The United States told a federal court that training AI models on copyrighted text is fair use. The Justice Department filed the statement on 1 September 2026 in the consolidated OpenAI copyright case in Manhattan. It argues the copying at the training stage is highly transformative under the first fair-use factor. The government says The New York Times is asking the court to narrow fair use too far. It warns that requiring paid licences would favour the largest technology companies and legacy publishers. It also links a strong domestic AI industry to national security and cites two presidential executive orders. The filing draws a line at the output stage, where reproducing a work may not be fair use. It takes no position on whether a licensing system would be workable. The court has not ruled.

• filed 1 September 2026
• 4 statutory fair-use factors
• case 25-md-3143