It’s Gotta Be The Shoes
First published: August 13, 2026. Posted here for archival purposes.
Everything in Between” is about the systems, institutions, and practices that people build, “things” of a sort that sit in between us, between groups of us, between “us” and “them,” and between us and other systems and institutions that seem terribly far away: “the market,” “the state,” the universe, and so on.
In addition to my posts here, I co-host a podcast titled “Your Leadership Podcast,” which is available on Spotify and wherever fine podcasts are available. I write about law and legal education at TaxProf Blog and for several years co-hosted a podcast about technology and law titled “Your Future Law Podcast.“ My older blog about Pittsburgh and renewing cities, Pittsblog, is still available online, as is my original blog about law, technology, and governance.
Columbia University’s law school recently published the latest in a string of non-blockbuster “AI use” policies for its law students. The Columbia policy is here. It comes on the heels of similar announcements from law schools at the University of Chicago (which offered a grand “AI Strategy Statement” titled “Rethinking Legal Education in the AI Era”), the University of Texas (in a “Dear colleagues” letter from the dean), and the University of California, Berkeley (simply a policy). The dean of the law school at Suffolk University, Andrew Perlman, has collected and archived AI policies and practices across many but not all US law schools. 163 schools, to be specific as I write this, based on public sources (the archive was built largely with AI, of course), out of the 200 or so ABA-accredited US law schools.
I focus on the policies, because almost all of the academic and legal profession reaction to what law schools are and are not doing relative to AI focuses, likewise, on the policies. I’ve called out the four above because they’ve been the most salient in provoking reactions, judging anecdotally from comments on LinkedIn and Substack and even coverage in the New York Times.
As I wade back into writing here about systems and institutions other than soccer/football (FIFA’s post-World Cup shenanigans tempt me, strongly, to stay with that theme, so don’t go anywhere, sports fans!), I have a few comments about what’s happening with AI policies and law schools.
One, and hardly original to me:
The attention to policies misses a great deal of what is actually important to law students, to new law graduates, to law professors, law schools, and therefore to the legal profession as a whole, and to the communities and organizations that rely on law and lawyers. Which is: practice, meaning the many different ways in which law schools and law professors might introduce students to AI, explore AI with students, help students see the powers and limits of AI, and so forth. Policies have only so much weight relative to student and faculty behavior.
Individual behavior is one thing; collective and institutional behavior is yet another. On the school and faculty side of the proverbial ledger, and beyond or behind a give statement or policy, what are the institutional commitments to build and support teaching practices and research around AI? “How is the policy enforced” is one thing, but only a start. Institutional commitments might reach much more deeply (or not); they might amount to money and time and faculty and other effort directed to courses and experiences that embrace AI (either in a curriculum that is “about” AI or in a curriculum that weaves AI into other material, or both), or that keep AI at arm’s length or beyond, or (more likely, I suspect) that blend of those things by context.
In the abstract, policies are parsed and critiqued for being (at times) too restrictive. Berkeley’s policy seems to be the strictest version of “No AI For You,” a la Seinfeld’s “Soup Nazi.” Or, at times, too generous and undisciplined. Even Berkeley’s policy, like most, allows for opt-outs by individual faculty members. What Scylla takes away, Charybdis gives.
I am as aware as any university professor that organizing and implementing resource strategies to back up institutional commitments pose different and more complex problems for academic leaders (deans, but also motivated faculty voices) than nudging faculty members into formal alignment with an institutional policy. A university is something like an aircraft carrier: a multi-purpose system of systems, technologies, purposes, and people that is enormously powerful but that changes course only slowly, never all in the same direction, and never on the proverbial dime. (Also, some argue that aircraft carriers are obsolete in an era of smart weapons. That’s a conversation for a different time.) The faculty opt-out strategy is obviously key to getting faculty to accept the conduct and outcomes of a policy-making process, for pragmatic as well as ideological reasons, but that process requires (only) a lot of talking with colleagues and (if it’s done well) with present and former students. Changes of institutional course often require partnerships with provosts, and advancement offices, and community partners.
Which is why this initiative at UC Law SF (the San Francisco law school formerly known as the Hastings College of the Law, part of the University of California) is distinctive. The law school will require that all JD students complete an “AI-enabled lawyering lab.” Observers may disagree about the wisdom of that strategy (personally, I think that it’s a good thing). What impresses me is the fact that the school is putting time and money behind its formal vision. UC Law SF is a stand-alone law school.
Given what may turn out to be (and what I think are likely to be) disconnects in practice between what the policies permit and require and what actually happens in classrooms and among students, it will be interesting to watch patterns of evolution and change at both levels.
Two, and entirely about me:
My own law school (Pittsburgh) is not represented in the archive produced by Dean Perlman at Suffolk Law, largely because our law school has no institutional policy on AI use by students (or others, including administrative staff). To the best of my knowledge, there is no movement among my colleagues to produce one. That’s not a considered forward-looking strategy; it’s inertia. The University of Pittsburgh culture often takes after City of Pittsburgh culture, which I once characterized as “don’t just do something. Stand there.” Both Pitt and Pittsburgh are no longer as stuck in the cultural mud as they once were. The AI economy is as difficult to overlook locally as it is nationally and internationally. But neither Pitt nor Pittsburgh has the institutional dynamism or flair of Silicon Valley or New York or even many cities of Pittsburgh’s own size.
Yet as every teacher at almost every level and in almost every institution knows by know, student access to and use of AI systems is nearly universal.
