The Trump administration filed a statement of interest in The New York Times Co. v. OpenAI Inc., arguing that using copyrighted material to train large language models qualifies as fair use under copyright law. It’s the first time the federal government has taken a formal position in any of the major AI copyright cases, and it lands squarely on the side of the AI companies.

“The United States has a strong interest in continuing to develop a robust and competitive artificial intelligence industry that sets the standard for the practice and procedure of AI use globally,” the brief reads. The government isn’t a party to the case, but its position carries weight. Courts don’t have to follow it, but they often do.

The Times sued OpenAI and Microsoft in December 2023, alleging that the companies scraped millions of Times articles without permission to train GPT models. The newspaper says this constitutes copyright infringement. OpenAI has argued its training practices fall under fair use, the legal doctrine that allows limited use of copyrighted material without permission for purposes like criticism, commentary, and transformative works.

What fair use actually means

Fair use isn’t a bright-line rule. Courts weigh four factors: the purpose and character of the use, the nature of the copyrighted work, the amount used, and the effect on the market for the original. The Trump administration’s brief emphasizes the first factor, arguing that training AI models is “transformative” because the output doesn’t substitute for the original articles.

That’s a stretch. The Times has shown examples of ChatGPT reproducing Times articles nearly verbatim when prompted. If the model can spit out the original work on demand, it’s hard to argue the use is transformative. The administration’s brief doesn’t address this problem directly.

The government also claims that blocking AI training on copyrighted material would “stifle innovation” and harm US competitiveness. That’s a policy argument, not a legal one. Congress writes copyright law to balance incentives for creators with public access to knowledge. Courts interpret that balance. They don’t rewrite it based on which industry might benefit more.

Who this actually affects

The administration’s position matters most for the dozens of other AI copyright cases working their way through federal courts. Authors, visual artists, and music publishers have all sued AI companies over training data. If courts adopt the government’s reasoning, most of those cases fail at summary judgment.

It also signals how the Trump administration plans to approach AI regulation more broadly. The brief reads like an industrial policy document. It treats AI development as a national priority that justifies weakening intellectual property protections. That’s a choice, but it’s not the only choice. The EU is taking the opposite approach, requiring AI companies to disclose their training data and respect opt-outs.

The Times case is still in discovery. OpenAI has fought to keep its training data confidential, arguing it’s a trade secret. The court recently ordered the company to produce more information about what went into GPT-4. If it turns out OpenAI systematically scraped paywalled content or bypassed technical barriers, the fair use argument gets weaker.

What happens next

The next major milestone is OpenAI’s motion to dismiss, which the court will likely rule on in the next few months. If the motion fails, the case goes to trial. That won’t happen until late 2027 at the earliest.

In the meantime, expect more government briefs in other AI copyright cases. The administration has made its position clear. Whether courts agree is another question. Fair use is supposed to be decided case-by-case, not by blanket policy pronouncements.

The real problem is that copyright law wasn’t written with AI in mind. It’s a bad fit. Training a model on millions of works doesn’t look like traditional copying, but it doesn’t look like traditional fair use either. Congress could clarify the rules, but it hasn’t. So courts are stuck trying to apply 18th-century legal concepts to 21st-century technology, and the results are predictably messy.

For now, the government has put its thumb on the scale. The Times will argue that’s irrelevant. The court will decide who’s right.