Trump Administration Backs OpenAI in Landmark AI Copyright Fight
The Trump administration backs OpenAI's fair use defense in The New York Times copyright case, arguing limits on AI training could weaken US leadership.
Summary
On September 2, 2026, the Trump administration filed a 20-page brief supporting OpenAI in The New York Times’ copyright lawsuit in the US District Court for the Southern District of New York. It defends using copyrighted books, articles, and other media without permission to train large language models powering ChatGPT, Anthropic’s Claude, and Google’s Gemini. The government argues that transformative training can qualify as fair use and that restricting it would impede creative and scientific progress, prosperity, economic mobility, and US leadership in artificial intelligence, echoing President Donald Trump’s 2025 executive order.
Publishers contend that unauthorized training infringes copyright, but cases have largely favored AI companies. In 2025, Judge William Alsup ordered Anthropic to pay writers a $1.5 billion settlement because it obtained training books from illegal shadow libraries, not because training itself was unlawful. Alsup compared model training to a reader absorbing books before creating something different. The administration’s brief is nonbinding because its authors lack jurisdiction, but it could influence the court’s unresolved decision on whether OpenAI’s training is sufficiently transformative for fair use.
Positives
- The 20-page federal brief gives OpenAI direct government support for treating transformative AI training as fair use.
- AI copyright cases have largely favored model developers when training itself, rather than illegal acquisition, is challenged.
- Judge William Alsup’s 2025 Anthropic decision distinguished model training from piracy and compared it with a reader creating something different.
- The Trump administration argues permissive training rules support creative progress, scientific advancement, prosperity, economic mobility, and US AI leadership.
Risks & concerns
- The New York Times alleges OpenAI used copyrighted books, articles, and other media for training without permission.
- Anthropic’s use of illegal shadow libraries produced a $1.5 billion settlement for writers in 2025.
- Fair use remains unresolved because courts must decide whether AI training is sufficiently transformative.
- The administration’s brief is not a ruling and has no jurisdictional force in the Southern District of New York case.