OpenAI, Microsoft, The New York Times and a group of authors led by John Grisham and George R.R. Martin filed competing briefs Friday in Manhattan federal court, each asking a judge to rule in their favor on the fair use defense that will likely decide the future of AI training under copyright law. The filings frame a direct collision between the tech industry's claim that training on copyrighted works is lawful and copyright holders' argument that it threatens their livelihoods.
Judge Sidney Stein must now weigh the arguments in the consolidated litigation, which began with lawsuits filed in 2023. The Times accuses the companies of using millions of its articles without permission to train ChatGPT. The authors' group separately sued the same year over the use of their books. Dozens of similar complaints are pending across the country, all turning on the same core question: whether AI systems make fair use of copyrighted material by creating new content from it.
A split bench on what counts as fair use
The first two federal judges to address the issue reached the same conclusion through different reasoning. U.S. District Judge William Alsup in San Francisco called Anthropic's use of books for AI training "quintessentially transformative." Two days later, Judge Vince Chhabria ruled for Meta in a similar case and also labeled the use of copyrighted books transformative.
Chhabria added a warning, however. He said AI training would not be fair use "in many circumstances" and raised concerns that generative AI could "flood the market" with content that competes with human creators. That distinction now sits at the center of Friday's arguments.
Copyright holders cite market harm
The authors echoed Chhabria's concern directly. In their brief, they argued that AI is "diluting the market for books across the board" and that fair use "does not countenance such catastrophic threats to the incentive to create." The news outlets took a parallel position, telling the court that ChatGPT diverts users from their websites and displaces the market for their work. "Defendants' competitive exploitation of Plaintiffs' expressive works cannot be excused" under court precedent, the outlets said.
These arguments lean heavily on the fourth fair use factor - the effect of the use on the potential market for the copyrighted work. For legal professionals tracking Generative AI and LLM litigation, market harm has emerged as the most contested element across these cases.
Tech companies push back
OpenAI cited Alsup's ruling and told the Manhattan court that using copyrighted work for AI training is "among the most transformative many of us will see in our lifetimes." The company said its technology has not harmed writers. "The purpose of OpenAI's pretraining process was to derive broad, unprotectable statistical patterns related to language that can be used to create new text, not reproduce protected expression," OpenAI said in its brief addressing the authors' case.
Microsoft took a sharper tone, calling the plaintiffs' fears unfounded. "Plaintiffs began this case speculating that LLMs would destroy their very livelihoods," the company said. "Years of discovery later, the record is clear: Neither LLM training nor the use of LLMs in products substitutes for copyrighted books."
Why this matters for legal professionals
Judge Stein's eventual ruling will either solidify or fracture the early consensus that AI training qualifies as fair use. The outcome will shape licensing negotiations, litigation strategy, and compliance frameworks across the AI for Legal field. With discovery now complete and summary judgment motions filed, the case is moving toward a decision that could set binding precedent in the Second Circuit - or tee up the issue for Supreme Court review.
Your membership also unlocks:
