I help lawyers and legal organizations learn to incorporate AI better; by “better” I mean improving efficiency, lowering risk, and building durable skills and knowledge. My work draws on nearly a decade of law school teaching and close, day-to-day attention to how AI is reshaping legal practice.
An AI prompting tool that turns vague drafts into precise, well-scaffolded prompts. Users build a profile capturing their role, expertise, and output preferences; the refinery weaves that context into every prompt and explains each change, helping users internalize better prompt-writing habits.
Start refiningAn email-to-LLM assistant: forward a message to a private address and Claude performs a task against its contents — summarizing, extracting, answering — then replies in-thread. It handles attachments natively (PDFs, images, spreadsheets, calendar invites, and archives), authenticates senders with DMARC/DKIM gates, and never stores message bodies.
Request accessA tool designed to accompany a workshop for Brooklyn Law School faculty on diagnosing and redesigning law school assessments in light of students’ access to AI. The premise: effective assessment in an AI-saturated environment is a design problem, solved by aligning what is assessed with what students should actually be learning. The linked version is a static demo.
Try the demoA curated digest at the intersection of AI, law, education, and policy — surfacing developments, scholarship, and practical resources for legal educators and practitioners navigating generative AI.
Read the digestAn interactive tool for law school faculty and administrators to build, customize, and export a generative AI policy that prioritizes pedagogy, professional responsibility, and practice readiness.
Start buildingA macOS folder-watching agent that turns dropped PDFs, URLs, and saved webpages into a searchable research library. Claude (via API call) reads each new source in the background, writes a structured summary, and files it into a tagged library of plain markdown.
View on GitHubA comprehensive legal research framework that integrates generative AI tools into the traditional research process, providing a structured methodology for balancing technological efficiency with rigorous verification and ethical standards. Designed for 1Ls but suitable for all law students.
Explore the guideA comprehensive, annotated generative AI policy for law clinics, synthesizing scholarly research and pedagogical best practices into a modular framework for institutional adoption. Includes a sample multiple-choice quiz to test knowledge and comprehension.
Review the policyA selection of invited keynote addresses, panels, and workshops centered on the intersection of generative AI and the law. These engagements reflect my commitment to distilling the complexities of emerging tech into accessible frameworks for legal scholars, practitioners, and general audiences.
This New York CLE webinar for Justia moved past the familiar hallucinated-citation cases to the harder questions now facing New York practitioners: what 22 NYCRR Part 161 does and does not require by way of AI disclosure, “second-generation” failure modes such as real cases cited for the wrong proposition, and what “reasonable efforts” under Rule 1.6 mean when a vendor’s retrieval system holds client documents.
The first program in Brooklyn Law School’s faculty Summer AI Series. Using the Integrity by Design framework, this session treated assessment validity as a design problem: whether an assessment’s format still requires the cognitive work it is meant to measure once students have access to AI. Participants classified assessments as AI-resistant, AI-transparent, or AI-embedded, then worked through live redesigns of two faculty assessments. The Assessment Validator above was built as a companion to this session.
This three-session, full-day workshop for Mitchell Hamline School of Law faculty covers AI fundamentals, policies, and assessments. The session on AI fundamentals addresses capabilities and limitations, prompting, and use cases; the policies session included customizable policies for all course types grounded in pedagogical principles; and the assessments session covered the spectrum of AI-resistant, AI-transparent, and AI-integrative assessments for in-person, hybrid, and remote course formats.
A student-organized panel and networking event at Rutgers Law School on how generative AI is changing day-to-day legal practice and what that means for students about to enter the profession.
This workshop and CLE session for Rutgers Law School clinical faculty covers the responsible and ethical use, including necessary risks and safeguards for integrating Generative AI into clinical legal education.
In this presentation for Rutgers Law School senior leadership, I make the case for integrating AI literacy as an essential component of legal education curriculum and workflows.
Delivered at the RWJBarnabas Health Medical Humanities Conference, this presentation details the practical applications of Large Language Models for research, writing, studying, and skill development.
Featured speaker at the journal’s symposium, Guardrails for Green Tech: Legal Perspectives on AI’s Environmental Impact. This talk explores the environmental regulation and responsibility of AI by applying key lessons learned from the history of consumer recycling.
Panelist at the Rutgers Business Law Journal symposium on the near-term trajectory of AI in business law practice, including adoption, risk, and the professional responsibility questions that follow.
A four-session series covering: AI for drafting and reviewing hypos; AI assistance with recommendation and clerkship letters; AI for studying; and AI in writing courses and supervising student notes.
Covering ethical considerations, prompt engineering, and the limitations of LLMs, this remote presentation introduced members to fundamental generative AI skills.
This presentation (co-presented with Anna Elbroch) sought to inform educators and faculty about some best practices for integrating Generative AI tools into online and other educational courses.
Co-presented with Anna Elbroch, this session examined whether generative AI is a threat or an asset in distance-education courses, where students work unsupervised and AI use is hard to observe, and offered course-design responses that treat AI as part of the learning environment rather than something to police.
An introductory exploration of general AI models and prompting strategies, designed to highlight high-value use cases alongside the essential ethical guardrails required for legal work.
As a panelist at the Seton Hall Journal of Legislation & Public Policy symposium Artificial Intelligence for Lawyers and Law Students: Crutch, Craft, or Catalyst, I explained the academic integrity and professional responsibility implications of AI in law schools.
