How Law Schools Are Scrambling to Teach AI After ChatGPT Disrupted Legal Research

Quick Answer

How are law schools responding to AI tools like ChatGPT in legal research? Law schools are rapidly updating curricula to include AI literacy, prompt engineering, and technology ethics courses. Programs at Harvard, Stanford, Michigan, and others have introduced dedicated AI labs, revised legal writing requirements, and partnered with companies like LexisNexis and Thomson Reuters to integrate AI-powered tools directly into the classroom. The goal: train lawyers who can critically use and ethically govern AI in legal practice.

The Disruption: What ChatGPT Did to Legal Research

When OpenAI released ChatGPT in November 2022, legal academia held its breath. Within weeks, law students were using the tool to draft memoranda, summarize case law, and outline legal arguments. These tasks had long been the backbone of first-year legal writing courses.

The disruption was immediate and undeniable.

What changed overnight:

• Law students could generate a passable first draft of a legal brief in minutes
• Professors began receiving suspiciously polished assignments from students who had struggled all semester
• Legal research platforms like Westlaw and LexisNexis scrambled to integrate AI into their own products before students abandoned them entirely.
• Law review editors started questioning how to handle AI-assisted submissions.

A 2023 survey by the Journal of Legal Education found that over 72% of law students had used a generative AI tool for at least one academic task within six months of ChatGPT’s public release. More telling: only 18% of law schools had any policy addressing AI use at the time.

The gap between student behavior and institutional readiness was, and in many places remains, vast.

The Hallucination Problem That Rattled the Profession

The legal community got its loudest early warning in May 2023, when a Manhattan federal judge sanctioned attorneys at the firm Levidow, Levidow & Oberman for submitting a brief containing six completely fabricated case citations generated by ChatGPT. The attorneys had not verified the citations before filing.

The case became a flashpoint. Law school deans cited it in emergency faculty meetings. Bar associations issued ethics opinions. And suddenly, every legal writing professor in America had a very concrete reason to completely rethink how they were teaching research methodology.

The incident didn’t just reveal AI’s limitations. It revealed a gap in how lawyers were being trained to think about source verification in the first place.

How Law Schools Are Responding

Law schools have responded to the AI moment along a broad spectrum, from bold integration to cautious restriction. What’s clear is that the “wait and see” posture is becoming untenable.

The Three Camps

1. Early Adopters (Integrate and Lead): Schools like Stanford Law School’s CodeX: The Stanford Center for Legal Informatics and Harvard Law School’s Library Innovation Lab moved quickly to not just acknowledge AI but to build entire research initiatives around it. These institutions are producing scholarship on AI governance, launching legal tech clinics, and training students on the next generation of tools before those tools even reach the market.

2. Adapters (Reform and Integrate): The majority of ABA-accredited law schools fall here. These programs have revised academic integrity policies, added AI literacy modules to existing courses, and begun requiring students to learn specific platforms. The adaptation is real, but it is reactive rather than strategic.

3. Resisters (Restrict and Monitor): A smaller cohort of schools, particularly those with strong traditionalist faculties, has doubled down on AI-free exams, in-class handwritten assignments, and explicit bans on AI use in research submissions. Critics argue this prepares students for a legal profession that no longer exists.

Faculty Divide

The debate within faculties is sharp. Legal writing professors, who have spent careers teaching the craft of researching and constructing arguments from scratch, often worry that AI tools short-circuit the cognitive development that struggle produces. Clinicians and practice-oriented faculty tend to take the opposite view: if AI is in every law firm, students who can’t use it are at a disadvantage from day one.

That tension is unlikely to resolve fully, but it is forcing a productive reckoning with what law school is actually for.

New AI Courses and Curriculum Changes

What Schools Are Adding to Their Curricula

Across the country, law schools are introducing new coursework at a rate not seen since the integration of the internet into legal research in the 1990s.

Dedicated AI and Law Courses

Programs now offering standalone AI and law courses include:

• Harvard Law School: “Computers, Privacy & the Law” (expanded with generative AI modules)
• Stanford Law School: “Law, Technology & Culture” and CodeX fellowships focused on AI governance
• Georgetown University Law Center: “Artificial Intelligence and the Law”
• University of Michigan Law School: “Technology, Law, and Public Policy”
• Vanderbilt University Law School: “Lawyering in the Age of Artificial Intelligence”
• University of California, Berkeley School of Law: “Regulating Artificial Intelligence”
• New York University School of Law: AI-specific modules in the Engelberg Center on Innovation Law & Policy

These are not elective curiosities. At several schools, AI literacy components are being woven into required 1L courses, particularly legal research and writing, making them part of the foundational legal education experience.

