OpenAI rallies librarians, startups and Washington for its fair-use defense against the Times

OpenAI posted excerpts from amicus briefs filed in support of its position in The New York Times’ copyright lawsuit, arguing that training AI models on copyrighted works is fair use. The filers it highlights include library and disability groups, the Authors Alliance, startup advocates, law professors, independent creators, a Pulitzer Prize-winning journalist and the United States government, whose brief argues the Times’ position would severely hamper progress (our coverage of the government’s filing).

This page is advocacy. A party to the lawsuit selected the friendliest passages from briefs on its side, and OpenAI accuses the Times of chasing an undeserved payday. That’s a litigation posture, not a neutral summary. The full filings are on the public court docket, and the Times and its allies have arguments that don’t appear here, including internal documents in which OpenAI and Microsoft described a “doom loop” for the web (unsealed filings).

The library and accessibility coalition makes a point that often gets lost: restricting training doesn’t only hit big labs. It also hits medical researchers, educators, people with disabilities and anyone relying on open-source models built by others. Public Knowledge connects AI training to fair-use cases that protected image search, Google Books and accessible digital books for readers with print disabilities. If training continues that lineage, a sweeping loss for AI developers could entrench the biggest players, who can afford licenses, and freeze out everyone else.

Fair use under 17 U.S.C. § 107 weighs four factors, including the purpose of the use and its effect on the market for the original. Courts in the Second Circuit, where Google Books was decided, have treated highly transformative uses favorably. Several briefs argue that training produces intermediate copies that aren’t similar to outputs and don’t substitute for the originals.

The Authors Alliance asks the court to reject the idea that aggregate competition from noninfringing AI output counts as market harm under the fourth factor. Copyright, it argues, protects against substitution for your protected expression, not against new work competing for the same audience. Doctrinally, that’s strong. Practically, it should worry anyone who writes for a living. If AI output floods the same market and erodes the value of original reporting, and the law says that doesn’t count, newsrooms take real economic damage with no remedy. Wikipedia’s operator has proposed a middle path: training can be fair use, but serving readers a substitute for the original shouldn’t be (its filing).

The Foundation for American Innovation suggests national-security value should weigh in fair-use analysis. That’s an expansive idea and a dangerous precedent: copyright analysis that bends toward whatever is strategically useful to the state isn’t a principle to apply to every future technology.

The case for training as transformative use is strong, and AI that learns broadly from public knowledge is worth protecting. An honest version of OpenAI’s argument would also admit that journalism is under real economic pressure from these systems, and that a court win doesn’t settle whether the people producing the underlying work can survive. The likeliest lasting outcome is a legal win for training paired with growing pressure for licensing deals and output-side limits on regurgitation. Other publishers are pressing similar suits (Seattle Times and Newsday). More in our OpenAI coverage and regulation coverage.