A founding proposal Nevada’s law school · Las Vegas
UNLV Boyd Center for AI & Legal Writing
The proposed Center would bring together a dedicated team to study AI in legal writing, evaluate AI tools, train lawyers and judges, develop standards for responsible use, and expand access to justice.
The Center will build on Write.law’s national legal-writing and AI training platform. Figures reported by Write.law
Select “Scan the AI draft” to check claims, citations, confidential information, and lawyer responsibility.
Interactive demonstrationStart here
Argument / Draft 01
The Ninth Circuit has that a public agency must provide individualized notice before changing access to essential benefits.
“Notice must be calibrated to the recipient’s circumstances.”
The denial was especially harmful because .
Claim support, citation accuracy, confidentiality, and lawyer responsibility will be checked in sequence.
Illustrative interface · Not legal advice · A responsible lawyer must verify sources and conclusions
Why independent evaluation is needed now
Lawyers are adopting AI, courts are addressing errors, and investment is increasing. The profession needs independent research on when AI improves legal work, when it introduces risk, and what effective supervision requires.
of 1,300+ legal-industry respondents reported using general-purpose AI for work; survey fielded Sept.–Oct. 2025
8am 2026 Legal Industry Report · vendor survey ↗02100+AI-related sanctions decisions analyzed across more than 60 U.S. jurisdictions, 2023–2026
Lucci, forthcoming Capital University Law Review ↗0317–33%hallucination rates in a preregistered Mar.–Apr. 2024 evaluation of three legal AI research tools
Stanford RegLab / Journal of Empirical Legal Studies ↗04$3.9Bprojected global legal AI market by 2030, up from an estimated $1.4B in 2024
Grand View Research market forecast ↗Survey results, empirical studies, sanctions research, and commercial forecasts answer different questions. Each figure is labeled and linked to its source.
Independent research on AI and legal writing
The Center would study how AI changes legal research, reasoning, drafting, and review. Working with practitioners, Boyd would compare work with and without AI to identify where it improves quality and access, where it introduces risk, and how lawyers should supervise its use.
Independent research on AITest how AI affects accuracy, reasoning, clarity, efficiency, and learning by comparing legal-writing tasks completed with and without AI.
Standards for responsible AI useUse evidence about AI to develop professional competencies, model policies, verification procedures, and guidance for selecting and supervising legal AI tools.
Education in AI and legal writingTeach students, lawyers, judges, and faculty how to use AI in legal research and writing, evaluate its output, protect confidential information, and retain professional judgment.
AI tools for specific legal needsWork with engineers, practitioners, and communities to develop and evaluate AI tools for defined legal-writing tasks, with clear limits and human oversight.
AI tools for distinct legal tasks
The Center would work with engineers, legal technologists, courts, legal aid organizations, firms, and intended users to develop and evaluate AI tools for public legal information, judicial writing, legal practice, rural communities, and professional education.
Existing tool developed by Professor Regalia at Write.law
Prompt Pilot asks clarifying questions, identifies available choices, marks decisions that require professional judgment, and leaves the lawyer responsible for the instruction and the result.
Draft an argument that the agency violated due process.
Before drafting, identify the strongest counterargument, list any missing facts, and distinguish verified authority from assumptions. Ask me to resolve each judgment call.
The proposal calls for an in-house AI and product lead and a team of software engineers, researchers, and partner lawyers.
Every project would require authoritative sources, security, accessibility, independent evaluation, and accountable professional review. Prompt Pilot is an existing Write.law tool developed by Professor Regalia; the other examples are proposed projects.What the Center would teach and study
Lawyers must know how to define an assignment, verify sources, protect confidential information, supervise AI-assisted work, measure results, and recognize when AI should not be used. The Center would teach and empirically study these competencies.
Decide when professional judgment is required, when review must occur, and who is responsible for the final work.
Identify required approvals, issues that require escalation, and decisions that cannot be delegated.
Compare quality when expert review occurs at different stages of the work.
A named lawyer inspects the sources and reasoning, not only the prose.
H Human review and responsibility must be designed into the process.
Begin with three defined programs
Measure how AI changes accuracy, reasoning, clarity, trust, time, learning, and equity in legal writing. Compare work completed without AI, by AI alone, and by people using AI to identify benefits, failures, and effective oversight.
Initial deliverable → public report on AI’s effectsDefine what lawyers, judges, students, and supervisors should know about directing AI, verifying its output, protecting confidential information, disclosing its use, and retaining professional responsibility.
Initial deliverable → proposed national standard and training programDevelop and evaluate an AI-assisted public tool—such as a custody-forms guide—with Nevada courts or legal aid, then publish evidence of its benefits, risks, and limits for other jurisdictions.
Initial deliverable → evaluated public-service tool06 Why independent leadership matters
Vendors will influence practice through their products. Independent law schools should also evaluate those products and develop evidence-based professional standards.Why Boyd is positioned to lead ↓
A setting for the proposed Center
The proposal calls for recruiting researchers, engineers, and practitioners around defined AI projects. Interested faculty would be invited to collaborate through specific projects and agreed roles.