What These Courses Cover

A typical AI and law curriculum in 2024 addresses:

• How large language models (LLMs) work at a conceptual level sufficient for a lawyer to ask the right questions
• Prompt engineering for legal tasks: how to construct queries that yield useful, accurate outputs
• Hallucination and verification: the professional responsibility obligations that flow from AI errors
• AI-generated evidence and discovery: how courts are handling AI in litigation contexts
• Bias in AI systems and implications for equal justice
• Emerging regulatory frameworks (EU AI Act, U.S. executive orders, state-level legislation)
• Contractual and IP issues surrounding AI-generated work product

Legal Writing Courses: The Biggest Battlefield

Nowhere is the AI debate more acute than in legal writing: the foundational first-year course that teaches students to research, analyze, and communicate legal arguments.

Legal writing professors are navigating a genuine pedagogical paradox:

• If they ban AI entirely, they may be teaching students to work in ways that no longer reflect practice
• If they allow AI without scaffolding, they may be depriving students of the intellectual struggle that builds legal reasoning skills.
• If they try to detect AI use through software, they risk academic integrity processes built on unreliable tools.

The emerging consensus, though not yet universal, is a scaffolded integration model: students learn the underlying skills manually first, then learn to use AI tools to augment those skills, and then are assessed on their ability to evaluate AI-generated work product critically.

Think of it as analogous to how math education handled calculators: you learn long division first, then the calculator is a tool, not a shortcut.

AI Tools Being Taught in Law Schools

The Legal Research Platforms

The legacy legal research platforms have not stood still. Both Westlaw and LexisNexis have made major AI investments, and law schools, many of which have institutional licensing agreements, are beginning to teach these tools formally.

Thomson Reuters CoCounsel (formerly Casetext): Acquired by Thomson Reuters in 2023 for $650 million, Casetext’s CoCounsel product is now being integrated into Westlaw’s product suite. It can draft legal documents, review contracts, and conduct research across Westlaw’s case law database with natural language queries. Several law schools have begun including CoCounsel in their legal research curricula.

LexisNexis Lexis+ AI: LexisNexis launched Lexis+ AI with a strong emphasis on citation verification—a direct response to the hallucination crisis. The tool includes a feature that shows users the underlying sources behind every generated statement, which law schools are increasingly teaching as a model for responsible AI use.

Westlaw Precision and AI-Assisted Research: Thomson Reuters has integrated AI research assistance directly into Westlaw, allowing natural language queries that pull verified case law. The difference from raw ChatGPT use: the results are drawn from a curated, verified legal database, not the open web.

General-Purpose AI Tools Being Addressed

Beyond legal-specific platforms, law schools are addressing student use of:

• ChatGPT (GPT-4 and later models): Used for drafting, brainstorming, and summarizing; subject to hallucination risk when used for legal citations
• Claude (Anthropic): Increasingly used for long-document analysis and legal memo drafting; notable for longer context windows
• Google Gemini: Integrated with Google Workspace, making it a practical tool for document-heavy workflows
• Microsoft Copilot: Appearing in firm environments through Microsoft 365 integration; relevant for students heading into BigLaw

Law School AI Labs and Clinics

Several schools have created dedicated infrastructure for AI exploration:

School	Initiative	Focus
Stanford Law	CodeX – The Stanford Center for Legal Informatics	AI governance, legal tech R&D
Harvard Law	Library Innovation Lab	Access to justice, AI and legal research
Georgetown Law	Institute for Technology Law & Policy	AI regulation, policy
Suffolk University Law	Institute on Legal Innovation & Technology (ILIT)	Practical legal tech training
Vanderbilt Law	Program on Law & Innovation	AI in practice
Chicago-Kent	Center for Access to Justice & Technology	AI for underserved populations

The Ethical and Professional Responsibility Challenges

What the Rules of Professional Conduct Say…and Don’t Say

The American Bar Association’s Model Rules of Professional Conduct were not written with generative AI in mind. But several existing rules apply directly to AI use, and bar associations are racing to issue formal guidance.

ABA Model Rule 1.1 Competence: Lawyers must maintain competence, which the ABA has interpreted to include understanding “the benefits and risks associated with relevant technology.” As AI becomes a standard tool in legal practice, competence arguably requires knowing how to use it—and knowing its failure modes.

