Introduction
The development and deployment of artificial intelligence (“AI”) is rapidly changing the study and practice of law. AI tools are now capable of completing complex tasks and generating content with previously unthinkable speed and accuracy. These tools have become firmly embedded in the work and learning that lawyers and law students do every day—e.g., using AI to conduct legal research, test arguments, perform due diligence, and draft case summaries and memoranda.
It is impossible to predict how AI will continue to evolve. Nonetheless, one thing is certain: AI is driving a fundamental shift within the legal profession. Columbia Law School is well positioned to innovate during this period of transformational change. Our location in one of the world’s hubs of innovation, investment, and public service; our long-standing connections to the practicing bar; the intellectual strength of our faculty with expertise at the intersection of law and technology; and the curiosity and engagement of our exceptional students, all set us apart. The Law School also draws from the considerable strength of Columbia University, which is at the forefront of AI research and applications.
The Law School’s approach to AI is grounded in the unique purpose of a legal education, which includes: (1) teaching students how to think critically, deeply, and rigorously; and (2) preparing them for leadership in a rapidly evolving profession.
In Fall 2025, the Law School’s Generative AI Task Force (the “Task Force”) was charged with thinking about how to train students to reason carefully and independently, without relying on AI, while also equipping them with the AI fluency that they will need to be successful in their careers. After consulting widely among faculty, students, alumni, and staff, the Task Force coalesced around the idea that students should be equipped with “the adaptive conceptual frameworks necessary to navigate a shifting landscape, rather than merely train[ed] on specific tools that may be obsolete by the time they graduate.”
This is a compelling notion—even as technological fluency becomes an essential component of legal practice, the effective and ethical use of AI will depend on the quality of the thinking among those who guide its use. Building on the Task Force’s work, we have identified the following guiding principles:
- A rigorous legal education must cultivate critical thinking, deep understanding, careful reasoning, and independent judgment.
- In order to use AI tools effectively, students must develop the intellectual habits of mind and professional capacities that underpin excellent lawyering.
- AI has enormous potential to enhance legal education and legal work, and the Law School has an affirmative obligation to help students learn how to use AI tools responsibly and ethically.
- The Law School will approach AI with intellectual curiosity, humility, and a willingness to experiment. Because AI technologies are evolving rapidly, we will continually assess our policies and practices; endeavor to learn from experience; and refine our approach in collaboration with faculty, students, alumni, and the professional community.
From these principles flows an institutional strategy for incorporating AI both in and outside of the classroom. The Law School does not aim to adopt a blanket ban on AI tools nor do we intend the Policy below to be an unqualified endorsement of its use. Instead, we seek to craft an approach that balances and advances the following goals:
- Continuing to prioritize the elements of legal education that AI cannot replicate, including critical thinking, sound judgment, intellectual curiosity, emotional intelligence, adaptability, resilience, and the ability to analyze complex problems and draw well-reasoned conclusions from a variety of data sources with confidence and clarity.
- Establishing a shared baseline of AI knowledge among students by implementing AI training during new student orientation and offering ongoing workshops and programming for returning students.
- Phasing in intentional AI-related introductory learning opportunities through specially designed AI modules embedded within selected 1L courses and the legal research and writing curriculum.
- Adopting an AI academic policy that provides guidance regarding the use of AI on exams, papers, and other credit-bearing assignments.
- Maximizing flexibility in the classroom by allowing faculty to determine the specific approach to AI that best suits their pedagogical goals and course content, including deciding if and how students are permitted to use technology in class.
- Creating pathways for deep intellectual exploration through specialized AI courses, seminars, and colloquia offered at the Law School.
We expect that students will encounter a range of approaches to AI across the curriculum. Some instructors may choose to incorporate AI tools extensively in their classes, while others may limit or prohibit their use in order to develop particular skills or advance specific learning objectives, particularly in the 1L year. Experiencing this range of approaches is itself an important part of preparing students for the realities of modern legal practice, where they will encounter varied professional norms and technological environments, remaining grounded in the foundation of what it means to think, and act, like a lawyer.
