Magistrate Judge Ona T. Wang of the U.S. District Court for the Southern District of New York is set to hold an in-person conference tomorrow in the ongoing New York Times Co. v. OpenAI copyright lawsuit. This confidential settlement conference will address a controversial court order requiring OpenAI to preserve all chat outputs indefinitely.
The order impacts approximately 400 million ChatGPT users, as it involves retaining data from both temporary chats and those marked for deletion by users. The preservation aims to identify potential inclusion of copyrighted material from the New York Times within these chats.
Concerns have been raised regarding user privacy, as many have made specific choices about their data, expecting deletions to be honored. Users' private conversations on topics ranging from health to personal matters could be retained unnecessarily, posing risks of misuse or abuse.
OpenAI has suggested a process to sample chat logs, comparing those not slated for deletion with others, to assess whether retention is necessary. The upcoming conference seeks agreement on this sampling approach.
The court and involved parties are encouraged to balance discovery needs with privacy concerns. If the preservation order continues in its current form, they must establish a protective order ensuring robust notice to users, tight security measures, limited use, and storage duration for these records.
The Center for Democracy & Technology (CDT) has argued against broad mandatory data retention by digital service providers due to user risk concerns—a stance supported by Congress's refusal to pass related legislation even in criminal contexts.