About

Courts Rule AI Chat Logs Are Discoverable and Must Be Preserved

Published
Score
10

Why it matters

Courts are treating AI chatbot conversations as ordinary electronically stored information subject to discovery and preservation obligations. Recent judicial decisions make clear that prompts and outputs from ChatGPT, Claude, and similar platforms are not a protected category—they must be preserved once litigation is reasonably anticipated, and deletion or loss can trigger sanctions or adverse inferences. The legal framework is straightforward: existing civil discovery rules apply. AI chats do not become privileged simply because they are later shared with counsel.

The scope of who must comply is broad. Litigants, companies, employees, and AI platform providers like OpenAI now face preservation duties. One federal court has already issued a preservation order requiring OpenAI to retain output logs in an active case, signaling how courts intend to operationalize these requirements. The precise contours of what must be preserved—and for how long—remain unsettled in many jurisdictions.

Practitioners should act now. Organizations need to inventory which AI tools employees use, update litigation holds to explicitly include AI chat histories, and coordinate with IT and vendors on log retention and retrievability. The risk is immediate: as generative AI becomes embedded in workplace workflows, the volume of potentially discoverable AI communications will only grow. Companies that fail to preserve these records face real consequences in pending or foreseeable disputes.

mail Subscribe to Law And Technology email updates

Primary sources. No fluff. Straight to your inbox.

Also on LawSnap