Chicago Law Dean on AI: You Won’t Be Able to Entirely Avoid It

Like most schools worried about academic integrity, the University of Chicago Law School used to discourage its first-year students from using generative AI but now it has crept into that first-year curriculum. Despite its overall inevitability, William Hubbard, a professor and deputy dean, says he’s surprised by how often he has to encourage AI-skeptical law students to at least try it out.

_{Listen here and subscribe to On The Merits on Apple Podcasts, Spotify, Megaphone, or Audible.}

Hubbard’s school has seen how law firms, and especially large law firms, have embraced this new technology and it’s followed suit, adding several AI-focused classes to its course offerings. Hubbard says the University of Chicago’s students need to graduate with at least a basic familiarity with AI—specifically when it is and isn’t appropriate to use in a legal setting.

He spoke to Bloomberg Law editor Jessie Kamens for our podcast, On The Merits, about what the legal industry wants law students to learn about AI and how his law school is going about teaching it.

Do you have feedback on this episode of On The Merits? Give us a call and leave a voicemail at 703-341-3690.

This transcript was produced by Bloomberg Law Automation.

Host (Jessie Kokrda Kamens):

Hello, and welcome back to On the Merits, the news podcast from Bloomberg Law. I’m your host, Jessie Kamens. It’s back to school time, and so we’re going to be bringing you a couple of episodes about law students, and specifically, what law schools are teaching them about AI.

We’ve all heard stories about lawyers getting in trouble when they file an AI-written brief that’s riddled with errors and fake citations. Today’s guest wants the next generation of attorneys to know enough not to make these kind of mistakes.

William Hubbard is a deputy dean at the University of Chicago Law School. In the past few years, his school has created several new courses on AI and the law. But the curriculum isn’t just about warning students about AI nightmares. Law firms, especially large law firms, are embracing this new technology in a big way. And that means law students will need to be comfortable using it before they arrive at the firm.

That’s why Hubbard says he’s surprised that despite the stereotype of young people being tech-fluent, some of his students are AI skeptics who need to be convinced to even try it out. We’ll get to that in a bit. But first, he talks about why the University of Chicago has been so quick to add AI to its curriculum.

Guest (William Hubbard):

We wanted to move very quickly to be responsive to what we saw as something that was going to be more than a short-term phenomenon, something that had the potential, at least, to fundamentally change how legal practice is done. And we were able to identify alumni and friends of the law school who were really well-positioned to teach our students about AI.

We have instructors who are experts in legal writing and advocacy who have really embraced the use of generative AI in their teaching. In addition to that, we also wanted to revisit our first-year legal research and writing curriculum. In the last couple of years, we’ve had to confront the fact that there are now generative AI tools out there that can cite--46--cases, sometimes hallucinate, but can often cite--46--cases very effectively, and can write with a lot of fluency.

So what we’ve moved to now over the last year, and then especially this coming academic year, is integrating the use of generative AI into our first-year legal research and writing curriculum. So we’ve shifted from a posture two years ago of saying, we don’t want you using AI at all as a first-year law student, to a posture of saying, as you begin the legal research and writing program, don’t use AI. Do everything by hand. We’re going to teach you how to use the tools that we’ve been teaching students to use for years now. But in the second half of that year, you are allowed to use generative AI. And we’re going to give you the training and the tools to make judgments about when you want to use generative AI in your research, and in your drafting, and in your editing.

Host:

In terms of the first half of the course, of the first-year legal reading and writing course, where you’re doing everything the old-fashioned way with the legal research tools that we had five years ago, how important is it to still develop those skills? Is that just sort of a relic that you’re keeping? Or do you think it has some real value to students that will be practicing in the next couple of years?

Guest:

Oh, absolutely real value. In fact, our view is that those skills are essential, and in some ways are going to be as important as they’ve ever been. And the reason is that I can imagine in a year, two years, three years, more and more legal practice settings moving to a model where even junior associates are in the position that, when I was in practice 20 years ago, the mid-level and senior associates were in.

What I mean by that is, you’re not the one doing the first draft. You’re now the one reviewing the first draft. And you’re making judgments about what’s good and what needs to be changed in that first draft.

Host:

Students from the University of Chicago Law School, they go to some of the best firms in the country. And I’m sure that the law school has good relationships with a lot of firms. What do you hear from firm leadership about what they want associates to know coming out of law school?

Guest:

Yeah, so as you know as well as I do, generative AI tools are something that law firms, especially the biggest law firms, have invested a huge amount of time and money into understanding and developing and implementing and rolling out for their attorneys.

In our conversations with folks at law firms of all sizes, especially the largest law firms, they are very interested in integrating these generative AI tools into their everyday practice. But what does that mean for law graduates coming out of UChicago Law? Most importantly, it actually doesn’t mean changing anything fundamentally.

There’s this mantra. It’s almost a cliche. We’re here to teach you how to think like a lawyer. And nothing about generative AI has changed that. The ability to argue, to make judgments, to separate the wheat from the chaff, those are all skills that are no less valuable in a world of generative AI. And you need those skills because you’re going to be reviewing AI output, as we’ve discussed.

What they do want to see is students having familiarity with tools. It’s not that we need to train them in any specific tool or have them understand any of the nuances of prompt engineering or something like that. The technology is moving too fast for us to rebrand ourselves as some sort of school that’s going to train you in how to use this product or that product.

Rather, the goal is for students to have enough experience with generative AI tools to understand how to review their output, when to recognize their limitations, how to recognize their limitations, and to understand what it means to use AI in an ethical and effective way.

