The Justice Department urged a Manhattan federal court to rule that OpenAI did not violate copyright law when it trained artificial intelligence systems on articles from The New York Times and other publishers. The filing, submitted late Tuesday, marks the first time the department has intervened in the escalating legal battle over how AI companies use copyrighted material.
The government argued that developing AI is critical to national security and that the training process sufficiently transforms written works into new material protected under copyright law. The benefits of AI, the filing said, "far outweigh any competitive harm."
The government's national security argument
Stanley Woodward Jr., the associate attorney general for the Justice Department, called the filing "a historic statement of interest" in a social media post. He added that President Trump had made clear that "AI dominance is critical to promote national security, prosperity, and economic mobility for all Americans."
The Justice Department's position ties AI development directly to American tech dominance. By framing the issue as a matter of national interest rather than a narrow copyright dispute, the government has raised the stakes in a case that could set binding precedent for the entire generative AI and LLM industry.
The Times pushes back
The New York Times sued OpenAI and Microsoft in 2023, alleging the companies used its journalism without permission to build systems that now compete with the publisher. Graham James, a spokesman for The Times, said the Justice Department was siding with a handful of "trillion-dollar AI companies" at the expense of American creators.
The case is one of several lawsuits testing whether AI training on copyrighted works falls under fair use. A ruling against OpenAI could force fundamental changes to how models are built, potentially requiring licensing deals with every publisher whose content appears in training data.
Why this matters for IT and development professionals
For developers and IT teams building on top of large language models, the outcome of this case will shape what training data is legally available. A ruling that restricts unlicensed training could shrink the pool of high-quality text used to improve models, affecting the performance of tools integrated into enterprise workflows. It may also accelerate the shift toward synthetic data and carefully curated training sets - changes that will land directly on engineering roadmaps. Professionals working in AI for IT and development should watch how courts define transformation in the context of model training, because the legal definition will determine what data pipelines look like in production.
Your membership also unlocks:
