OpenAI is resisting a demand from the New York Times for 20 million private ChatGPT conversations, calling it a user privacy invasion.
The request is part of the ongoing copyright lawsuit the Times filed against OpenAI. OpenAI has publicly framed its resistance as protecting users, not protecting itself in litigation, and says it is accelerating new security and privacy protections in response. The company provided no specifics about what those protections involve or when they would roll out.
The framing matters more than the filing. By positioning this as a fight for users rather than a routine objection to broad discovery, OpenAI gets to wear the good-guy badge while still pursuing its legal interest in keeping those conversations out of evidence. The 20 million figure, if accurate, suggests the Times is casting a wide net, not targeting specific interactions.
The same company that asks the public to trust it with sensitive daily queries is now arguing in court that disclosing those queries would harm users. Both things can be true. The convenience of that argument, though, is hard to ignore.