So I wrote my own policy. “Opt-out” permissions even at schools that have adopted institutional rules mean that I could do this, I think, if I were teaching at Berkeley, Texas, Chicago, or Columbia. I will be curious to learn, if I learn, about the range of opt-out settings adopted there. There is an archive to be built of individual faculty practices, at law schools and elsewhere.
I teach upper-level elective courses only. They are relatively small scale. I don’t give exams of any sort. I have long permitted my students to consult any material and any person of their choosing. I am not trying to assess what my students know; I am trying to assess their progress in becoming legal professionals. So I treat them that way: as junior legal professionals. AI will be a part of their professional experience (and if they are employed in part-time legal work during law school, in all probability AI is part of that experience); therefore it should be a part of their apprenticeship, meaning law school. So the TL;DR version of my policy is this: students may use the AI systems that they wish to but are expected to disclose the character of their use in some meaningful detail. The policy is new for 2026-2027, so I do not yet have answers to some obvious questions: how much detail? What happens if I deem the detail to be inadequate? What happens if they use AI but don’t disclose?
It’s an experiment.
Three, looking ahead, and again, not inventing a theme but bringing it into relief in my field:
One obvious explanation for all of the attention to policy rather than practice, both among law professors and their critics, is that addressing AI via policy insulates the faculty and the practitioners from the sense that we (the collective “we,” meaning the multiple wings of the profession) don’t in fact know what we’re doing. We did, but now we don’t.
I mean that – “we don’t know what we’re doing” – in a nearly literal sense, and I also mean it mostly in a positive, optimistic sense.
One key thread linking statements by Berkeley, Texas, Chicago, and now Columbia is a strong and unbreakable commitment to protecting and privileging the critical cognitive skills of lawyers-to-be. The commitment is not limited to law and lawyers and law schools, but it is especially sharp in a field that has signed on – for decades – to the principle that the purpose of law school is to teach its graduates to “think like a lawyer.” Folks as old as I am will remember this legendary line from the fictional Professor Kingsfield: “You teach yourselves the law, but I train your minds. You come in here with a skull full of mush, and you leave thinking like a lawyer.”
Shaping mush into discipline is what law professors do, even if we don’t do it with the ruthlessness of Harvard Law School in the 1970s and before. (Kingsfield was a piece of work, but by all accounts his questioning style was not unrepresentative of the real Harvard Law School.) It’s what law professors know how to do. For the most part, it is both what law professors take pride in doing and pretty much all that law professors know how to do. “We have to keep AI from corrupting that discipline-forming process” is also “we have to protect our field, and our jobs, careers, and identities,” even if that latter statement is a difficult truth.
AI policies are the tips of institutional icebergs. Below the waterline are big questions: what should law schools and law professors do, looking forward, given AI? Stay the cognitive course? Steer in additional directions? Who does the steering, keeping the course or changing it, and when and where, and at what pace? If any specific school is a proverbial aircraft carrier, what are its resources, purposes, and missions, and personnel? To whom is it accountable, and how? I have a recent post up on the TaxProf Blog that elaborates somewhat on what I called existential themes.
Those are not rhetorical questions, at least not entirely. Whether they are rhetorical or real, they are difficult questions, with no simple or obvious answers. I hear them asked rarely, among my colleagues. (And I have not even waved at the finances of law schools. Over here, I did.) And to be clear, I am using law here only as a case study. The entirety of the higher education edifice is facing one version of these questions or another, or perhaps all of them. And the only way forward is, as Anne Lamott famously wrote, bird by bird. One step at a time, in purposeful action in multiple directions, even purposeful re-commitment to classic values and practices. Policy isn’t enough.
Commentators of a certain age might get to this point and yield the floor to Pogo: I have met the enemy and he is us.
But I am more optimistic than that. The university’s decentralized character, its multi-purpose-aircraft-carrier-character, is its superpower. If it has a superpower. Where do we go? What do we do? Like Dorothy in the Wizard of Oz, we – the collective we, the citizens of the department, the center, the school, the college, and the university – have the power to decide, if only we seize it and exercise it.
Years ago, talking with a senior, experienced, and widely-respected law professor friend, I explained that I had given up the traditional law school practice of assessing my students via single comprehensive end-of-semester “issue spotting” exams. Instead, during the semester I gave them (and give them) a series of written assignments that are meant to test my students’ abilities to form themselves into junior lawyers, dealing with incompleteness and ambiguity. (Anyone can read my assignments for themselves at my course pages, and anyone can read my explanation and justification here.) I judge their work accordingly. I was a practicing lawyer for a long time myself, and I’ve been a client. I think that I have a sound basis for evaluating the early career “lawyering” capabilities of my students.
My law professor friend replied, in all sincerity and in good faith, mustering a career’s worth of dignity, pride, and expertise: I can’t do that. I don’t know how to do that.
It’s an imaginative leap from that anecdote to this one, but stay with me, please:
Way back in the late 1980s, Nike launched an ad campaign built around Michael Jordan and directed by (and co-starring) Spike Lee. The gag was a blend of Spike Lee’s performing in his Mars Blackmon character from “Do the Right Thing” and Mars’ relentless insistence that Michael Jordan’s scoring prowess was based on the Nike “Air Jordan” shoes that he wore.
Rather than on, say, MJ’s relentless drive not only to perfect his talent but also to change the way that the game was played. Basketball before MJ was one thing; basketball after MJ has been another thing entirely. Michael Jordan is not the only player to have changed the sport, but he did change the sport.
AI.
Changing what teachers and schools do? Or do we change ourselves?
It’s gotta be the shoes.
The original commercial is too good to omit.
Thanks for sticking with me.