Berkeley Law’s Artificial Intelligence Policy, effective summer 2026, helps students by providing clear default guidance about what is permissible and impermissible, while preserving room for instructors to allow AI use when it fits the goals of the class.
Is Berkeley Law banning AI?
No.
What is the goal of the new policy on the use of AI at Berkeley Law?
The goal is to do two things at once: teach students how to use AI effectively and make sure that exams, papers, and other assignments measure the student’s own work.
Students need to learn how to use AI, but they also need to learn how to think, write, and analyze as lawyers. AI can help lawyers do many things more effectively, but it cannot replace the judgment we are trying to teach and evaluate. When a student is evaluated on an exam or for a paper, it should be for that student’s work. We don’t want the student to be graded by turning in the work of another student, a lawyer, or a generative AI tool.
The policy adopted by the faculty seeks to give students and instructors clearer guidance about what is allowed and what is not.
Can faculty allow AI use?
Yes, any faculty member can adopt a different policy that fits the goals and format of their class, simply by providing students with written advance notice.
That flexibility matters because classes have different goals. In one class, the point may be to assess whether students can reason through a problem on their own. In another, the point may be to teach students how to use AI carefully and ethically, or to pilot ideas for how to expand access to legal services using technology.
The default rule is there for clarity. Faculty can depart from it when AI use makes sense for the class.
How does this policy help prepare students for their careers, as AI grows as a force in the legal field?
Berkeley Law is on the vanguard of law and technology. The school has incorporated AI into legal writing classes, offers courses about AI and the law, and has a broad set of law and technology courses taught by full-time faculty and lecturers from practice. In fall 2026 alone, we will offer 22 law and technology courses. Berkeley Law also has 12 technology-law-affiliated student groups.
As Berkeley Law’s Artificial Intelligence hub page shows, the school’s faculty, degree programs, research centers, and executive education programs are helping equip lawyers with the knowledge and skills to lead in this changing legal landscape. The page also highlights how Berkeley Law’s AI policy keeps thinking at the center of good lawyering, while its AI course offerings give students the toolbox they will need as practitioners.
Dean Chemerinsky has also announced the formation of a Berkeley Law AI Leadership Committee. The committee will focus on preparing students for a profession that is changing quickly, supporting the responsible use of AI in legal education, advising on policy and guidance as AI evolves, connecting faculty and centers already working in this area, and strengthening Berkeley Law’s leadership in AI and the law.
Students need to graduate with the cognitive capacity to use AI effectively, but not as a substitute for their own exercise of legal analysis, problem-solving, and judgment. With the fundamental skills to think for themselves as lawyers, students will be better prepared to deploy and evaluate AI tools where appropriate.
How was the policy developed, and what concerns shaped it?
The faculty adopted an AI policy two years ago. We received many comments on it from students and faculty, especially as to areas that were not clear or affected by rapid changes in AI.
A faculty committee was formed and proposed the new policy. It was discussed at a faculty meeting on May 6 and adopted.
Experience under our old policy shaped the need for greater clarity and thus the examples used. The prior policy raised difficult interpretive questions, including distinguishing AI grammar checking from AI composition. It also raised concerns about students presenting existing ideas as their own without attribution, perhaps not recognizing that AI tools had recreated other authors’ ideas.
Why require students to complete some work without AI?
We want to be grading the students’ work.
Students who are able to perform key legal tasks without the support of AI will understand the fundamental principles of effective lawyering. Without those basic cognitive and analytical skills, they will not be equipped to provide value to their clients and the legal system, including by using AI effectively and avoiding AI-induced hallucinations, plagiarism, and other mistakes.
How will policy violations be detected and adjudicated, and what are the consequences?
Violations of the policy are a student conduct violation and will be handled the same as all student conduct violations under the Academic Honor Code.
Detection will depend on context and circumstances. For example, citation to cases that don’t exist is a strong indication of use of AI. It is the student’s responsibility, as it is for any lawyer, to check citations to make sure they are accurate. Courts have sanctioned lawyers for citing non-existent cases or quoting non-existent passages. Enforcement of the rule will depend on the facts of each situation and will be assessed on a case-by-case basis.
Instructors are also adopting assessment techniques that put a premium on independent student thinking, including closed-book exams without access to the internet or computer files, oral presentations, and iterative writing projects requiring students to explain their work process.
What should students do if they are unsure whether a tool or use is allowed?
Students should ask the instructor.