See the proposed team ↓Beginning fall 2026, every first-year Boyd student will receive practical instruction in responsible generative and agentic AI use.
Read the announcement ↗Since its founding in 1998, Boyd has developed nationally recognized programs without centuries of inherited structure. It has demonstrated that it can create new programs quickly and deliberately.
About Boyd Law ↗Las Vegas is a global convening city, UNLV has established university-industry partnerships, and Nevada’s legal community is closely connected. These conditions support national meetings and practice-based research.
Explore UNLV innovation ↗The first 36 months
The Center would build credibility through careful research, useful programs and tools, strong partnerships, and independent conclusions.
Phase 010–12 months
Phase 0212–24 months
Phase 0324–36 months
Founding director
Professor Regalia will serve as the Center’s founding director. A Boyd professor and co-founder of Write.law, he works at the intersection of legal writing, AI, professional judgment, and training for legal teams. Through Write.law, he has already developed courses, workshops, coaching, consulting, and AI tools for firms, courts, agencies, legal departments, schools, and individual lawyers.
The Center will pair that existing practitioner network and product experience with Boyd’s academic independence and public mission. It will not start from a blank page.
A real delivery team
Secure architecture, model evaluation, product strategy, and accountable delivery
Developers and fellows recruited for AI projects, with opportunities for interested UNLV collaborators
Researchers and practitioners recruited to test accuracy, confidentiality, usability, and risk
Firm studies, custom tools, private pilots, executive training, and implementation support
10 · Founding partnerships
Boyd intends to establish the world’s first center dedicated specifically to AI and legal writing. We are seeking a limited group of firms and legal organizations whose founding commitments can launch its people, research, technology, education, and public-service work.
A founding partnership can do more than support research. It can give a firm a defined place in the Center’s national work—and a practical path to work with its team.
For a qualifying commitment, naming rights for the Center or a major program may be available, subject to university approval and a definitive gift agreement.
01 · Institutional and marketing value
Agreed recognition as one of the organizations that made a first-of-its-kind center possible—at launch, online, and in selected programs and events—consistent with university policy and the gift agreement.
Designated leaders invited to private executive roundtables, founding-partner sessions, research briefings, and the Las Vegas Forum on AI & Legal Writing. Partners can raise practice questions and help shape convening agendas.
Support identifiable work for courts, legal aid, rural and solo practitioners, students, and the public, with recognition where appropriate. Philanthropic treatment will follow the applicable UNLV or Foundation process, agreements, and law.
02 · Applied support for your firm
Test how AI changes the research, drafting, review, and approval of briefs, contracts, client communications, and other documents. Compare results with your existing process and measure AI’s benefits, risks, and demands on lawyer supervision.
For suitable projects, scope, design, prototype, and evaluate tools around the firm’s users, documents, policies, workflows, and security requirements. Production rights, licensing, maintenance, intellectual property, and cost would be governed separately.
Priority—and when appropriate, exclusive pilot—access to selected pre-release tools and demonstrations, with testing before any decision to adopt at scale.
Private programs for innovation leaders, partners, supervisors, associates, and professional staff, tailored to the firm’s practice areas, tools, policies, workflows, and risk profile.
Founding support funds the Center’s public mission. Firm-specific research, software development, licensing, and training would be separately scoped so that deliverables, confidentiality, security, data use, intellectual property, and fees are clear. Boyd would independently control academic research methods, publication, and conclusions.
Founding donors · Law firms and legal departments · Technology partners · Courts, bars, and public institutions
Existing tool developed by Professor Regalia at Write.law
Prompt Pilot asks clarifying questions, identifies available choices, marks decisions that require professional judgment, and leaves the lawyer responsible for the instruction and the result.
Proposed Nevada public-service project
A guided, plain-language app could present complex custody forms in an understandable sequence, check for missing information, explain unfamiliar terms, and identify questions that require legal aid or court staff.
Proposed legal-writing review tool
A writing tool could connect each proposition to cited authority or record evidence, flag unsupported statements, identify unnecessary private details, and record the responsible lawyer’s approval.
Proposed tool for judges and clerks
A judge-controlled tool could review structure, support in the record, citation accuracy, public comprehension, and plain-language explanations without replacing judicial reasoning.
Decide when professional judgment is required, when review must occur, and who is responsible for the final work.
Analyze and improve the existing process before deciding where AI can help.
A useful instruction identifies the audience, objective, permitted sources, missing information, and constraints.
AI-generated statements must be verified. Every material proposition needs support in authority or the record.
Selecting an AI tool and deciding what information it may receive are professional-responsibility decisions.
Speed is only one measure. Evaluation should also address accuracy, clarity, trust, learning, access, and equity.
Responsible AI use includes recognizing matters in which delegation creates more risk than benefit.