The Trump administration just threw its weight behind OpenAI in the most closely watched AI copyright battle yet. A 20-page brief filed in the Southern District of New York argues that training large language models on copyrighted material—including New York Times articles—qualifies as fair use, and that restricting the practice would cripple American competitiveness.
The government’s position is blunt. “The United States has a strong interest in continuing to develop a robust and competitive artificial intelligence industry,” the brief states, pointing to an executive order Trump signed last year calling for global AI leadership. The filing warns that “constraining LLM development under a misunderstanding of fair use doctrine” would hinder both scientific progress and economic mobility.
This lands at a pivotal moment. Courts have so far leaned in AI companies’ favor on training-data questions. Last year, Judge William Alsup ordered Anthropic to pay a $1.5 billion settlement to writers—but not for training on their work. The fine targeted Anthropic’s use of pirated shadow libraries. On the training question itself, Alsup was sympathetic, writing that LLMs train on works “not to race ahead and replicate or supplant them—but to turn a hard corner and create something different.”
The administration’s brief isn’t binding. The judge in the NYT case will make the actual call. But a statement of interest from the Justice Department signals to courts how the federal government views the stakes, and it arrives just as publishers were gaining momentum in arguing that training on their archives without payment constitutes infringement at scale. The fair use question—whether ingesting millions of articles to build a commercial product is “transformative”—remains unresolved. What’s now clear is which side the White House is on.