There is no question that this period of transition presents significant challenges and opportunities for the Law School. The technology itself is evolving rapidly, as is our understanding of its uses and limitations in the context of legal education. In such a dynamic environment, the Law School’s goal is not to produce a static set of rules, but to cultivate a community capable of learning and adapting along the way. Since no policy can adequately resolve all questions that will inevitably arise, we anticipate continuing to evaluate and update our norms and policies as we learn together. This is work that we must undertake in a spirit of shared curiosity and ownership.
Indeed, an explicit goal of this Policy is to foster dialogue and discussion, both in the classroom and beyond. We encourage conversations between faculty and students, in particular, and believe strongly in fostering an environment in which our community can think and grapple with AI’s benefits and drawbacks openly and productively.
Policy Text
Effective August 1, 2026, Columbia Law School (the “Law School”) adopts the following default academic policy outlining the use of AI by students in the teaching and learning context for the 2026–27 academic year. This Policy supersedes the earlier Interim Policy on Generative AI in its entirety.
- Statement of Purpose: The purpose of this Policy is to provide a framework to guide the use of generative artificial intelligence (or “AI”) by students in the teaching and learning context. It establishes default expectations in a manner that supports the Law School's educational purpose, foregrounds academic integrity, and prepares students to use AI responsibly and ethically in the study and practice of law.
- Scope: This Policy governs the use of AI by Law School students in all academic work, including, but not limited to, examinations, papers, research memoranda, assignments, clinic submissions, and journal work.
- Definition: AI in this context includes any machine-based tool designed to brainstorm, formulate, ideate, write, edit, translate, or restructure language from user questions and prompts. AI includes stand-alone programs (e.g., ChatGPT, Claude, Copilot, Gemini, and GeminiLM); AI-powered and -assisted legal research tools (e.g., Harvey, Legora, NormAI); custom-built solutions (e.g., using open weight models); and AI tools embedded in programs such as Grammarly or Microsoft Copilot and research platforms (e.g., Lexis, Westlaw, Bloomberg Law).
- Default: The Rules set forth in Sections 8 through 11 of this Policy are considered defaults. Individual instructors may adopt alternative course or assignment-specific policies as long as they do so in writing, at a minimum in the initial course syllabus, and with advance notice to all students. To the extent that a faculty member does not communicate an alternative this Policy will apply.
- Reference to Applicable University Policies
- This Policy supplements the University’s Generative AI Policy, as amended from time to time. The University’s Generative AI Policy governs the use of generative AI tools by faculty, students, and staff in the performance of their functions for or on behalf of Columbia. It is the Law School’s expectation that all faculty, students, and staff are knowledgeable of, and in compliance with, its terms, particularly its data protection rules.
- All uses of AI must be consistent with the University’s policies on confidential and personal information.
- This Policy supplements the University’s Generative AI Policy, as amended from time to time. The University’s Generative AI Policy governs the use of generative AI tools by faculty, students, and staff in the performance of their functions for or on behalf of Columbia. It is the Law School’s expectation that all faculty, students, and staff are knowledgeable of, and in compliance with, its terms, particularly its data protection rules.
Summary Statement
Columbia Law School permits the use of AI as a tool to support learning and to perform specific functions as outlined herein. The foundational requirement of this Policy is that students must bear full intellectual responsibility for all work submitted for academic credit. At every stage of the writing process, the ideas, arguments, analysis, and expression in submitted work must reflect the student’s own judgment and thinking. Students must deliberately choose the arguments they advance, understand and be able to explain and defend every proposition in their submissions, and accept responsibility for their accuracy and conclusions.
AI may serve as a learning aid and critical interlocutor—helping students test ideas, improve organization, or refine expression. AI may not, however, replace the student's own legal reasoning, serve as an undisclosed ghostwriter, or otherwise replace the student’s role in originating the intellectual content of a submission. A student who submits work generated by AI without exercising independent judgment, or who cannot explain and defend the work as their own, has failed to satisfy the expectations of this Policy.
- General Guidelines: The use of AI is subject to the following guidelines:
- Student Responsibilities: Students alone are wholly responsible for the work they submit for credit.
