May 13, 2026 – Artificial intelligence (AI) is transforming the way lawyers practice law. It is also changing the way clients approach their lawyers and their legal issues.
How does a client’s use of AI affect lawyer-client privilege and the work product doctrine? On Feb. 17, 2026, two judges in two different jurisdictions came to two different conclusions.
What’s a lawyer to do?
Question
We’ve modified our fee agreement to include information about our firm’s use of AI and have developed firm policies regarding the responsible use of such tools.
But now we’ve hit upon a new issue related to AI: our client’s use of it. We’ve gotten a few submissions from clients that are obviously AI generated. Besides the fees incurred for us having to read it all, we’re concerned how those missives might be characterized in litigation down the road.
What’s the current landscape regarding this topic?
Answer
Neither attorney-client privilege nor the work product doctrine are ethics rules. Rather, they are evidentiary rules.
So why am I talking about them in an ethics column?
Because a lawyer has a duty to keep abreast of the risks and benefits associated with changes in technology,[1] and has a duty to explain a matter to the extent reasonably necessary to permit the client to make informed decisions regarding the representation.[2]
Given these ethical duties, it is imperative that a lawyer understand the technology their clients are using and the potential consequences of their use of that technology.
Recently, two courts came to two different conclusions about what the use of AI means for purposes of attorney-client privilege and the work product doctrine. The opinions happened to be issued on the same day.
Neither case is a Wisconsin case, and so neither is binding, but given the rapid increase in the use of AI, the same issues are likely to land before a court in Wisconsin at some point.
United States v. Heppner: Not Protected
In the first case,
United States v. Heppner,[3] a court ruled that written exchanges between a criminal defendant and a generative AI platform were not protected by attorney-client privilege or the work product doctrine.
In this case, a criminal defendant used Claude, Anthropic's generative AI platform, to prepare an outline of potential defenses and legal arguments, which he shared with his counsel. His counsel had not directed him to run the searches.
In reaching his decision that the GenAI documents were not covered by attorney-client privilege, the judge applied the usual three-part test for attorney-client privilege, that the communication must be:
between privileged parties;
intended to be and actually kept confidential; and
made for the purpose of obtaining or providing legal advice.
The court found that the client/defendant did not have an attorney-client relationship with Claude and that the communications were between two nonlawyers, not between a client and a lawyer. The court further found that privilege depends on a “trusting human relationship.”
In finding that the communications with Claude were not confidential, the court noted Anthropic’s privacy policy, which states that Claude’s inputs and outputs could be retained, used to train the platform, and disclosed to third parties.
Finally, the court ruled that because Claude expressly disclaims that it can provide legal advice, the communication could not have been for the purpose of obtaining legal advice. The court also rejected the notion that the materials were protected by the work product doctrine, noting that the doctrine protects only those materials prepared by counsel or at the direction of counsel.
Warner v. Gilbarco: AI are Tools
The same day
Heppner was decided, a second case dealing with similar issues was handed down.
In that case,
Warner v. Gilbarco, Inc.,[4] the plaintiff was representing herself. The defendant sought production of “all documents and information concerning [the plaintiff’s] use of third-party AI tools in connection with this lawsuit.”
The judge denied the request, finding that Rule 26(b)(3)(a) of the Federal Rules of Civil Procedure – the work product doctrine rule – protects documents “prepared in anticipation of litigation or for trial by another party or its representative.”
The judge rejected the defendant’s argument that the use of ChatGPT constituted a waiver of work product protection. In doing so, he drew a distinction, absent from the analysis in
Heppner, between attorney-client privilege waiver and work product waiver.
The court explained that “while the mere showing of a voluntary disclosure to a third person will generally suffice to show waiver of the attorney-client privilege, it should not suffice in itself for waiver of the work product privilege.” Work product waiver requires disclosure to an adversary or disclosure in a way likely to reach an adversary’s hands.
Further, the court held that AI is not a third person: “ChatGPT (and other generative AI programs) are tools, not persons, even if they may have administrators somewhere in the background.” Thus, the defendants could not get to the waiver question because there was no “person” to disclose the information to.
Two Competing Frameworks
In a blog article about the decisions, Jennifer Ellis, a lawyer who provides technology and AI consulting for lawyers and law firms, succinctly summarizes the position we are left in after the
Heppner and
Warner decisions:
What we have right now are two competing frameworks: one that treats AI platforms as third parties whose terms of service can destroy confidentiality, and another that treats them as tools, ‘not persons,’ through which a litigant processes their own mental impressions.[5]
It is unlikely that courts will agree upon a framework by which to analyze these situations anytime soon, although a lawyer should always be aware of any relevant precedent in their jurisdiction.
Given the unsettled landscape, lawyers would be wise to talk to their clients about which tools the clients are using, the purpose for which they are using the tools, the confidentiality afforded by those tools, and how the use of AI might weaken or defeat attorney-client and work product protections down the road.
Endnotes
[1] SCR 20:1.1 Comment [8] states, “To maintain the requisite knowledge and skill, a lawyer should keep abreast of changes in the law and its practice, including the benefits and risks associated with relevant technology, engage in continuing study and education and comply with all continuing legal education requirements to which the lawyer is subject.” ^
[2] SCR 20:1.4(b).
^
[3]
United States v. Heppner, No. 25 Cr. 503 (S.D.N.Y.)
^
[4]
Warner v. Gilbarco, Inc., No. 2:24-cv-12333, (E.D. Mich.)
^
[5] Jennifer Ellis, “Two Answers: When Does Using AI Waive Privilege?” JLE Blog, Feb. 21, 2026.
^