ABA Model Rule 1.3 Diligence: Relying on an AI tool without verification may violate the duty of diligence. The Levidow case made this concrete: an attorney cannot outsource case citation verification to an AI and then disclaim responsibility for errors.

ABA Model Rule 3.3 Candor Toward the Tribunal: Submitting AI-generated false citations, knowingly or unknowingly, potentially violates the duty of candor. Courts are now explicitly asking attorneys to certify whether AI was used in brief preparation.

ABA Formal Opinion 512 (2024): The ABA issued Formal Opinion 512 in 2024, providing guidance on generative AI use. Key points:

• Lawyers must take reasonable steps to verify AI-generated content
• AI use does not reduce lawyers’ professional responsibility obligations
• Supervision of AI tools is analogous to supervision of non-lawyer assistants
• Disclosure to clients of AI use may be required, depending on the engagement

Law schools are now incorporating these ethics obligations directly into professional responsibility courses, formerly a somewhat dry coverage of rule-book provisions that now has urgent, contemporary relevance.

Academic Integrity Policies Are Being Rewritten

The standard academic integrity policy says, “Don’t plagiarize, don’t cheat,” which was written for a world where cheating meant copying another human’s work. Generative AI scrambled those categories.

Law schools are now developing much more nuanced policies that distinguish between:

• Prohibited use: Submitting AI-generated work as one’s own without disclosure; using AI on explicitly AI-free assessments
• Permitted use with disclosure: Using AI for brainstorming, outlining, or initial drafting when disclosed and the student has substantially revised and is accountable for the final product
• Encouraged use: Using AI-powered legal research tools as one would use Westlaw—as a resource, not a ghostwriter
• Required use: Some clinics and practice-oriented courses are beginning to require students to use AI tools and demonstrate proficiency

This spectrum represents a significant evolution from the blunt “no AI” policies issued in panic in early 2023.

Impact on the Bar Exam and Licensing

The NCBE Is Watching and Moving

The National Conference of Bar Examiners (NCBE), which develops the Uniform Bar Exam (UBE), has acknowledged the AI moment explicitly. The bar exam tests competencies that the legal profession has identified as essential, and the question of whether AI changes those competencies is now actively under study.

The NextGen Bar Exam, which the NCBE is phasing in beginning in 2026, was already in development when ChatGPT launched. Its designers have been attentive to how AI changes the baseline skills a new lawyer needs.

Key considerations:

• Legal research is already tested differently: The NextGen Bar Exam de-emphasizes rote memorization of doctrine in favor of the ability to apply legal reasoning to novel fact patterns. It is a skill less easily replaced by AI.
• Written component: Bar essays test the ability to construct a legal argument under time pressure without AI assistance, which remains a core attorney competency
• Practical skills: Jurisdiction-specific performance tests (MPT) test document drafting from provided materials—skills that AI changes but does not eliminate

What Law Schools Are Doing to Prepare Students

Ironically, the bar exam’s AI-free environment means law schools must still teach students to do legal analysis without AI scaffolding, even as they teach AI proficiency.

The result is a two-track preparation model emerging at many schools:

1. Bar preparation track: Mastering doctrine, analytical frameworks, and argument construction under timed, AI-free conditions
2. Practice preparation track: Developing AI proficiency, verification skills, and professional judgment about when and how to use AI tools

Students who can move fluidly between these modes of thinking rigorously without AI and then use AI to amplify that thinking will be the most competitive graduates.

What Law Students Need to Know Right Now

If You Are Currently in Law School

1. Understand your school’s AI policy precisely. Policies vary dramatically, even course by course. A professor who explicitly permits AI-assisted drafting in one seminar may teach in the same building as a professor who treats any AI use as academic dishonesty. Read every syllabus. When in doubt, ask.

2. Learn the legal research AI tools, not just general AI. ChatGPT is not a legal research tool. Westlaw, LexisNexis, and their AI-integrated interfaces are. Employers at law firms will expect you to know the difference and to use professional-grade tools that provide citation verification.

3. Verification is now a core competency. Every piece of AI-generated legal content must be verified against primary sources. This is not optional. It is a professional responsibility. Practice doing it until it is automatic.

4. AI literacy is a hiring signal. Law firms are increasingly listing “familiarity with AI legal tools” in associate job postings. Summer associate programs at major firms include AI tool training. Being able to speak intelligently about AI in a callback interview is a competitive advantage.

