New York Times Pushes Back on Claim That Using Articles for AI Training Is 'Fair Use'
The U.S. government has sided with OpenAI, which used articles from The New York Times (NYT) to train its artificial intelligence (AI). The U.S. Department of Justice has determined that training a “large language model (LLM)”—the foundation of AI services such as ChatGPT—on newspaper articles qualifies as “fair use,” a doctrine that permits the use of copyrighted works without the rights holder’s permission. According to reports on the 2nd (local time), the Justice Department filed a statement of interest with the U.S. District Court for the Southern District of New York the previous day, urging the court to reject the argument that using NYT articles and other content for AI training violates copyright law. Even when it is not a party to a case, the U.S. government may submit a statement of interest laying out its position in litigation involving the public interest.
The lawsuit traces back to 2023. The NYT sued OpenAI and Microsoft (MS) for copyright infringement, alleging that the companies used millions of its articles without permission to train AI models such as ChatGPT and Copilot. The newspaper is seeking damages as well as a court order halting the unauthorized use of its content for AI training.
At the heart of the Justice Department’s argument is “transformativeness.” It contends that using vast bodies of copyrighted works as training material for LLMs is a “highly transformative” act, serving a purpose distinct from delivering the original works to readers. Fair use is a copyright law principle that allows, as an exception, the use of works without the rights holder’s consent, depending on factors such as the purpose and manner of use and the impact on the market for the original work. The department added that the act of copying works to train a model is separate from the model’s act of producing output for users.
At the end of its argument, the Justice Department invoked national security, framing the development of the AI industry as a matter directly tied to U.S. national security. Deputy Attorney General Stanley Woodward said on social media, “This administration will never allow the United States to be placed at a disadvantage against foreign adversaries because of a plainly wrong interpretation of copyright law.”
The Justice Department also took a dim view of proposals to require AI companies to pay licensing fees for news content, arguing that steep copyright costs would reduce competition in the LLM market and make it harder for smaller AI firms with less capital to enter.
The NYT pushed back immediately. NYT spokesperson Graham James said, “The government is siding with a handful of AI companies worth trillions of dollars at the expense of countless American creators whose work has been misappropriated.” He emphasized that “AI and creators can both thrive,” adding, “AI companies simply need to pay fair compensation for the content that makes their products possible.”
This is not the first time the Justice Department has filed such a statement. In June, it submitted a brief to a Tennessee court seeking the dismissal of a lawsuit filed by the NAACP and environmental groups over the construction of data centers for xAI, Elon Musk’s AI company.
