Generative AI and Attorney-Client Privilege: Caution on Waiver
The Rise of AI in Everyday Life
Artificial intelligence is now far from a futuristic concept confined to science fiction or research labs. For many, generative AI has now become woven into the fabric of their daily lives. People interact with AI multiple times a day, frequently without even realizing it, and its uses range from spam filters quietly protecting your inbox, to providing answers to countless questions that come across people’s minds. However, for those that utilize public AI platforms to provide answers to legal questions or generate documents, users should beware of the unintended consequences.
The Heppner Case: What Happened
In a decision with significant implications for any non-lawyer who uses artificial intelligence tools to research or analyze legal matters, the United States District Court for the Southern District of New York ruled on February 10, 2026, that documents generated through a public AI platform were not protected by the attorney-client privilege or the work product doctrine. Specifically, the court in United States of America v. Heppner, 25-cr-00503-JSR, granted the government’s motion to access documents that defendant Bradley Heppner created using the AI tool Claude before his arrest on federal fraud charges.
Heppner is charged with a variety of financial crimes. During the execution of a search warrant at Heppner’s residence, federal agents seized electronic devices containing documents that Heppner had generated using Claude relating to the Government’s investigation. Heppner asserted that these documents were privileged since he shared them with his attorney and because some of the information Heppner input into Claude was information he had learned from his attorney. However, the government filed a motion seeking a determination from the court that such documents were not privileged and were subject to government inspection, since such documents were not communications between Heppner and his lawyer, were not created for the purpose of obtaining legal advice from counsel, and were not confidential since they were created by a publicly accessible AI platform. The government also argued that the documents were not shielded by the work-product doctrine, because Heppner created the AI documents on his own initiative and not at the request of his attorney.
The Court’s Ruling on Privilege
The court held that these documents were not protected from disclosure to the government on the basis of either attorney-client privilege or the work-product doctrine. First, the court held that attorney-client privilege did not protect these documents from disclosure because (1) they were not communications between Heppner and his attorney since Claude was not an attorney, (2) the communications in the documents were not confidential since Claude is a public AI platform wherein the user accepts that any data that is input into Claude is subject to disclosure to third-parties, and (3) Heppner did not communicate with Claude for the purpose of obtaining legal advice, since the Claude platform expressly disclaims the provision of legal advice. Moreover, the court held that the documents were not protected by the work-product doctrine because (1) they were not prepared by or at the request of Heppner’s attorney, and (2) they did not reflect Heppner’s counsel’s strategy.
Key Takeaways for AI Users
This decision provides several important lessons to keep in mind when using these tools moving forward. First, communications with public AI platforms may not satisfy the requirements of attorney-client privilege because AI tools are not attorneys, inputs to them are not confidential, and most of them expressly disclaim that they provide legal advice. Second, taking the ruling a step further, it is not a stretch to conclude that sharing confidential attorney-client communications with a public AI tool can waive any privilege that would otherwise apply to those communications. While this decision is not binding on other courts, the analysis conducted by the Southern District of New York is not surprising and can easily be followed in other courts.
Clients should carefully consider these implications before using public AI platforms to analyze legal issues. While AI has become an undeniable part of our daily lives, there are serious risks associated with asking Claude, ChatGPT or a public equivalent to resolve your legal problem or provide legal guidance. Our attorneys here at Strauss Troy are monitoring the developments in this area of the law and here to help our clients and address any questions they have. If you would like to discuss any of these topics, please contact either of the authors.
Stephen S. Schmidt: ssschmidt@strausstroy.com or 513.629.9422
Andrew D. White: adwhite@strausstroy.com or 513.629.9466
