The Expert RecordThe standard for expert witnesses using AI

AI in the Courtroom · Expert witnesses and AI

Villanueva v. Las Vegas Metropolitan Police Department

The plaintiff's expert in an excessive-force case used ChatGPT to research and draft his report and disclosed it only at deposition. The district judge ordered the ChatGPT case-law summaries produced as facts or data considered, and held his prompts, a deposition outline and draft-report output protected.

Primary document readChatGPT case summaries ordered produced; prompts, outline and draft output protected September 29, 2026

What happened

Jose Villanueva sued the Las Vegas Metropolitan Police Department and an officer over his treatment while detained at the Clark County Detention Center, bringing a civil rights claim under 42 U.S.C. section 1983 and state-law claims. His expert witness, Tom Melton, was retained to give opinions on the use of force.

At his deposition on October 1, 2025, Melton testified that he had relied on ChatGPT to help research and draft his expert report. That use had not been disclosed in response to the defendants' earlier subpoena for "all data, documents, and other information considered by expert witness in forming the expert's opinions." When the defendants asked for the ChatGPT history afterward, the plaintiff declined and listed it on his privilege log as attorney and expert work product.

The plaintiff described the roughly 40-page log in three categories. Much of it asked ChatGPT for summaries of controlling excessive-force cases. Some of it asked for a deposition outline, which the expert apparently used to predict the questions he would face. The rest consisted of prompts to analyze and draft what could become portions of the report, such as the qualifications of the opposing expert. Melton testified that he sometimes copied and pasted ChatGPT output into his own report.

What the court did

On December 23, 2025, Magistrate Judge Albregts granted the defendants' motion to compel from the bench and ordered the whole log produced. He reasoned that the log should have been disclosed under Rule 26(a)(2)(B) because the expert considered it in forming his opinions, and that it was neither a draft report nor privileged. He did not review the log in camera.

On September 29, 2026, District Judge Anne R. Traum granted the plaintiff's objection in part and overruled it in part, after conducting her own in camera review. She divided the log rather than treating it as one document:

  • Produced: ChatGPT's case-law summaries. The order states that "The legal case summaries generated by ChatGPT constitute 'facts or data' that Melton considered and must be disclosed under Rule 26," and notes that the log is the only place the defendants can review any hallucinated citations the expert considered.
  • Protected: the expert's own questions and prompts, which the court found are not facts or data and would reveal his mental impressions; the deposition outline; and the output used to draft the report, which the court treated as protected draft material under Rule 26(b)(4)(B) "regardless of the form in which the draft is recorded."

The court also addressed waiver: "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 Melton's use of the program." It added that the defendants' concerns about the expert's methodology and reliance on ChatGPT are appropriate to raise at trial as impeachment.

Where it stands

The ruling resolves the discovery dispute in the district court. The same order overruled two unrelated objections concerning punitive damages discovery. Motions for summary judgment filed by the defendants in July 2026 remained pending as of the order. Whether the case summaries have since been produced, and whether they contain fabricated authority, was not located as of October 8, 2026.

Villanueva v. LVMPD reaches a different line from the one drawn in Conservation Law Foundation v. Shell, where a magistrate judge treated an expert team's prompts as part of a disclosable methodology. In Villanueva the prompts were protected and the outputs the expert relied on for legal substance were not. Both orders turn on what the expert did with the system, which is the question an expert's own records answer best.

The obligation it engaged

Clause 8 of the Standard, disclosure, is the first clause engaged. The ChatGPT use surfaced only when the expert was asked at deposition, after a subpoena that called for everything he considered. Clause 7, contemporaneous record, is what would have let the expert separate the categories the court separated: research, deposition preparation and drafting.

Clause 3, independent authority, explains why the case summaries were ordered produced. A summary of a court decision is a characterization of a source, and the Standard requires that it be confirmed against the decision itself before anything rests on it. Clause 2, non-delegation, bears on the copied passages: the report's analysis has to be the expert's own, whatever the court ultimately says about whether the draft output is discoverable.

Sources

What could not be confirmed

The magistrate judge's bench ruling is described from the district judge's order; the hearing transcript (ECF No. 69) was not opened. The magistrate judge's first name is not given in the order. Whether the case summaries have been produced, and what they contain, was not located as of 2026-10-08.

About this entry

This entry was checked against the primary document listed in the sources. The entry for Villanueva v. Las Vegas Metropolitan Police Department records conduct and what the tribunal did. It does not characterize a holding and it is not legal advice. Last checked October 8, 2026. More matters like this one are collected under expert witnesses and AI. If something here is out of date, write to [email protected].

Top