The U.S. Department of Justice filed a statement of interest on Tuesday backing OpenAI's position that training large language models on copyrighted text constitutes fair use under copyright law, marking what Reuters reports is the first time the federal government has directly weighed in on the wave of copyright lawsuits facing the AI industry10,12.
The 20-page brief, filed in the New York Times' lawsuit against OpenAI and Microsoft, argues that "the creative possibilities and public benefits" of training LLMs on copyrighted content "far outweigh any competitive harm". The Justice Department called AI training "exceedingly transformative"9 and warned that ruling in the Times' favor would "thwart creative and scientific progress while hindering American prosperity and economic mobility".
The case dates to December 2023, when the New York Times sued OpenAI and Microsoft in Manhattan federal court, alleging that OpenAI used millions of copyrighted articles without permission to train its AI models and that ChatGPT sometimes reproduces long portions of the newspaper's articles verbatim11. The lawsuit seeks "billions of dollars" in damages from both companies15. Other publishers, including MediaNews Group-owned newspapers the New York Daily News and the Chicago Tribune, digital media publisher Ziff Davis, and the nonprofit Center for Investigative Reporting, have joined the suit.
The DOJ brief frames the question in national security terms. Justice Department attorneys wrote that AI models "help national security officials analyze and draw inferences from (unprotected) real-world facts such as those conveyed in New York Times articles" and that the "creative benefits often overlap with national security interests". The brief also warns that restricting AI training in the United States "could give foreign rivals an advantage" and that "foreign adversaries won't be held back by copyright law".
The DOJ further argued that a ruling against OpenAI "would hamper competition in the market for LLMs, because only the largest technology companies might have the capital necessary to pay licensing fees". U.S. Associate Attorney General Stanley Woodward said "this Administration will never let the nation be at a disadvantage relative to foreign adversaries based on a plainly incorrect understanding of copyright law".
The brief draws a distinction between training and output, stating that "whatever legal questions certain output uses might raise should not bear on the transformative nature of the training use".
The New York Times pushed back. A spokesperson said "the Administration is siding with a handful of trillion-dollar AI companies at the expense of countless American creators whose work they stole". The spokesperson argued that AI companies "need to pay fairly for the content that makes their products possible, as copyright law requires" and that allowing companies to take content "without permission or compensation would undermine the sustainability of human-created content".
The Times has framed its position as a middle ground in which AI companies can continue developing models while compensating creators whose work they use for training.
ANALYSIS The DOJ's intervention carries no binding legal force as a statement of interest, but it places the executive branch's weight squarely behind the fair-use defense that OpenAI and other AI companies have advanced across multiple lawsuits. The national-security framing and the competition argument against mandatory licensing fees align the government's position with the broadest possible reading of fair use for AI training, while explicitly deferring the question of liability for model outputs.