5. Document your AI use during law school. Develop a habit of noting when and how you used AI tools in your academic work (where permitted). This will help you articulate your approach to AI in interviews and will prepare you for client disclosure obligations in practice.

If You Are Applying to Law School

Ask these questions in admissions interviews:

• Does the school have a formal AI literacy curriculum?
• Are there dedicated AI and law courses, and are they available to 1Ls?
• What is the school’s relationship with legal tech platforms like Casetext/CoCounsel or Lexis+ AI?
• How is the school addressing AI in its legal writing curriculum?
• Is there an AI or legal tech clinic?

Schools that can answer these questions specifically are ahead of the curve. Schools that respond vaguely or defensively may not have caught up yet.

The Future of Legal Education and AI

What the Next Five Years Look Like

AI will become infrastructure, not a novelty. Within five years, AI-assisted research will be as normalized in law school as Westlaw access is today. The question won’t be “should students use AI?” but “are they using it with enough sophistication and critical judgment?”

New specializations will emerge. Just as law schools developed specializations in IP, cybersecurity, and data privacy as those fields matured, AI law is becoming a discipline. Expect to see more LLM programs, clinics, and joint degree programs (J.D./M.S. in Computer Science; J.D./M.S. in Data Science) catering to lawyers who want to work at the intersection of AI and law.

Assessment will evolve. The traditional law school exam, which is a timed, closed-book essay, was already under pressure before AI. Expect more oral examinations, open-book exams with explicit AI parameters, portfolio-based assessment, and clinic evaluations that test judgment over memorization.

Access to justice implications will shape the field. One of the most significant long-term questions is whether AI can democratize legal services. If AI tools allow non-lawyers or under-resourced legal aid organizations to provide higher-quality assistance, law school curricula will need to grapple with the implications for the unauthorized practice of law, for the role of lawyers, and for what legal education is preparing students to do.

Regulatory law will be essential knowledge. The EU AI Act, emerging U.S. federal frameworks, and state-level AI legislation mean that lawyers in virtually every practice area will need baseline AI regulatory literacy. Law schools that treat AI governance as a specialty rather than a general competency may be underpreparing their graduates.

FAQs: AI in Law School

Q: Can law students use ChatGPT for legal research? A: It depends on the school and course policy. However, even where AI use is permitted, ChatGPT should not be used as a primary legal research tool because it is not connected to verified legal databases and is prone to fabricating case citations. Law students should use AI-integrated platforms like Westlaw or Lexis+ AI for actual legal research.

Q: Are law schools banning AI? A: Some courses and professors have banned AI use for specific assignments, but outright school-wide bans are rare and increasingly untenable. The trend is toward nuanced policies that distinguish between prohibited, permitted, and encouraged uses of AI depending on the learning objective.

Q: Will AI replace lawyers? A: The professional consensus is that AI will transform legal practice but not eliminate lawyers. Tasks involving judgment, strategy, client relationships, courtroom advocacy, and novel legal questions remain distinctly human. However, AI will likely reduce demand for certain entry-level tasks, meaning law graduates need stronger differentiated skills to remain competitive.

Q: How is AI affecting the bar exam? A: The bar exam currently does not allow AI use. The NextGen Bar Exam (phasing in from 2026) is designed around competencies that are more durable against AI displacement—legal reasoning, analysis, and judgment rather than doctrinal memorization. Law schools must still prepare students to perform without AI assistance on the bar exam.

Q: Which law schools are best at teaching AI? A: Stanford (CodeX), Harvard (Library Innovation Lab), Georgetown (Institute for Technology Law & Policy), Vanderbilt (Program on Law & Innovation), and Suffolk (ILIT) are among the most recognized leaders. However, many regional and state law schools are also developing strong programs. Prospective students should research specific courses, clinics, and faculty research agendas rather than relying solely on rankings.

Q: What should I look for in a law school’s AI curriculum? A: Look for dedicated AI and law courses available to 1Ls, faculty conducting research in AI governance or legal technology, institutional partnerships with legal research platforms, AI-integrated clinics, and clear and thoughtful AI use policies that reflect genuine engagement with the issues rather than panic or dismissal.

Q: Is AI proficiency required for law firm jobs? A: Increasingly, yes. Major law firms are investing heavily in AI tools and expect associates to use them. Firms, including Allen & Overy (now A&O Shearman), Clifford Chance, and many AmLaw 100 firms, have either deployed or announced AI tools for associate use. Graduates who can demonstrate AI literacy and sound professional judgment about AI use will have a competitive advantage.