OpenAI Slams Court Order that Lets NYT Read 20 Million Complete User Chats

Why It Matters

The ruling will set a precedent for how far courts can compel AI companies to disclose user interactions, shaping privacy protections and discovery practices in future AI‑related litigation and influencing public confidence in generative‑AI services.

Summary

OpenAI petitioned a U.S. District Court in New York to overturn a November order that compels the company to provide 20 million de‑identified ChatGPT conversation logs to the New York Times and other news plaintiffs in a copyright lawsuit. The firm contends the logs, which contain full multi‑turn exchanges, are far broader than the prompt‑output pairs originally requested, that over 99.99% are irrelevant, and that the disclosure threatens user privacy and sets a risky discovery precedent. OpenAI offered a targeted sampling method and highlighted upcoming client‑side encryption, but the magistrate judge affirmed the production deadline, leaving the dispute over wholesale data disclosure unresolved.