OpenAI fights order to hand over 20 million private ChatGPT conversations

OpenAI is seeking judicial intervention to overturn a recent ruling that mandates the company to release 20 million user conversations to The New York Times and other plaintiffs in a copyright infringement lawsuit. The tech firm argues that while it previously offered a portion of these conversations as a compromise, the court's order is excessively broad. In court documents filed with the US District Court for the Southern District of New York, OpenAI emphasized the nature of the logs in question. "These logs encompass entire conversations, with each entry representing a full exchange of multiple prompts and responses between a user and ChatGPT," the company stated. They further explained that revealing such comprehensive logs poses a greater risk of exposing sensitive information compared to disclosing isolated prompt-output pairs. OpenAI contends that an overwhelming majority of the 20 million chats—over 99.99%—are irrelevant to the ongoing case. The firm has requested the court to annul the order and compel the news plaintiffs to engage with OpenAI’s suggestion for pinpointing pertinent logs. Additionally, OpenAI may pursue an appeal in a federal appellate court. In a communication to its users, OpenAI highlighted that The New York Times is insisting on access to these private conversations in an effort to find instances where users attempted to bypass their paywall using ChatGPT. This situation raises privacy concerns for users, especially as ChatGPT interactions have been discovered in Google search results and the Google Search Console, a tool for tracking web traffic. To address these privacy issues, OpenAI has announced plans to implement enhanced security measures, including client-side encryption for user messages, aimed at safeguarding user data more effectively.