Artificial Intelligence Guide for Law & Law School: AI & Law School

AI Gets Its First Law School A+s, by Andrew Blair-Stanek, Donald G. Gifford, Mark Graber, Guha Krishnamurthi, Jeff Sovern, Donald B. Tobin, and Michael P. Van Alstine, June 10, 2025 (SSRN)

Seven law professors at the University of Maryland had OpenAI's new reasoning model, o3, take their Spring 2025 final exams with the "reasoning effort" setting on high. They graded the model's responses using the same curve applied to their students. In contrast to results from three semesters ago—when the older GPT-4-turbo model would have scored between a B+ and a D—o3 achieved significantly higher marks: three A+s, one A, one A-, two B+s, and one B.

Should Law Students Be Using AI — Even on Exams?

Law professor Robert Diab explores whether law students should be allowed to use AI during exams. While critics argue that it undermines independent legal analysis, others contend that legal education should reflect the realities of AI use in practice. Diab proposes a middle ground: instead of prohibiting AI, assess how effectively students use it.

Bye-Bye, Bluebook? Automating Legal Procedure with Large Language Models

Yale Law student Matthew Dahl conducted a study to evaluate whether leading AI models can accurately apply The Bluebook’s citation rules. His findings highlight that AI tools continue to struggle with the level of procedural precision required in legal practice.

The 'Why' & 'How' of Artificial Intelligence in Legal Scholarship

In a first-of-its-kind issue, the Texas A&M Journal of Property Law has published a collection of law review articles explicitly written with the assistance of generative AI. This foreword outlines the editors’ rationale for embracing AI-assisted scholarship and presents a practical framework for transparency, authorship standards, and editorial review. A valuable read for students and editors engaging critically with the role of AI in legal writing.

ChatGPT Goes to Law School, Jonathan H. Choi, Kristin E. Hickman, Amy Monahan, & Daniel Schwarcz, Journal of Legal Education (forthcoming, 2023)

This study evaluates how well ChatGPT can perform on law school exams without human help. Using real exams from the University of Minnesota Law School, the AI was tested on 95 multiple-choice and 12 essay questions and graded blindly through standard procedures. ChatGPT earned an average grade equivalent to a C+, passing all four courses. The paper discusses the implications for legal education and practice, and offers example prompts and guidance on using ChatGPT for legal writing support.

This piece offers actionable and detailed advice for lawyers and law students on harnessing the power of AI large language models (LLMs) such as GPT-4, Bing Chat, and Bard for legal research and writing.

Should Law Students Be Using AI — Even on Exams? by  Professor Robert Diab, March 26, 2025

In this article, Professor Robert Diab explores the ongoing debate over permitting law students to use AI tools during exams. Critics argue that AI use compromises the assessment of a student’s individual comprehension and analytical skills, particularly in foundational legal subjects. Conversely, a more practical view acknowledges AI’s increasing relevance in legal practice and supports its inclusion in exams to better mirror real-world conditions. Diab proposes a middle ground: instead of banning AI, legal education should evolve to evaluate students on how effectively they use these tools—focusing on their skill in crafting precise prompts and critically assessing AI-generated responses.

AI is Coming for our Classrooms! A Survival Guide for the AI Apocalypse, RIPS Law Librarian Blog, Julie Tedjeske Crane, January 8, 2024. This article examines the role of artificial intelligence in legal education, emphasizing that AI should be seen as an opportunity rather than a threat. The author advocates for integrating AI and technology into law school curricula to better prepare students for the modern legal market. Key points include the potential of AI to enhance legal training, the importance of ethical considerations, and the necessity for future lawyers to understand AI’s implications in practice. The article concludes that embracing AI is crucial to ensuring law students are equipped for the evolving legal landscape.

AI and The Bluebook: Why ChatGPT Falls Short of Traditional Algorithms for Bluebook Legal Citation Formatting (LawNext)

The article examines why ChatGPT struggles with formatting legal citations in accordance with The Bluebook. It points out key limitations, including the AI's inability to consistently apply required formatting styles like italics and small caps, as well as its frequent mistakes in generating, abbreviating, or structuring citations accurately. These shortcomings make it unreliable for precise legal citation work.

ChatGPT, Professor of Law,Tammy Pettinato Oltz, 2023 U. Ill. J.L. Tech. & Pol'y 208 (2023)

In this paper, the author tests ChatGPT’s ability to assist law professors with common tasks by running four service-related and three teaching-related prompts through the free version of the model. ChatGPT performed excellently on service-related tasks and moderately on teaching-related tasks. The paper concludes that ChatGPT can generate near-complete solutions for routine tasks and serve as a useful starting point for more complex tasks.

The Ultimate Study Partner: Using A Custom Chatbot To Optimize Student Studying During Law School, Sean A. Harrington, SSRN (draft working paper), Jun. 7, 2023

This article examines how artificial intelligence can enhance law students' study experiences by supplementing traditional teaching methods. The author explores various AI applications, such as Large Language Models, that can turn static study materials into interactive learning tools. Students can upload documents like outlines, lecture notes, and textbooks into AI systems to generate summaries, practice questions, essays, and even visual aids for memorization. The article highlights AI’s potential in creating personalized study outlines, simplifying legal texts, converting materials into audio or visual formats, and automating spaced-repetition flashcards. The author argues that AI can help law students gain deeper insights, identify knowledge gaps, and improve memorization and application skills, positioning AI as an "ultimate study partner" that enhances legal education.

