Using tutors to integrate AI into legal education

In an age of powerful large language models, what use are traditional legal tutors? That's not such an easy question. Many law schools use senior law students to assist faculty in teaching first year students difficult material. The idea is primarily to provide first years with an accessible and non-threatening opportunity to get help from students who, while perhaps not at the professor level in terms of knowledge, nonetheless know enough to provide structure, organization, and dispel the more common misconceptions. It also provides the talented students serving as tutors a chance to learn by teaching others, a time tested technique.
But in an era where challenged students can get high quality help from AI, which is available 24/7 and also not threatening, are student tutors really needed anymore? The base knowledge contained by ChatGPT, Gemini and the gang of high end large language models is often enough to answer questions as well as student tutors (and possibly professors at times). The advent of "study and learn" mode from ChatGPT and Gemini roughly emulates the soft Socratic method so beloved by law faculty because it forces the learner to do the work rather than just be a passive recipient of knowledge. The ability to feed course documents to large language models means that their responses are "grounded" and less prone to hallucination than they would be otherwise.
I guess, however, I am not willing yet to give up on student tutors without a fight. I'm not yet willing to relegate my students to either a Cubby subscription or AI without human instruction or support. Students tend to like the tutors (more than me sometimes). And I will confess to some ego gratification from developing a cadre of constitutional law proto-scholars groomed by me to think in the "right way" about the subject.
This spring, I plan to run an experiment in my constitutional law class that redefines the role of student tutors—one I suggest other faculty should consider as well, hence this blog post. Rather than serving primarily as domain experts, these tutors will act as AI coaches. For this role, emotional intelligence is a key qualification alongside a solid understanding of the law, since true emotional connection remains difficult for AI to emulate or for many humans to accept from AI. (But cf. this.) Here's the job description I just posted.
Job Title: Constitutional Law Tutor
Position Overview
This role moves beyond traditional tutoring to place you at the intersection of legal education, technology, and mentorship. As a tutor, your primary responsibility is not to re-teach doctrine but to act as a strategist and guide. You will coach 1L students on the process of learning, empowering them to use cutting-edge AI tools—including a custom AI-powered essay evaluator—responsibly and effectively. You will also serve as an essential source of mentorship and support, helping 1Ls build the skills and resilience needed to succeed in law school. This position offers a unique opportunity to develop skills in leadership, pedagogy, and legal technology that are increasingly valued in the legal profession. This is not to say that you can not or should not assist students with constitutional law doctrine; you definitely should. But the primary goal is to equip students to learn the doctrine actively.
Key Responsibilities
- Lead AI Onboarding: Co-lead an "AI for Law Students" bootcamp that trains 1Ls in essential skills like prompt engineering, building a personal knowledge base with tools like NotebookLM, Study and Learn mode, and critically evaluating AI outputs.
- Facilitate "Skills Gyms": Plan and lead weekly, interactive workshops focused on higher-order skills such as issue-spotting, argumentation, and legal analysis. These sessions can use current events as a vehicle for demonstrating the relevance of the material under consideration.
- Coach AI-Powered Writing Improvement: Guide students in using an AI-powered essay evaluation tool. Your role will be to help them interpret the tool's feedback, identify patterns in their writing, and develop concrete strategies for improvement.
- Promote Responsible AI Use: Actively teach students to identify the limitations of AI, spot errors, and develop a healthy skepticism that prioritizes critical thinking and their own analytical abilities.
- Serve as a Peer Mentor: In one-on-one office hours and group settings, offer guidance, empathy, and encouragement to students navigating the stresses of the 1L year.
Why This Role is a Valuable Career Move
This position is designed to provide you with tangible skills and experiences that will enhance your own knowledge and marketability.
- Develop Demonstrable AI & Legal Tech Skills: You will gain practical experience using and coaching others on legal technology. This provides you with a valuable credential and talking point for interviews, demonstrating your fluency with the tools that are reshaping legal practice.
- Gain Concrete Leadership and Management Experience: Planning workshops, leading group discussions, and mentoring 1L students are valuable leadership activities. You are not just a teaching assistant; you are a coach managing a cohort of developing professionals.
- Achieve True Mastery of the Subject Matter: The best way to deepen your understanding of a complex subject is to teach it. Coaching others on how to spot issues and structure arguments in Constitutional Law will sharpen your own analytical skills to a level that will benefit you in advanced courses and on the bar exam.
- Refine Your Communication and Feedback Skills: Learning to give clear, constructive feedback is a critical skill for any lawyer, whether dealing with clients, colleagues, or junior associates. This role provides a structured environment to hone that ability.
Qualifications
- A 2L or 3L student who has excelled in constitutional law. There will be a preference for students who were enrolled in Professor Chandler's course but this is not a strict requirement.
- A student with high emotional intelligence, patience, and a genuine desire to help peers succeed.
- An individual with a strong curiosity about the intersection of law and technology.
- An excellent communicator who is comfortable leading group discussions and providing constructive feedback.
We will see what happens. I am very hopeful it will excite students for precisely the reasons stated in the job description. If I were a hiring partner, I would be extremely impressed by a student who had not only learned constitutional law enough to receive the honor but had also learned how to integrate modern legal tech that is becoming pervasive in practice. Still, some students who had excelled in constitutional law and might feel particularly motivated to teach separation of powers or the first amendment during these challenging times – may be disappointed by the shift in roles and not apply for the position at all. It's possible I will lose some favored student tutors who might otherwise challenge themselves with office hour discussions of remedies in Establishment Clause cases or the tensions between the 15th amendment and new notions of race blindness in cases involving the Voting Rights Act. And some students might well prefer the work my tutors did for them in years past: organizing presentations on matters covered in class (perhaps at lightning speed). Some students, I expect, may see being steered to law AI as the equivalent of being put into some hideous phone tree during a customer service adventure. And some, despite my repeated warnings against dependency and passivity in AI learning, will treat "relegation" to AI as an invitation to "learn" passively, falsely believing that seeing tokens roll across a chatbot screen is the equivalent of having wrestled with the material for oneself. I will urge the tutors to warn daily of the risks of misusing AI in that way. Until Mark Zuckerberg's Ray-Bans get a lot better or lawyers all get chipped, people in the legal profession still have to have legal materials actually in their brains.
There is also the matter of personal vulnerability. If knowledge-focused student tutors can be replaced by AI coaches, where does that leave the professor? I cling to the belief that my human colleagues and I offer something irreplaceable—that students can only fully benefit from AI within the framework that human professors provide. Simply giving students a course outline and saying "have fun with ChatGPT's study mode" would likely fail many students, though I suspect some would thrive with this approach, especially if their professor wasn't a strong teacher. (I've certainly learned a lot from conversations with AI myself.) While I don't like justifying my job this way, I worry that for many students, relegating instruction entirely to AI would feel like a return to COVID-era remote learning—where the loss of personal connection and human guidance often diminished real understanding.
Will changing the role of the tutor succeed? I'm starting this experiment with my usual enthusiasm. Ask me in May 2026 about the results.
Note: A friend with whom I discussed this post feared that AI might give students the wrong answer with the beguiling confidence that only the very best models can display. It might indeed. But I am afraid that is also the case with student tutors. Moreover, students will be taught during the "boot camp" described in the job description to use grounded AI such as NotebookLM or RAG systems. Under those circumstances, the probability is low that 2026 state-of-the-art AI will hallucinate on the sort of general issues of mainstream American constitutional law that I cover in my class.