Harvard Law School doesn’t need to endorse AI to prepare students for it.
The rise of artificial intelligence has been a non-negotiable reality, with its use infiltrating a range of industries and professions. The legal profession is no exception: According to two recent surveys, 41 percent of law firms report active generative AI use, while 83 percent of individual lawyers report using AI tools. However, professional adaptation need not demand educational integration — and HLS should draw a firm line between the two.
As its newly convened AI committee navigates the technology’s place at the law school, it should steer a strict course, keeping AI out of the classroom to protect the independent reasoning, writing, and judgment that traditional legal education seeks to cultivate.
The case for that boundary begins with the transformative power law school is supposed to bestow.
Law school is a unique intellectual experience, and those granted the opportunity to pursue a legal education should be expected to engage fully in the reasoning it demands. Law school trains students to reason through ambiguity — to sit with dense, indeterminate, and edifying cases, extract their holdings, weigh competing interpretations, and construct arguments of their own.
The struggle and contemplation are not unfortunate byproducts of learning the law; rather, they are the very processes needed to understand it. In the same vein, legal reasoning requires more than the surface-level summaries AI can regurgitate. When done well, reasoning helps students discover gaps in their own interpretations and understanding.
These are precisely the features of a legal education that AI threatens to short circuit. AI can summarize cases, identify arguments, synthesize doctrine, and produce or refine legal writing almost instantaneously. When students feel welcome to outsource those practices, they bypass the intellectual exercise meant to make them into thoughtful lawyers.
Even if you take this to be the musings of a Luddite, many of Harvard’s peer institutions have sounded similar alarms about AI. MIT’s recent report on AI and education promotes “leaning into learning,” while warning that chatbots can trigger “cognitive surrender.”
Law schools have begun putting similar concerns into practice. University of California, Berkeley School of Law now forbids AI use for conceptualizing, outlining, drafting, revising, translating, or editing any work submitted for credit. Similarly, the University of Chicago Law School is piloting restrictions on technology across its core first-year classes, banning electronic devices from classrooms to preserve the “effortful struggle” it views as essential to legal education.
Rather than just deliberating on AI’s place at HLS, the committee should take an active role in prohibiting AI from interfering with foundational learning. At the very least, it should work to insulate first-year core classes from AI outsourcing, preventing students from developing the habit of turning to a chatbot on arrival.
This is not to say that the advent of AI should be completely ignored within the walls of HLS. HLS could, for example, offer periodic AI proficiency workshops to illuminate how current lawyers use the technology to assist their work in efficient ways.
The key is to separate AI proficiency from curricular courses and law school work, preserving the distinction between learning how one could use AI in legal practice and relying on AI to complete the work through which students learn the foundations of law.
We must also confront the fact that legal education carries higher stakes by way of being a profession. The reasoning and doctrine students are taught are practice for work with tangible consequences. Whether it’s a client’s freedom, a company’s liability, a family’s custody arraignment, or more of the like, legal education prepares students to bear real ramifications for decisions made on someone else’s behalf.
Mastering prompt engineering is no substitute for the independent judgment and articulation that legal practice demands.
We should not risk AI interfering with the foundational knowledge and skills needed to do these jobs effectively and authentically. Once those foundations are secure, graduates can decide when and how to enlist AI in their work.
In other words, law school should teach students how to think like lawyers before AI teaches them how not to.
Ava M. Ribaudo ’27, a Crimson Editorial editor, is a Government concentrator in Leverett House.
Want to keep up with breaking news?Subscribe to our email newsletter
Have a tip for The Crimson?Share it confidentially
