AI + Privilege

Attorney-client privilege and work-product protections in the AI-age: a moving target?

In a recent case, an expert used ChatGPT to assist with his research--this came out at a deposition and the opposing party then demanded production of the expert's ChatGPT history (prompts and outputs).

At first, the magistrate judge overseeing the discovery dispute ordered the production of the expert's entire ChatGPT history. The Article III judge reversed, noting:

"1) questions and prompts to ChatGPT, 2) use of the program to create a deposition outline, and 3) creation of a draft report are all protected from disclosure." Villanueva v. Las Vegas Metro. Police Dept., Case No. 24-CV-125, 2026 WL 2942348, 3 (D. Nev. Sept. 29, 2026).

The interesting part, to me, is this: "Because ChatGPT is not a person, but was an application used in this instance as a research, drafting, and editing program, the Court finds that any privileged information is not waived by [the expert's] use of the program." Id. at 5.

This hints at where I think most courts will land--like they did with using email, Westlaw, and other electronic tools now ubiquitous in the law: tools used as an extension of a lawyer (or expert) to aid in the litigation (or practice of law) do not break the privilege or work-product chain.

This makes sense.

It encourages the use of technology to support litigation (and trial development) and creates some protections against that technology being owned and run by third-parties not automatically waiving privilege and work product protections.

Could you still waive privilege via these tools? Sure. Just not automatically. I hope more courts embrace this concept.

♥️✌🏻🔥