On November 13, 2025, OpenAI challenged a ruling by a federal court in New York, opposing the court's order requiring OpenAI to provide over 20 million anonymized ChatGPT conversation logs to publishers including The New York Times.

The order stems from allegations by The New York Times and other publishers that OpenAI trained its AI language model using their articles without authorization, constituting copyright infringement. The publishers sought the dialogue data to investigate whether ChatGPT replicated their copyrighted content when generating responses.

Although the court stated the requested logs would undergo “de-identification” and be delivered under a legal protective order, OpenAI maintains this demand exceeds reasonable bounds. The company asserts the lawsuit cannot proceed at the expense of users' fundamental privacy rights, arguing the logs could expose substantial private information. It further claims “99.99% of the records are irrelevant to the copyright infringement allegations in this case.”