Are AI Conversations Protected Like Lawyer Conversations?

A New York federal court recently ruled that conversations with an artificial intelligence (“AI”) platform are not protected by attorney-client privilege or the work product doctrine. United States v. Heppner, 25 Cr. 503; 2026 WL 436479 (S.D.N.Y. Feb. 17, 2026).

Attorney-client privilege normally protects confidential discussions between a client and their lawyer. It allows people to speak openly with their attorney without worrying that those communications could later be disclosed in court.

“Related to but distinct from the attorney-client privilege” is the work product doctrine, which “at its core … shelters the mental processes of the attorney, providing a privileged area within which he can analyze and prepare his client’s case.” Id. at *3. “The doctrine “provides qualified protection for materials prepared by or at the behest of counsel in anticipation of litigation or for trial.” Id.

In Heppner, a defendant used an AI tool to help think through a legal strategy after learning he was under investigation. He later shared those AI-generated notes with his lawyer and argued that they should be protected.

The court disagreed for three main reasons:

1. AI is not a lawyer. Privilege only applies to communications between a client and their attorney. Talking to an AI platform does not create that relationship.
2. The conversations were not confidential. The AI platform’s privacy policy allowed it to collect and potentially share user data, meaning the user could not reasonably expect the conversation to remain private.
3. The AI was not providing legal advice. The platform clearly states that it does not give legal advice and encourages users to consult a lawyer.

Because of these factors, the court ruled that communications were not protected by attorney-client privilege or the work product doctrine, meaning they could be used as evidence.

But closer to home, another federal court reached a different result in Warner v. Gilbarco, Inc., No. 2:24-cv-12333, 2026 WL 373043 (E.D. Mich. Feb. 10, 2026). In Warner, the defendants made a discovery request for all information concerning any use of AI by the plaintiff. The court characterized this request as “a fishing expedition,” finding that defendants sought to “compel Plaintiff’s internal analysis and mental impressions – i.e., her thought process –rather than any existing document or evidence.” Id. at *4. Although work product privilege can be waived, this plaintiff’s use of ChatGPT didn’t have that effect because “the work-product waiver has to be a waiver to an adversary or in a way likely to get in an adversary’s hand.” Id. (emphasis in original). “ChatGPT (and other generative AI programs) are tools, not persons, even if they may have administrators somewhere in the background.” Id. (emphasis in original).

The Warner opinion doesn’t address the attorney-client privilege because the plaintiff in that case was representing herself.

What This Means for Clients

These decisions highlight an important point: information shared with public AI tools might not be confidential. If you are dealing with a legal issue, it is safest to discuss the matter directly with your attorney rather than using AI tools to analyze or draft a legal strategy.

Even though the plaintiff in Warner seems to have dodged this bullet, neither Warner nor Heppner have yet been subject to appellate review.  Also, both of these cases were decided under the federal rules of evidence. A Michigan state court might reach a different result in Michigan’s rules of evidence.

We will continue to follow both Warner and Heppner as courts across the country wrestle with these emerging issues.