AI In Law School: A Leap into The Unknown, Law360, Daniel Connolly, August 2, 2023

The narrative highlights a divide in legal academia over the use of AI. During a comedy event at the University of Memphis Law School, skepticism—particularly toward ChatGPT—was expressed through humor. Law student Tiffany Odom-Rodriquez raised concerns that AI could encourage plagiarism and weaken core legal skills. While some institutions, such as Arizona State University, support the use of AI in law school applications, others like the University of Michigan discourage it. Although AI's ability to handle routine tasks is appreciated by some, a recent survey shows that only 9% of law students currently use generative AI, and just 25% intend to in the future, reflecting ongoing worries about its accuracy and possible effects on legal careers.

Learning the Law with AI: Why Law School Students Are Tentative about Using ChatGPT, Sarah Wellen, LawNext, June 2nd, 2023 

A LexisNexis survey reveals that many lawyers are already integrating generative AI into their work, using it for research (59%), improving efficiency (54%), drafting documents (45%), and writing emails (34%). In contrast, law students remain wary. Only 9% currently use generative AI, and just 25% plan to adopt it in the future. Their concerns center on AI’s potential to generate inaccurate or outdated information, which could undermine the reliability of legal research. Students also worry about academic integrity, fearing that AI could facilitate plagiarism or cheating. Additionally, they argue that AI lacks the capacity to grasp nuanced legal reasoning, which could hinder the development of essential critical thinking skills. The growing use of AI in routine legal work also sparks anxiety over job prospects for entry-level lawyers.

ChatGPT Is an OK Law Student. Can It Be an OK Lawyer?, Bloomberg Law, January 27, 2023

This Bloomberg Law opinion piece explores the potential impact of ChatGPT on the legal profession. It notes that ChatGPT scored around the level of a C+ student on four law exams at the University of Minnesota Law School, while performing better on a business exam at Wharton. While the article concedes that ChatGPT is unlikely to replace human lawyers, it highlights possible benefits, including lower costs, faster service, and greater access to legal assistance. At the same time, it raises important concerns about ethical implications, quality control, and liability associated with relying on AI for legal work.

Legal Scholarship Through the Lens of Generative AI, Darkly
by Andrew Martineau and Loren Turner, March 25, 2025 (SSRN)

This article explores GPT-4's engagement with law review articles, highlighting its limitations in independently summarizing content but demonstrating improved accuracy when given the full text. Retrieval-augmented generation (RAG) may enhance AI reliability through more automated methods, though issues like algorithmic bias, intellectual property rights, and the broader effects on legal scholarship remain significant. Law librarians need to thoughtfully evaluate these concerns when managing how institutional research is accessed and utilized by AI technologies.

ChatGPT and the future of legal education and practice, Marjan Ajevski, , Kim Barker, Andrew Gilbert, Liz Hardie, & Francine Ryan, The Law Teacher (2023)

The launch of ChatGPT in November 2022 marked a significant moment in AI development, prompting wide interest in its potential impact on legal education and practice. This article explores how law schools can adapt assessments to challenge generative AI while also educating students on its workplace applications. It emphasizes the need for legal educators to navigate the complexities of open-source AI and recognize its potential influence on both education and practice, urging the integration of AI technology into teaching methods.

AI Now
by Rachelle Holmes Perkins, May 24, 2024 (SSRN)

This Article argues that every law professor bears an unavoidable responsibility to understand generative artificial intelligence. This duty arises from the essential role faculty members play across three interconnected areas: teaching, scholarship, and institutional governance. No member of the law faculty is exempt—whether their focus is on instruction, research, clinical work, or administration, their responsibilities inevitably touch at least one of these domains. This obligation does not depend on a professor’s personal interest in using AI in their own work. Rather, it reflects the pressing and multifaceted influence that law professors have on shaping both future legal professionals and the broader structure of the legal system.

Language Models, Plagiarism, and Legal Writing
by Michael L. Smith, August 16, 2023 (SSRN)
The author contends that "those urging the incorporation of language models into legal writing education leave out a key technique employed by lawyers across the country: plagiarism. Attorneys have copied from each other, secondary sources, and themselves for decades. While a few brave souls have begun to urge that law schools inform students of this reality and teach them to plagiarize effectively, most schools continue to unequivocally condemn the practice...(but) continued condemnation of plagiarism is inconsistent with calls to adopt language models, as the same justifications for incorporating language models into legal writing pedagogy apply with equal or greater force to incorporating plagiarism into legal writing education as well."

How to Use Large Language Models for Empirical Legal Research
by Jonathan H. Choi, August 13, 2023 (SSRN)

This Article showcases how large language models (LLMs) can be applied to the analysis of legal documents, offering guidance on best practices while examining both their capabilities and limitations in empirical legal research. In a straightforward classification task involving Supreme Court opinions, the study finds that GPT-4 matches human coder performance and significantly outperforms earlier NLP classifiers. Notably, its accuracy shows no meaningful gains from supervised training, fine-tuning, or tailored prompting.

Re-Evaluating GPT-4's Bar Exam Performance
by Eric Martínez, May 8, 2023 (SSRN)

This paper examines the methodological difficulties in substantiating GPT-4's widely cited claim of scoring in the 90th percentile on the bar exam. It presents four sets of findings indicating that, while GPT-4 represents a significant advancement over GPT-3.5, OpenAI’s reported UBE percentile scores appear to be overstated.

Learning the Law with AI: Why Law School Students Are Tentative about Using ChatGPT, LawNext, 6/2/23

Law school students have varying levels of enthusiasm for using AI, particularly ChatGPT, as an educational tool. While some believe it can assist in research and writing, others worry it might limit critical thinking skills and raise ethical issues, including plagiarism. Legal educators must carefully consider the benefits and drawbacks of AI integration to enhance legal education effectively.