AI
Trump Administration Files Amicus Brief Supporting OpenAI in New York Times Copyright Case
The federal government has thrown its weight behind AI companies, arguing that training large language models on copyrighted material constitutes fair use under U.S. law.
WIRED — AI
The Trump administration has moved to bolster OpenAI's legal position in its ongoing copyright dispute with the New York Times, submitting a federal brief that frames AI training on published works as a legitimate fair use practice rather than infringement.
Government's Stance on AI Training#
In a letter filed this week, the Justice Department asserted that the nation's competitive position in artificial intelligence hinges on permitting companies to train models on copyrighted content without explicit permission. The administration characterized such training as "extraordinarily transformative," a key legal standard in fair use doctrine. Federal lawyers contended that restricting large language model development would damage U.S. technological progress and global standing.
The government further argued that OpenAI's models do not directly compete with Times journalism, another factor courts weigh when assessing fair use claims. The administration also drew an analogy to human learning, suggesting that prohibiting AI systems from training on copyrighted text would be as illogical as forbidding writers from studying other authors' work—noting the example of a young Joan Didion studying Ernest Hemingway's prose.
Industry and Creator Reactions#
The move has polarized stakeholders. A New York Times spokesperson countered that the brief disregards the intellectual property rights of creators whose work underpins AI systems. The Authors Guild, which has filed its own lawsuit against OpenAI, called the filing "replete with faulty arguments" about copyright law.
Legal experts suggest the brief carries considerable weight despite being non-binding. An intellectual property attorney noted that Judge Sidney H. Stein, overseeing the case, would "almost certainly" take the government's position seriously given its institutional authority.
The Broader Copyright Battle#
This intervention arrives amid a sprawling legal landscape. Earlier rulings have produced mixed outcomes: Meta prevailed in one case, though the judge left open the possibility that unauthorized training on copyrighted materials could be illegal in different circumstances. Anthropic, by contrast, lost its case and agreed to pay authors $1.5 billion in damages—the largest copyright settlement in U.S. history—though the court did find the AI training itself constituted fair use.
