US Government Sides With OpenAI Against the New York Times: The DOJ's Fair-Use Bombshell

For the first time in the wave of AI copyright litigation that has swept American courtrooms since 2023, the United States government has picked a side — and it is not the side of the newspapers. In a statement of interest filed in Manhattan federal court on Tuesday, the Justice Department backed OpenAI’s fair-use defense against the New York Times and other publishers, arguing that training large language models on copyrighted texts generally does not violate copyright law. The filing, first reported Wednesday, is the government’s first intervention in any of the dozens of cases brought by authors, publishers, music labels, and news outlets over AI training data.

What the brief actually says

At the core of the filing is a single sentence with sweeping consequences: “The United States has a strong interest in this court rejecting any argument that training LLMs on copyrighted texts violates copyright law” based on concerns including scientific advancement and national security.

The government’s argument runs on three tracks. First, it frames AI training as “extraordinarily” transformative — adopting the very language tech companies have used since the earliest fair-use defenses were sketched. Beyond the subject matter of the litigation, the brief argues, “LLMs are already helping researchers across fields achieve major breakthroughs,” and “constraining LLM development under a misunderstanding of fair use doctrine would thwart such creative and scientific progress while hindering American prosperity and economic mobility.”

Second, the brief leans hard on geopolitics. “AI dominance is critical to promote national security, prosperity, and economic mobility for all Americans,” U.S. Associate Attorney General Stanley Woodward said in a statement posted to X announcing the filing. “This Administration will never let our Nation be at a disadvantage relative to our foreign adversaries based on a plainly incorrect understanding of copyright law.” The message to the court is unmistakable: in the administration’s view, a ruling against OpenAI would be a ruling against American AI leadership.

Third, the government is timing its entry carefully. A statement of interest carries advisory rather than binding legal weight — but it arrives as the Times case, first filed in December 2023, moves toward a landmark trial, and as dozens of parallel suits against Anthropic, Meta, and other labs wait on the sidelines for signals about how the first judges to reach the merits will rule. The first two judges to consider the fair-use question issued diverging rulings last year, leaving the doctrine unsettled and the stakes enormous.

The case it lands in

The Times’ lawsuit accuses OpenAI and its largest financial backer, Microsoft, of using millions of newspaper articles without permission to train the models behind ChatGPT. It has grown into one of the nastiest discovery fights in tech litigation: publishers asked a judge in July to sanction OpenAI over allegations it “chose obstruction” rather than produce training datasets and ChatGPT logs, and OpenAI has separately lost battles to keep anonymized user chat logs out of discovery. Against that combative backdrop, the government’s brief reads less like neutral legal analysis and more like a thumb on the scale.

Notably, the same day the brief surfaced, Commerce Secretary Howard Lutnick told G20 officials meeting in North Carolina that their countries should embrace fair use and let AI companies train on creators’ work — while finding ways to “protect artists.” The synchronized messaging suggests a coordinated administration push to export the fair-use position internationally, not merely to influence one courtroom.

Why it matters

Three consequences stand out.

Precedent leverage. Even advisory, a DOJ brief gives every defendant in the pending copyright wave — Anthropic, Meta, Mistral, and the rest — a citable government position supporting their core defense. Judges weigh the government’s view on matters of public interest, and the DOJ chose the highest-profile case in the country as its vehicle.

The national-security reframing. Copyright litigation has traditionally turned on the four fair-use factors: the purpose and character of the use, the nature of the work, the amount taken, and the effect on the market. The government’s brief imports an entirely new vocabulary — foreign adversaries, economic mobility, scientific progress — that could pull courts toward a policy-weighted analysis rather than a doctrinal one.

Chilling effect on licensing. Publishers have bet that litigation pressure would force labs into licensing deals; some, like the Associated Press and Axel Springer, signed them. A government-endorsed fair-use right to train would remove much of that leverage, strengthening the hand of holdouts who calculate that fighting is cheaper than licensing.

The other side of the argument

Critics, including the publishers’ legal teams, counter that transformative purpose does not excuse free-riding, and that the market-effect factor — ChatGPT answers substituting for newspaper visits — cuts against fair use no matter how innovative the training process is. They also note the irony of an administration professing to defend creators against Big Tech while backing the largest AI company in the world against the newspaper of record. The Times and OpenAI both declined to comment on the filing Wednesday.

What happens next is procedural: the publishers will respond, Judge Sidney Stein will weigh the statement alongside the actual motions, and the case inches toward a trial that could define the legal boundaries of AI training for a generation. The government has now made clear which outcome it prefers — and how far it is willing to go to get it.