Loading...
Loading...
A U.S. federal court ruling in US v. Heppner (S.D.N.Y., 2026) is forcing a rethink of how lawyers and clients use AI chatbots. Judge Jed Rakoff ordered disclosure of dozens of documents created through Anthropic’s Claude in a securities-fraud case, rejecting claims that the exchanges were protected by attorney-client privilege. The decision is prompting law firms to warn clients that chatbot conversations can become discoverable evidence and that feeding attorney communications into AI tools may waive privilege. Firms are updating policies, contracts, and workflows to limit AI use or route it through controlled, vetted systems.
US v. Heppner (S.D.N.Y. 2026) no attorney-client privilege for AI chats [pdf]
A U.S. federal judge ordered a defendant to hand over documents generated by Anthropic’s Claude, ruling that communications with AI platforms don’t create attorney-client privilege and can be sought by prosecutors. The case raises questions about whether using cloud AI services or AI-enabled productivity tools (Google Docs, Gmail drafts, medical bots) can waive legal protections, since sharing privileged information with third parties typically undermines privilege. Lawyers warn this ruling could expose sensitive chatlogs in criminal and civil discovery and complicate use of SaaS AI features; some suggest local models for private interactions. The decision highlights an urgent need for clearer rules on AI, privacy, and privilege in legal and regulatory contexts.
A federal judge ruled this year that AI chat logs are not covered by attorney-client privilege, prompting U.S. lawyers to warn clients against treating chatbots as confidential advisors. The ruling arose in the securities-fraud case of former GWG Holdings chair Bradley Heppner, where Judge Jed Rakoff ordered disclosure of 31 documents generated with Anthropic’s Claude after prosecutors argued AI exchanges weren’t protected. Major law firms have begun advising clients and adding contract clauses cautioning that sharing lawyer communications with AI or using chatbots to draft materials can waive privilege. The development matters because it reshapes evidence risks in criminal and civil cases and forces legal teams to alter workflows and policies around AI use.
US v. Heppner (S.D.N.Y. 2026) no attorney-client privilege for AI chats [pdf]