There’s all sorts of traps one can fall into as a lawyer if they’re not careful about understanding the limitations of AI. The most obvious examples, of course, are lawyers who rely uncritically on the output of generative AI. But there’s other things to be concerned about as well. You don’t want AI tools retaining or training off of a work product that is potential attorney-client communications, sensitive information.

And so understanding the ways in which you, as a user of AI, need to ensure that you’re protecting yourself and your client when you’re using it, those are the kinds of skills that we want to emphasize in law school.

Host:

This generation of students, they’re obviously a lot more tech-fluent than you or I would have been going to law school. Do you find that they are more aware of the pitfalls of this kind of technology than maybe someone who’s not used to using this sort of thing and just starts using it in their practice? Or is there too much trust between young people and the technology? Do you notice a trend one way or the other?

Guest:

Yeah, that’s a really great question, because to tell you the truth, I’ve been surprised as I’ve learned more about generative AI and have interacted more with students about generative AI.

I initially thought that what the law school had to do was really emphasize the negative about generative AI, because I was initially very worried that students would be too quick to embrace it and make all sorts of mistakes. What I discovered was something quite a bit more nuanced than that.

In fact, I think we have some students who are very sophisticated users of generative AI and, in fact, are so sophisticated that they do understand a lot of its limitations and how to utilize it in ways that can avoid a lot of the mistakes that users might make. Then there are other students who might need a little bit of training on understanding the pitfalls, but they’re very comfortable using generative AI.

And then there’s a whole other group of students that are quite resistant to using generative AI. And in fact, one of the biggest shifts that I’ve had to make in my own approach to talking to students about generative AI is recognizing that for a lot of students, I actually have to encourage them to use generative AI, because they have heard, they’ve read all the stories, they’ve heard all the news about all of the ways in which lawyers have gotten into trouble using generative AI. And they say to themselves, look, I don’t want to get myself in trouble, and I’m here to learn anyway. I’m going to just focus on doing things the old-fashioned way.

And for those students, I actually have to tell them, look, I completely respect that approach, but AI is going to be a part of legal practice from now on. And you need to at least dip your toes in those waters so that you have some understanding of how AI can be used, because you might be expected to use it at your next job.

And so there’s really this full spectrum from people who are completely embracing it to people who are very, very skeptical and really need to be encouraged to give it a try.

Host:

During their time at Chicago Law School, is this something that they could avoid completely or dive into as a specialty? Like, what is the average student coming out going to have?

Guest:

Yeah, that’s something we’re actually thinking pretty hard about right now. We have a committee on AI, a faculty committee on AI. We’re working very closely with the librarians at the law school who have really been following developments in AI and legal research technology.

What we are exploring right now is exactly the question of how much of what we’re putting together is going to be required versus elective in the curriculum. And I think where we’re headed is a setup where you’re not going to be able to avoid AI entirely. It’s going to be part of your, it already is, frankly, it already is part of your first year legal research and writing curriculum. So you’re going to get AI in that context, no matter what.

And then what we’re going to see over time is, even in classes that don’t have the letters AI in the title, and as I mentioned, we have six or seven of those already, but even in classes that don’t have AI in the title, what you’re going to see is faculty integrating AI in different ways.

In the clinical setting, I think, is a particular setting where generative AI is really going to become part of every student’s experience as they go through the clinics over time. We already have some clinical faculty that are thinking very creatively about the ways that they can use existing AI tools and customize existing AI tools so that they can make their own clinical practice more efficient.

And I think this relates to something that a lot of the students at the University of Chicago Law School are really interested in, which is, what do you do as a lawyer when you don’t necessarily have the resources of the largest law firms who are able to create their own custom versions of generative AI tools?

Host:

It’s so interesting because when you’re talking about the clinical setting, a lot of what we talk about with lawyering is about the drafting and the research and making that faster, but you’re also talking about new ways to solve problems, which is sort of a new way of thinking like a lawyer. How can I use this tool to help more people?

Guest:

Yeah, I’m so glad that you mentioned that because there are two really different ways of thinking about generative AI than I’ve heard in most of the conversations about generative AI so far that I’ve been thinking about a lot recently and that I think are really important for the University of Chicago.

One of them is to say exactly what you said. Is there more to the use of generative AI than just using it to do more legal research or faster legal research like we’ve always done? Can we creatively design new generative AI tools to serve aspects of the legal community that haven’t been served as well by traditional tools?

And then the other thing that I’ll mention about this is it’s easy to think about the ways in which AI can help lawyers do things like legal research and legal writing better. But one of the things that I think we should focus on as a profession is how can AI actually make the law more human?

And I think one of the real possibilities for AI that has me feeling very optimistic about the legal profession is that to the extent that AI can help lawyers be more efficient at some of the tasks that take us away from our clients, things like going and doing legal research, going and writing memos, if AI helps us do those things faster, what that means for the profession, I think, is it actually frees up more of lawyers’ time and mental energy for the aspects of the profession that really are about human relationships and about serving the needs and building the relationships between lawyers and clients, lawyers and judges, lawyers and the broader community that we’re part of.

And I think that’s an aspect of what AI offers to the legal profession that I think hasn’t been emphasized, but I think is really important for the future.

Host:

That was William Hubbard, a deputy dean at the University of Chicago Law School. And that’ll do it for today’s episode of On the Merits. For more updates, visit our website at news.bloomberglaw.com. Once again, that’s news.bloomberglaw.com.

The podcast today was produced by myself, David Schultz, and Elleiana Green. Our editors were Chris Opfer and Alessandra Rafferty, and our executive producer is Josh Block. Thanks for listening. We’ll see you next time.