TL;DR: A Nevada federal district court’s September 29, 2026 order in Villanueva v. Las Vegas Metropolitan Police Department holds that AI-generated case summaries relied on by an expert are discoverable under Rule 26, while the expert’s AI prompts and deposition outlines remain protected as draft materials or attorney work product. The decision, issued by Judge Anne R. Traum, signals a growing split among courts on how to treat AI inputs versus outputs in expert discovery. For trial teams, the takeaway is to implement clear protocols for AI use, preserve materials with careful labeling, and anticipate targeted scrutiny of AI-influenced portions of expert opinions.
What happened and why it matters
In Villanueva v. Las Vegas Metropolitan Police Department, a civil rights action under 42 U.S.C. 1983 filed in the District of Nevada, the court confronted a central question in AI-assisted expert practice: when an expert uses a Generative AI platform to assist in researching, drafting, or organizing a report, which components are subject to discovery? On September 29, 2026, District Judge Anne R. Traum issued an order addressing discovery of the expert’s AI materials (ECF No. 112). The key holding was nuanced:
- AI-generated case summaries that the expert relied upon in forming opinions are discoverable as “facts or data” under Rule 26, requiring production of those AI-generated outputs.
- The expert’s questions and prompts to the AI platform themselves are protected, as they reveal the expert’s mental impressions or trial strategy and fall within draft material or attorney work product protections.
- AI-generated deposition outlines used for preparation are also protected, while the AI-generated content fed into the final opinion may still be subject to scrutiny if it functioned as input to the expert’s conclusions.
The court’s careful delineation—outputs that inform an opinion versus inputs that shape the process—illustrates the current complexity of AI discovery in federal practice. The Nevada decision is one of the clearest post-Conservation Law Foundation line items to date that attempts to separate AI prompts from AI-assisted outputs for privilege and discovery purposes. The ruling was reported in depth by legal outlets and summarized by practitioners, noting that two federal courts have issued divergent approaches to the discoverability of AI prompts and AI-produced materials. (docs.justia.com)
For trial teams, the practical upshot is immediate: if an expert has used AI to generate case summaries, citations, or other substantive input that informed the conclusions, those AI outputs may be fair game in discovery. In contrast, the prompts, search queries, and deposition questions that the expert used to drive the AI research remain protected, at least in this jurisdiction, as materials that reveal strategy or mental impressions. Counsel must therefore prepare to justify privilege or draft-protection for inputs while ensuring that crucial AI-generated outputs relied upon by the expert are available for cross-examination and impeachment if needed. (mayerbrown.com)
Practical implications for working litigators
- Discovery planning and preservation: Counsel should issue clear preservation directives for AI-related materials at the Rule 26(f) stage. Separate the categories of AI outputs (summaries, analyses, synthesized documents) from inputs (prompts, queries, deposition outlines) in retention and privilege logs. This helps align with the court’s approach in Villanueva and reduces the risk of inadvertent waivers.
- Expert retention agreements: Include explicit language about AI use, specifying which AI-generated outputs are treated as potential evidence and which inputs remain privileged. Consider stipulations that address chain-of-custody for AI-produced materials, version control, and the ability to impeach with AI-derived content if necessary.
- Deposition strategy: Anticipate deposing experts about the extent of their AI reliance, the nature of the AI outputs used, and how those outputs influenced their opinions. Use targeted questions to reveal whether AI-generated case summaries or other inputs fed into the opinion. Prepare to challenge or defend the discoverability of specific outputs consistent with the Villanueva framework.
- Cross-examination and impeachment: Be ready to examine the reliability and provenance of AI-generated summaries, including potential “hallucinations” or incorrect citations. The district court in Villanueva emphasized that even protected AI drafts may be scrutinized at trial, and the use of AI does not entirely shield a party from impeachment through AI-generated content. (mayerbrown.com)
How to implement best practices in light of Villanueva
- Create an AI-use protocol for experts: Document every step of AI engagement, including the tools used, prompts entered, the types of outputs produced, and how those outputs contributed to the final opinions. This fosters transparency and helps with privilege analysis during discovery.
- Separate inputs from outputs in disclosures: When preparing expert reports, clearly distinguish between AI-generated content that informed conclusions and the input materials that guided the AI process. Label and produce AI outputs that informed opinions, while protecting or redacting prompts or internal queries that reveal strategy.
- Use version control and logs: Maintain dated, auditable records of AI sessions and results, with a focus on outputs that become part of the expert’s testimony. If a court later orders production of AI prompts, you will have a structured, defensible record.
- Prepare for jurisdiction-specific variance: Villanueva adds to a developing landscape with some circuits or districts tilting differently on prompts versus outputs. Monitor how your venue treats AI materials and adjust discovery tactics accordingly.
Objection Academy’s role in ongoing AI discovery readiness
Objection Academy supports trial teams by building muscles for objection handling, evidence-based trial readiness, and AI-assisted preparation grounded in real-world courtroom needs. In the wake of Villanueva, practitioners can leverage Objection Academy to practice objections to AI-generated materials, simulate cross-examinations focusing on AI reliance, and strengthen oral advocacy around the admissibility and credibility of AI-produced outputs. The platform’s emphasis on realistic drills, courtroom realism, and disciplined review aligns with the discipline required to manage AI discovery in complex cases. As the AI era evolves, integrating practical objection training with AI-assisted evidence preparation remains a core advantage for trial teams pursuing faster readiness and stronger trial performance.
What to watch next
The Villanueva decision is part of a broader, unsettled federal landscape regarding AI in discovery. While the Nevada order provides a concrete, jurisdiction-specific rule, other courts have issued different conclusions about the discoverability of AI prompts and AI-assisted materials. The Mayer Brown analysis notes that Conservation Law Foundation v Shell Oil Co. produced a contrasting result in Connecticut, where the court treated AI prompts differently. Expect continued district-level fragmentation and a growing body of persuasive authority that practitioners will need to track for each case’s factual and procedural posture. Staying ahead means building robust internal protocols now, so that when a court orders or refuses production of AI-related materials, the team can respond with clear, defensible positions that preserve essential information while protecting legitimate trial preparations. (mayerbrown.com)
Sources
- Federal Court Splits On Disclosure Of Expert Witness AI Prompts, Mayer Brown, October 09, 2026. (Summary of Villanueva v. LVMPD and the Nevada order, including the court’s distinctions between discoverable AI outputs and protected prompts.) (mayerbrown.com)
- Order in Villanueva v. Las Vegas Metropolitan Police Department, No. 2:24-cv-00125-ART-DJA (D. Nev. Sept. 29, 2026) (ECF No. 112). Justia Dockets & Filings. (Official court record detailing the ruling and its reasoning.) (docs.justia.com)
Note: This analysis focuses on a timely development within the last 45 days and reflects public, verifiable sources. As the law evolves, fresh opinions may refine or diverge from Villanueva’s framework, so ongoing vigilance remains essential for trial teams.