Trump Administration Backs OpenAI’s Fair-Use Defense in New York Times Case

The Trump administration has intervened in The New York Times’ copyright suit against OpenAI, filing a 20-page brief that supports the company’s argument that training large language models on copyrighted works can qualify as fair use. The government has cast the dispute as a test of whether the United States can sustain a competitive AI industry, not solely as a conflict over a publisher’s material.

The brief was filed in the U.S. District Court for the Southern District of New York, where the Times’ case against OpenAI is pending. It defends OpenAI’s unlicensed use of copyrighted material for language-model training and argues that constraining that work through an incorrect application of fair-use doctrine would damage U.S. creative, scientific and economic progress.

A copyright defense recast as industrial policy

Fair use is a copyright exception that can permit use of protected work without permission in particular circumstances. Here, the central question is whether using works to train a model is sufficiently transformative: whether the training produces something different rather than simply reproducing the original material. OpenAI has advanced that position, and the government’s filing supports it.

The administration’s rationale reaches beyond the parties. It says the United States has an interest in a robust, competitive AI sector and frames the ability to ingest copyrighted text as important to scientific advancement, economic development and national security. The filing has been described as the first formal U.S. government intervention in an AI copyright case involving a major publisher.

Constraining LLM development under a misunderstanding of fair use doctrine would thwart such creative and scientific progress while hindering American prosperity and economic mobility.

U.S. government brief

What the intervention cannot settle

The filing is not a judicial decision, and its authors have no jurisdiction over the case. Its practical significance is therefore persuasive rather than binding: it gives the court the federal government’s view of fair use and the policy consequences it sees in a restrictive interpretation.

Three questions the brief leaves with the court

• Whether OpenAI’s training use qualifies as fair use under the facts of this case remains for the court to decide.
• Whether the Times can prove its allegation that millions of its copyrighted articles were used without permission remains part of the underlying 2023 lawsuit against OpenAI and Microsoft.
• The Times’ July sanctions motion, which alleges misconduct in discovery and data handling, is a separate dispute within the litigation.

Earlier cases show a narrower fault line

The government enters a legal landscape where the treatment of model training and the source of training data can diverge. Judge William Alsup ordered Anthropic to pay a $1.5 billion copyright settlement to writers, but the penalty concerned the company’s use of illegal shadow libraries, not the act of training its models. One account of earlier AI copyright litigation characterizes those cases as generally favorable to AI companies on training.

That distinction does not answer the Times case. The company’s allegations concern the use of its articles, while the government’s position centers on the legal character and national consequences of training. The court now has both arguments in front of it, alongside a formal federal view that an adverse interpretation of fair use could impede U.S. AI development.