The Scholarship Lab
Expanding legal scholarship by showing the process and preserving the record.
The Scholarship Lab publishes transparent, inspectable experiments in AI-assisted legal scholarship, preserving the prompts, outputs, verification, limitations, and human judgment behind the work.
Ways the Lab Publishes Work
Most scholarship shows only the destination. The Lab can publish the destination, the route, or both, whatever makes the work most useful and most honest.
The Final Work
A completed scholarly article, essay, or analysis, published in the usual way, for authors who want the end product to stand on its own.
A Companion Living Edition
Alongside a published article, JLETI can release a Living Edition that reproduces the AI-generated text, discloses the model and prompt used, audits a sample of citations, and includes the author's preface.
A Process-Only Record
Sometimes the process is the contribution. The Lab also publishes standalone records: prompt logs, model outputs, comparisons, citation audits, and verification notes, even when there is no finished article attached.
Companion Living Editions and process records can also be versioned and updated over time as tools, models, and the law change, keeping the record accurate as well as durable.
The AI Scholarship Lab is a publication space for experimental work on artificial intelligence, legal scholarship, authorship, and verification. It is designed for projects that do not fit neatly within traditional article formats, and it is for work that uses AI in a meaningful, transparent, and scholarly way, not text published simply because AI generated it.
See the process, not just the product.
Most AI-assisted work is shown only after it has been cleaned up, corrected, and finalized. What usually disappears is the process behind it: which system was used, how it was prompted, what worked, what failed, and what a human had to catch, question, verify, or revise.
The AI Scholarship Lab preserves that record. It makes visible how AI-assisted legal work was created, where the technology helped, where it failed, and how the author responded. By publishing the work and the workings behind it, the Lab creates scholarship others can inspect, learn from, test, and build on, rather than a polished final product with its method erased.
Dean Andrew Perlman's Article
JLETI's inaugural Living Edition pairs Dean Andrew Perlman's article with a version generated by Claude Fable 5 in a single prompt.
A Living Archive of AI's Legal Outputs
Artificial intelligence is advancing faster than the legal profession can document it. Models are released, updated, restricted, replaced, and retired quickly, and when a model changes or disappears, what vanishes with it is more than a product.
The specific outputs it generated, the prompts that produced them, the way it reasoned, the citations it invented or got right, its characteristic strengths and failures, all of it can become difficult or impossible to recover. We are, in real time, losing the record of what these systems could and could not do.
The AI Scholarship Lab is, in part, an answer to that loss. It functions as a durable, citable archive of AI in legal work, preserving not only AI-assisted scholarship, but the underlying artifacts that give it meaning: the prompts, the model and version used, the outputs as generated, author notes, citation audits, and the human verification performed around them.
An archive like this matters across the profession:
Legal Education
Students and faculty can study how specific systems actually reasoned about real legal problems: where they were reliable, where they hallucinated authorities, and how prompting and verification change the result.
The Practice of Law
Practitioners need an honest, time-stamped record of how tools behaved at a given moment, so claims about reliability and professional responsibility rest on evidence rather than marketing.
Scholarship and Reproducibility
A finding about an AI system is only meaningful if others can see what was tested and how. Archiving prompts, outputs, and methods keeps AI-era scholarship inspectable.
Accountability and Policy
As AI systems attract regulatory, institutional, and professional scrutiny, a neutral scholarly record of their documented capabilities and limits becomes a public good for courts, regulators, educators, and the public.
The Historical Record
This is a formative period for law and technology. The Lab preserves primary sources from it, so future scholars need not reconstruct what today's systems did from memory, marketing, or press releases.
What We Are Looking For
Transparent, inspectable, and useful experiments.
Submissions should be transparent, inspectable, and useful. Authors should be prepared to explain what was tested, how AI was used, what role humans played, how claims or citations were verified, and what limitations remain.
JLETI created the AI Scholarship Lab to expand what serious legal scholarship can look like in an era when legal knowledge is increasingly shaped by models, prompts, platforms, and human-machine collaboration, and to preserve that work before the tools that shaped it disappear.
Examples of Submissions
The AI Scholarship Lab welcomes experimental legal scholarship in a variety of forms, including:
AI-Generated Scholarship Experiments
A generative AI system produces a legal article, essay, argument, or theory, accompanied by an editor's note, author's note, prompt disclosure, verification statement, or critical commentary.
Human-AI Collaborative Scholarship
The author uses AI in research, drafting, synthesis, critique, revision, or verification, and reflects on how that collaboration shaped the work.
Prompt-Based Legal Scholarship
Prompts, prompt sequences, model outputs, and human interventions are treated as part of the scholarly record: prompt logs, annotated outputs, comparison tables, or commentary.
Model Comparison Studies
A project comparing how different AI systems analyze the same legal problem, summarize doctrine, identify authorities, or reason through legal issues.
Verification and Citation Audits
A project testing the accuracy, reliability, hallucination risk, citation integrity, or doctrinal soundness of AI-generated legal analysis.
Annotated AI Outputs
AI-generated legal analysis presented with human annotation, correction, source verification, or editorial explanation.
Tool-Based Research Workflows
Documentation of how AI tools were used in a research, teaching, writing, litigation, compliance, policy, law-library, or legal-education workflow.
Living or Versioned Scholarship
A project designed to be updated over time, with version histories, update notes, public correction logs, or continuing editorial review.
Teaching and Training Experiments
A project using AI to train students, faculty, lawyers, law librarians, judges, or others in legal research, verification, professional responsibility, or AI literacy.
Scholarly Demonstrations
A submission that performs an idea rather than merely arguing it. For example, an essay on AI-generated scholarship that includes the AI-generated article, the prompt used, an editorial verification note, and invited commentary.
Submission Requirements
Submissions to the AI Scholarship Lab should include, where applicable:
Editorial note: The AI Scholarship Lab is a publication category, not a shortcut around editorial review. Submissions remain subject to JLETI's review, verification, disclosure, formatting, and publication policies.
Submit to the AI Scholarship Lab
Use this category for experimental legal scholarship involving AI, authorship, verification, prompt records, model outputs, the preservation of AI's legal record, or new forms of scholarly publishing.