- Students have an affirmative duty to consult their instructor in every course before using any AI tool if there is any doubt whether its intended use is permitted.
- Students are responsible for assessing the value and independently verifying the validity of any AI-generated sources, authorities, and other outputs. If a student uses AI to identify relevant sources or authorities, the student must carefully review those authorities before citing them in written work.
- All submitted work must represent the student’s own deliberate thinking and judgment. Students must be able to explain and defend decisions, both pertaining to the ideas and arguments, as well as how those ideas and arguments are articulated.
- Regardless of whether AI use is disclosed, under no circumstances may a student submit AI-generated work as their own independent work product when doing so materially misrepresents the extent of their own effort and contributions.
- Students have an affirmative duty to consult their instructor in every course before using any AI tool if there is any doubt whether its intended use is permitted.
- Disclosure: Students are required to disclose any use of AI that contributes to a written submission or other work product.
- Students should always be transparent and err on the side of disclosure when their work relies on the output of an AI tool. Individual instructors may prescribe additional requirements, including protocols for citing AI outputs.
- Disclosure serves an educational purpose, in that it enables instructors to understand how students are using AI, provide more meaningful feedback, and engage students in conversations about the appropriate and effective use of AI tools. It also encourages students to reflect critically on when AI enhances their learning, when it may impede the development of important skills, and how to exercise sound judgment in deciding whether and how to rely on AI. By making AI use visible rather than hidden, the Law School seeks to foster a culture of openness, experimentation, and shared learning.
- Students should always be transparent and err on the side of disclosure when their work relies on the output of an AI tool. Individual instructors may prescribe additional requirements, including protocols for citing AI outputs.
- Student Responsibilities: Students alone are wholly responsible for the work they submit for credit.
- Permitted Uses of AI: Unless expressly indicated otherwise by the course instructor, AI is permitted for the following purposes:
- Learning and Understanding: Students are permitted to use AI tools to help learn legal concepts; summarize their own notes or materials; generate examples or hypotheticals; serve as a mooting partner; identify and define legal terms; study for examinations; and otherwise support their comprehension and mastery of course content.
- Research: Students are permitted to use AI tools to conduct research, including identifying potentially relevant authorities; summarizing cases and scholarship; and organizing research results for further investigation.
- Writing Support and Error Correction: Students are permitted to consult AI to test arguments, solicit criticism, and explore alternative articulations. Students may also use AI to improve their written work, including for clarity and reliability, alternate phrasing, and tone and concision. Students are also permitted to use AI tools to assist in identifying and correcting spelling and grammatical errors. In all of these cases, students are required to independently and rigorously assess the feedback generated by AI, using their own critical judgment to decide whether and how to implement any suggestions.
- Learning and Understanding: Students are permitted to use AI tools to help learn legal concepts; summarize their own notes or materials; generate examples or hypotheticals; serve as a mooting partner; identify and define legal terms; study for examinations; and otherwise support their comprehension and mastery of course content.
- Prohibited Uses of AI: Unless expressly indicated otherwise by the course instructor, AI is prohibited for the following purposes:
- Original Legal Arguments and Analysis: Students may not rely on AI to perform the legal analysis, reasoning, or judgment that an assignment is designed to assess.
- Examinations: Except where advance permission is granted in writing by the course instructor in the course syllabus and/or examination instructions, use of AI during any examination is prohibited.
- Every word of a student’s examination answer must be produced by that student without assistance from AI (including for English translation purposes). This prohibition encompasses all examinations, irrespective of format (i.e., whether proctored or take-home, online or handwritten).
- Students are not permitted to input current examination questions, prompts, or fact patterns from any course into an AI tool.
- Every word of a student’s examination answer must be produced by that student without assistance from AI (including for English translation purposes). This prohibition encompasses all examinations, irrespective of format (i.e., whether proctored or take-home, online or handwritten).
