Judge: Claude chats not privileged in Heppner case

The Heppner ruling is one piece of a broader pattern in which courts are being asked to decide whether conversations with AI tools deserve the same protections as traditional legal communications. Judge Rakoff relied heavily on Anthropic's consumer privacy policy, which allows data collection, training on inputs and outputs, and disclosure to third parties including regulators. That undercut any reasonable expectation that the chats were confidential.

At the same time, the law is not settled. On the same day as the Heppner ruling, a Michigan magistrate in Warner v. Gilbarco ruled that a self-represented plaintiff's ChatGPT chats were protected work product, calling AI tools "tools, not persons." A Colorado court reached a similar result in Morgan v. V2X in March. Both of those cases involved people representing themselves, which is a different situation from Heppner's.

The practical lesson is that a privacy policy is a promise about what a vendor intends to do, not a promise about what a court will allow. If a copy of a conversation exists on a provider's servers, it can potentially be preserved and handed over as evidence.