Internal documents and depositions from the ongoing lawsuit between The New York Times and OpenAI suggest the AI developer had tools to track copyrighted material before the case began. This contradicts earlier claims by the company that it could not effectively monitor training data.
Discovery revealed a database containing approximately 78 million ChatGPT conversations monitored for copyright compliance. The New York Times has filed motions for sanctions, citing these findings to argue that OpenAI’s previous statements regarding its tracking capabilities were misleading.
The lawsuit, originally filed in December 2023, alleges unauthorized use of articles published between 1928 and 2024 to train AI models. As of September 2026, about 160 regional publications have joined the legal action against OpenAI and Microsoft.
The case is currently at the summary judgment stage before Judge Sidney H. Stein in the Southern District of New York. The US Department of Justice has submitted a brief supporting the fair use defense of OpenAI, while publishers continue to push for licensing agreements.
