Court decision on AI shows the pitfalls of using it in the legal context without proper safeguards
Most of the cases recently regarding use of AI either discuss attorneys (even at “big” firms) using it carelessly and not double checking AI output for “hallucinations” (fake case cites usually) or for privacy issues or even suits with allegations that AI encouraged a teen suicide. But this case was different and stood out to me as I know a number of clients that use AI regularly and likely in the legal context. In a recent federal court ruling in February 2026 in New York in US v. Bradley Heppner a federal court held in a published decision that a non-attorneys use of an AI chatbot to prepare reports to outline a defense strategy against the US government, which was prosecuting him for securities fraud wire fraud and making false statements, was NOT protected by the attorney client privilege.
In connection with Heppner’s arrest agents from the FBI executed a search warrant at his home and seized numerous documents and electronic devices. There were numerous documents that memorialized communications that he had with the generative AI platform operated by a private company. Mr. Heppner's counsel did not instruct his client to go to the AI platform to create a defense strategy but his counsel later argued after the search warrant was executed, that Heppner was preparing these reports in anticipation of a potential indictment. The government of course then moved for an order allowing them to examine all of the AI documents arguing that they were not protected by the attorney-client privilege nor the attorney work product doctrine.
The court held the attorney-client privilege “attaches to, and protects from disclosure, communications (1) between a client and his or her attorney (2) that are intended to be, and in fact were, kept confidential (3) for the purpose of obtaining or providing legal advice." Courts construe the attorney-client privilege narrowly because it operates as an exception to the rule that "all relevant proof is essential".
Applying these principles, the AI Documents lacked at least two, if not all three, elements of the attorney-client privilege.
First, the AI documents were not communications between Heppner and his counsel. Heppner did not, and indeed could not, maintain that the AI chatbot was an attorney.
Second, the communications memorialized in the AI Documents were not confidential. This is not merely because Heppner communicated with a third-party AI platform but also because the written privacy policy to which users of the AI platform consent provides that the AI company collects data on both users' "inputs" and the AI "outputs," that it uses such data to "train" the AI, and that the company reserves the right to disclose such data to a host of "third parties,". Essentially AI users do not have substantial privacy interests in their "conversations with [another publicly accessible AI platform] which users voluntarily disclosed" to the platform. For these reasons, Heppner could have had no "reasonable expectation of confidentiality in his communications" with the AI.
Third, Heppner did not communicate with AI for the purpose of obtaining legal advice. This issue perhaps presented a closer call because Heppner's counsel asserted that Heppner communicated with AI for the "express purpose of talking to counsel." Heppner's counsel also conceded, Heppner did not do so at the suggestion or direction of counsel. Moreover, even assuming that Heppner intended to share these communications with his counsel and eventually did so, it is black-letter law that non-privileged communications are not somehow alchemically changed into privileged ones upon being shared with counsel.
Finally, the AI Documents do not merit protection under the work product doctrine because, even assuming, arguendo, that they were prepared "in anticipation of litigation," they were nevertheless not "prepared by or at the behest of counsel," nor did they reflect defense counsel's strategy. Heppner was not acting as his counsel's agent when he communicated with AI. Because the AI documents were not prepared at the behest of counsel and did not disclose counsel's strategy, they did not merit protection as work product.
So the takeaway from all this, especially in potential civil litigation, is to let your counsel interface (if at all) with AI about your case, unless of course you are willing to loose all claims of attorney client privilege over your most intimate thoughts about the legal viability of your case or defense.