The Trump administration has stepped into a major copyright case involving OpenAI. The move could shape how courts view AI training going forward.
The New York Times sued OpenAI over the use of copyrighted articles to train ChatGPT. The newspaper argues the company used its content without permission.
On September 2, 2026, the U.S. government filed a 20-page brief supporting OpenAI's position. The filing was submitted in the U.S. District Court for the Southern District of New York.
The brief argues that the United States needs a strong and competitive AI industry. It states that limiting AI training under a misread of fair use law could slow American progress.
The filing references an executive order signed by President Trump last year. That order calls for removing barriers to AI leadership in the United States.
What the Brief Says
The core legal question is whether training AI models on copyrighted material counts as fair use. Fair use is a part of copyright law that allows limited use of protected works without permission.
Courts often look at whether the use is "transformative" enough to be legal. The government's brief argues that AI training meets that standard.
The brief states that treating AI training as copyright infringement would hurt economic growth. It also claims restrictions would slow scientific progress in the country.
This filing is not a ruling. The authors of the brief do not have the power to decide the case themselves.
Legal experts say government briefs like this one can still shape how a judge thinks about a case, even without formal authority.
Past Rulings on AI and Copyright
Most copyright cases involving AI training have so far favored AI companies. Judges have generally accepted the argument that training counts as fair use.
Anthropic faced a separate lawsuit from a group of writers over its training data. That case did not result in a fine for the training process itself.
Instead, Anthropic was ordered to pay 1.5 billion dollars for using illegal shadow libraries to obtain books. Judge William Alsup approved that settlement last year.
In his ruling, Judge Alsup compared AI training to a person reading books before writing their own work. He said the AI did not copy the material but used it to create something new.
The New York Times case against OpenAI also includes claims that the company hid evidence during the trial. The newspaper raised this accusation in July 2026.
OpenAI has not issued a new public statement following the government's filing. The case is expected to continue in the coming months.
The outcome could affect how other AI companies handle copyrighted training data. Several similar lawsuits are still working through U.S. courts.
