy."

That piracy, according to the filing, included reproducing and then deleting book content found on LibGen. The plaintiffs note that courts have consistently held that obtaining "for free" something one would "ordinarily have to buy" is not fair use, and argue books are high-quality data because they contain organized writing and structured thoughts.

As evidence of harm, they cite rising sales of AI-generated fiction on platforms such as Kindle e-readers and quote OpenAI machine-learning technician Tarun Gogineni's viral 2025 tweet that if Song of Ice and Fire author George R.R. Martin "dies early, GPT-5 will autocomplete his series."

OpenAI puts ChatGPT's alleged regurgitation rate at 0.00007%

The September 4 filing from OpenAI argues the use was fair because it was "highly transformative." Training a useful LLM requires hundreds of billions to tens of trillions of words from diverse sources, the defendants write, mined for general linguistic patterns rather than creative expression.

The defendants say ChatGPT does not display copies of books and that its content is no different from excerpts or summaries available free online. They put the alleged regurgitation rate at 0.00007% and say the plaintiffs' expert used millions of prompts, producing at most a 1,899-word excerpt from A Game of Thrones, or 0.62% of the book.

OpenAI likens its position to Authors Guild v. Google (2015) and Kadrey v. Meta Platforms (2025), where fair use prevailed, and says competitive harm from new, non-infringing works is not what copyright protects against. The plaintiffs counter that the copying was nontransformative and served a competing commercial product, leaving rights holders no chance to license AI training.

October oppositions, November replies

Opposition briefs are due in early October and reply briefs in early November. A date for a decision on the motions is not named in the filings, and key material stays sealed: the number of books at issue and the deposition evidence the defendants cite on harm to the author plaintiffs are both redacted in the public versions.