The U.S. government has officially taken OpenAI’s side in a significant copyright lawsuit brought by The New York Times concerning the use of copyrighted materials to train large language models (LLMs). In a 20-page legal brief submitted by the Trump administration, the government emphasized the national interest in fostering a competitive AI industry that leads globally. The brief highlighted the importance of AI innovation for U.S. economic growth, arguing that restricting training on copyrighted content misinterprets fair use doctrine and could hinder progress in AI development.

The legal dispute centers on whether AI companies, including OpenAI, can lawfully use copyrighted books, articles, and other published works without permission to train their language models. Many publishers have challenged this practice as copyright infringement. Key to the debate is the concept of fair use, which allows certain uses of copyrighted materials without consent if deemed transformative. The government’s brief argues that training AI models on copyrighted data is a highly transformative process akin to reading for creative innovation, supporting OpenAI’s defense in this murky area of law.

This development comes amid a growing number of legal cases and public scrutiny of AI training practices. While some rulings, such as a $1.5 billion settlement involving Anthropic last year, penalized companies for pirating copyrighted works, courts have not disallowed the act of training AI models on such materials. The government’s support for OpenAI underscores the desire to maintain U.S. leadership in the AI sector despite ongoing copyright tensions, although the case itself is still under trial in the Southern District of New York.

Though the brief is not a judicial ruling, it is a notable intervention and could influence the court’s considerations. It reflects broader governmental efforts, including an executive order signed by President Trump in 2025, to build America’s AI capabilities without overly constraining innovation through restrictive copyright interpretations. The outcome of this case may set precedent on how AI companies can source training data and navigate intellectual property rights moving forward.