On February 17, 2026, a federal judge from the Southern District of New York held that an AI-chat made by a criminal defendant in anticipation of a meeting with his attorneys was not protected by either attorney-client privilege or the work product doctrine. This ruling—which the court acknowledged answers “a question of first impression nationwide”—has serious implications for anyone using AI tools in the context of legal matters. Understanding whether privilege attaches or whether you might waive a privilege is critical to identifying your evidence at a trial. A more fact-dependent analysis — and careful consideration of the role of AI within the attorney-client relationship — supports that AI search evidence may sometimes qualify for privilege.
In United States v. Heppner, Heppner used an artificial intelligence application called, “Claude,” in anticipation of a potential indictment. Without direction from counsel, Hepner input information he had learned from counsel into Claude and “prepared reports that outlined defense strategy [and] what he might argue with respect to the facts and the law.” He delivered Claude’s outputs to counsel, likely thinking that he was helping counsel with merely hypothetical arguments. Those outputs, in turn, influenced counsel’s strategy “going forward.” Heppner had used Claude for the “express purpose of talking to counsel.”
Key implications of this ruling include:
No Privilege: Communications with AI may be treated as conversations with a third party, not a confidential advisor.
Discoverable Evidence: Prompts and outputs from publicly available AI platforms (like ChatGPT or Claude) can be subjected to discovery in lawsuits.
Waiver Risk: Using AI to analyze legal exposure or draft strategies can waive attorney-client privilege.
It is critical that an attorney know the extent to which a client may have made admissions to certain facts at issue, and the attorney should be made aware of the precise AI activity:
Disclaimer: The caselaw above is from the Southern District of New York. While Federal law was applied, the case is not directly binding on Indiana Courts. The case signals a direction in how trial courts will treat AI conversations.