OpenAI slams court order that lets NYT read 20 million complete user chats

TL;DR

OpenAI is seeking to overturn a court order to produce 20 million ChatGPT user chats to The New York Times.
The company argues that producing entire conversations is too broad and poses privacy risks, as most chats are unrelated to the lawsuit.
OpenAI previously offered 20 million chats as a counter-proposal but objects to the court-mandated production.
The New York Times claims the chat data is necessary to hold OpenAI accountable for copyright infringement.
Concerns about ChatGPT data privacy have been heightened by its appearance in Google search results.
A judge ordered the production of de-identified chats, stating that existing protective orders and anonymization adequately address privacy concerns.
OpenAI contends that de-identification does not remove all potentially private information.
The AI company believes the court's reliance on a previous case (*Concord Music Group, Inc. v. Anthropic PBC*) is misplaced due to differences in the nature of the data.
OpenAI plans to develop advanced security features, including client-side encryption, to enhance user privacy.
The New York Times stated OpenAI's user communication is misleading and omits facts, emphasizing that no user privacy is at risk.

Continue reading the original article