- Generation of Written Work: Work submitted for academic credit must be student-composed, not AI-generated. Student-composed work reflects ideas, arguments, analysis, and expression that are the product of the student's own intellectual creativity and judgment. Students must deliberately choose the positions they advance; understand and be able to explain and defend every material proposition in their submissions; and accept responsibility for the accuracy, reasoning, and conclusions reflected in their work. Revising, editing, or paraphrasing AI-generated text does not, by itself, transform AI-generated work into student-generated work. Even when assisted by AI tools, students bear full intellectual responsibility for every proposition in their submitted work.
- Classroom Participation and Presence: Except where permission is granted by the course instructor, students may not use AI for any purpose during a class session.
- Original Legal Arguments and Analysis: Students may not rely on AI to perform the legal analysis, reasoning, or judgment that an assignment is designed to assess.
- Clinics and Externships: Clinics and externships are unique pedagogical experiences that require special consideration with respect to AI use due to the sensitivity of client and other personally identifiable information. In addition to the general guidelines and permitted and prohibited uses set forth above in Sections 7 through 9, the following also govern AI use in all clinic and externship settings.
- Client Work. Except where expressly prohibited by the course instructor (or in the case of an externship, the field placement supervisor) students may use AI in performing client work, subject to the general guidelines in Section 7 above and to the confidentiality requirements below. Permitted use may include studying and deepening understanding of legal concepts, brainstorming, and proofreading for spelling and grammatical errors. In all cases, students remain solely responsible for the accuracy and quality of all work products, and must be able to explain and defend any submission as their own.
- Confidentiality. Confidential information pertaining to client matters may not be input into an AI tool or system without the express permission of the course instructor or field placement supervisor.
- “Confidential information” is defined broadly under the New York Rules of Professional Conduct as information gained during or relating to the representation of a client, whatever its source, that is (a) protected by the attorney-client privilege, (b) likely to be embarrassing or detrimental to the client if disclosed, or (c) information that the client has requested be kept confidential.
- Because determinations of the scope of confidentiality are extremely fact-specific, students should treat all client information as confidential unless faculty, instructors, and/or supervisors have explicitly instructed otherwise.
- “Confidential information” is defined broadly under the New York Rules of Professional Conduct as information gained during or relating to the representation of a client, whatever its source, that is (a) protected by the attorney-client privilege, (b) likely to be embarrassing or detrimental to the client if disclosed, or (c) information that the client has requested be kept confidential.
- Client Work. Except where expressly prohibited by the course instructor (or in the case of an externship, the field placement supervisor) students may use AI in performing client work, subject to the general guidelines in Section 7 above and to the confidentiality requirements below. Permitted use may include studying and deepening understanding of legal concepts, brainstorming, and proofreading for spelling and grammatical errors. In all cases, students remain solely responsible for the accuracy and quality of all work products, and must be able to explain and defend any submission as their own.
- Translation: Effective written and oral communication in the English language is a core competency for law students and practicing lawyers. Columbia Law School’s curriculum, therefore, includes assignments that are designed not only to assess legal analysis, but also to develop the communication skills required for the study and practice of law.
- Translation of Non-English Sources into English: Students are permitted to use translation tools to translate non-English primary sources into English.
- Submission of a Student’s Own Writing Translated from a Language other than English: Students may not use any translation tools to translate their own writing in a non-English language into English for the purposes of submitting this translation as their own original written work product.
- Translation of Non-English Sources into English: Students are permitted to use translation tools to translate non-English primary sources into English.
- Violations: Alleged violations of this Policy should be reported to the Associate Dean of Registration and Student Services (the “Associate Dean”), who will conduct an inquiry to determine whether there is a reasonable basis to believe that a violation has occurred. Where a reasonable basis does exist, the Associate Dean may propose an informal resolution or, in consultation with the Vice Dean for Academic Affairs, refer the matter to the Center for Student Success and Intervention.
Conclusion
As Columbia Law School navigates this period of technological transition, our commitment to a rigorous legal education remains steadfast. This Policy seeks to advance core institutional values—fostering critical thinking and independent judgment—while carefully integrating the AI that are becoming essential to the practice of law. We recognize that the technological landscape is rapidly evolving, and we are committed to an iterative and thoughtful process of review and refinement. Moving forward, the Law School will continue to evaluate our progress and update our guidance accordingly.