This brief ethics overview for legal writing professors (co-presented with Kirsten Davis) analyzed the implications of ABA Formal Opinion No. 512 for Generative AI ethics in both legal writing and ethics instruction.
This keynote address (co-presented with Hon. Scott Schlegel) at the State Bar of Wisconsin Annual Meeting & Conference addressed the specific benefits and risks of using artificial intelligence in the practice of law, including hallucinations, deepfakes and authentication of evidence, and ethical duties.
In this CLE session at the State Bar of Wisconsin Annual Meeting & Conference, this presentation provided an overview of generative AI utility in legal practice, covering the mechanics of effective prompting, diverse workflow applications, and the critical security warnings inherent to the technology.
This presentation advocated for the integration of generative AI into law school curricula, exploring how faculty and administrators can master the technology’s mechanics, ethics, and practical applications to effectively adapt teaching and assessment for an AI-augmented legal landscape.
Presented to the Rutgers Law School Faculty Committee on Technology and AI, this session offered a framework and sample language to help faculty write course-level generative AI policies that are clear to students, consistent across the curriculum, and grounded in learning objectives rather than blanket prohibition.
A comprehensive look at the capabilities and limitations of LLMs, featuring a deep dive into prompt design, practical office integration, and the primary ethical risks facing practitioners.
Focused on improving legal writing, this California CLE webinar highlights essential techniques drawn from the distinct writing styles of renowned Supreme Court Justices.
“If a machine can easily do everything we are asking our students to do, then we are not asking enough of our students.”
— David S. KempI contribute to the legal conversation through a mix of formal journal articles and digital editorials. My work often explores the intersection of emerging technologies, equity, and ethics, distilling complex legal frameworks into clear, accessible insights for scholars, industry leaders, and the public alike.
Legal Writing (forthcoming Mar. 2027)
This essay argues that AI research tools, by delivering a finished analysis in minutes, let first-year students bypass the struggle through which research judgment develops: monitoring what they do not yet know, working through productive difficulty, and regulating their own process. It proposes sequencing conventional research before AI at both the curricular and the assignment level, enforced through a structured research log that records a pre-research plan, the provenance of every source, and how each AI-identified authority could have been found and verified by conventional means. A sample research plan and log is included as an appendix.
51 Rutgers Comput. & Tech. L.J. SE81 (2026)
Drawing on the failures of consumer recycling, this article argues that AI’s environmental impact cannot be solved through individual responsibility or voluntary corporate commitments alone. Instead, it proposes a regulatory framework that aligns market forces with sustainability by using mandatory disclosures, resource-weighted fees, and producer responsibility to reward resource efficiency.
49 Seton Hall J. Legis. & Pub. Pol’y 633 (2025)
This article explores the integration of generative AI into legal education, examining its history and the potential for faculty to utilize it for pedagogical, administrative, and scholarly purposes. It further addresses the risks of student over-reliance on AI, proposing that law schools should treat AI as a collaborative tool and shift toward assessments that prioritize process and diverse lawyering skills over traditional exams.
Verdict (Sep. 2023)
The article argues that ChatGPT’s well-known failures in legal research — such as hallucinating cases and misrepresenting the law — can be turned into a pedagogical advantage, by having law students use AI-generated output as a starting point for learning how to verify legal information using reliable tools like Westlaw and LexisNexis.
Verdict (Aug. 2023)
The article argues for the integration—rather than prohibition—of generative AI tools, specifically ChatGPT, into legal education. It suggests that guided AI instruction on how to harness this disruptive technology prepares law students for the evolving landscape of legal practice.
24 AMA J. Ethics E773 (2022) (with Charles E. Binkley and Brandi Braud Scully)
Written with surgeon and bioethicist Charles E. Binkley and Brandi Braud Scully, this article examines whether machine-learning clinical decision support can counter surgeons’ tendency to overestimate operative risk for patients based on race and socioeconomic status, and warns that a model trained on historically biased outcomes data will reproduce the bias it was meant to remove.
Shorter and more frequent pieces on AI, law, and legal education appear on my blog, Learning Machines. The latest posts:
I spent nearly a decade teaching in law schools — at Rutgers Law School, UC Berkeley School of Law, and UC Law San Francisco — where I built curricula, assessment frameworks, and training programs around one of legal education’s most pressing questions: how to prepare lawyers for a profession being reshaped by artificial intelligence. That work included six semesters of a course I designed, first called Emerging Tools & Technology in the Practice of Law and most recently AI Skills for Lawyers, centered on building the kind of transferable judgment and problem-solving skills that hold up as tools and circumstances change.
I now bring that same approach — grounded in learning science, professional responsibility, and close tracking of a field that shifts weekly — to the practice side, working on AI innovation within a global law firm. The transition from teaching to practice sharpened a conviction I’ve held throughout: the people building and deploying these tools need the same rigor around pedagogy, ethics, and institutional design that the best law schools demand.
Alongside that work, I have served since 2011 as managing editor of Oyez and Justia’s Verdict, and since 2025 as Secretary of the Board of Directors of the Center for Computer-Assisted Legal Instruction (CALI).
My approach to training and curriculum development draws on how memory works, how skills transfer, and how judgment develops under pressure. That foundation shapes everything from how I sequence a training program to how I design individual exercises.
I remain actively engaged with legal education and welcome conversations with law school administrations, faculties, and legal organizations about AI pedagogy, curriculum design, and the institutional questions law schools are working through right now.
Juris Doctor
Bachelor of Arts, Psychology
A complete record of my academic appointments, publications, presentations, and professional experience.
Download CV