Judge Orders OpenAI to Hand Over 20 Million ChatGPT Logs in Copyright Case

A U.S. magistrate judge has ordered OpenAI to provide 20 million anonymized ChatGPT chat records to plaintiffs in a major copyright lawsuit filed by The New York Times and other media organizations. The order, issued by Magistrate Judge Ona Wang, requires the company to submit the logs by Friday and states that user privacy will be protected through anonymization and existing court safeguards.

OpenAI has asked the court to reconsider the decision, arguing that producing the data would expose personal user conversations and violate privacy commitments. The company stated that nearly all of the requested material—about 99.99%—is unrelated to the plaintiffs’ claims. It warned that the ruling could allow future litigants to demand access to vast amounts of sensitive user information without relevance to their cases.

The media plaintiffs contend that access to the chat logs is necessary to determine whether ChatGPT replicated their copyrighted material and to challenge OpenAI’s claims that the evidence against it was fabricated. They maintain that the anonymized data will remain protected under the court’s confidentiality order.

The order is part of a broader legal dispute in which news organizations accuse OpenAI and its partners of using their content to train AI models without authorization. The case represents one of several ongoing lawsuits worldwide addressing the intersection of copyright law and generative AI technology.