March 2026 / Episode 9 / Under 25 minutes
Welcome to Keeping Up to DataSM, a space in which we discuss, analyze, and contextualize trends and perspectives in the current law school admission cycle.
SUSAN KRINSKY: Welcome back to Keeping Up to Data. I’m Susan Krinsky, executive vice president at LSAC®, with an update on the 2026 application cycle and a discussion with Assistant Dean Sarah Zearfoss, of the University of Michigan Law School, to talk with us about the use in law schools, specifically law school admission, of that technology known as artificial intelligence.
As I record this podcast, we are just over a third of the way into the month of March and slightly more than 80% of the way through the cycle, based on volumes. By this time last year, we had seen 81% of the final applicant count and 85% of the applications that were eventually submitted. As of today, the number of individuals applying to law school — that is, applicants — is 12% higher than last year at this time and almost 35% higher than two years ago. As of today, 69,555 individuals have applied to at least one law school, and when we look at the average number of applicants over the past four years, that’s a 30.2% increase. As for applications submitted, there are almost 511,000 applications in our system — 14.6% more applications than last year at this time, and 41.7% more than two years ago. Compared to the prior four-year average, applications are up almost 34%.
Both applicants and applications have slowed down pretty dramatically over the past month. As schools’ application deadlines approach, that’s probably not surprising. We and others have been encouraging individuals interested in applying to law school to apply earlier in the cycle, and that does seem to be happening. In addition, it’s pretty well known that this is a particularly competitive cycle, as was last year’s, so people who thought about applying late in the cycle last year may have moved to early this year, and people who thought they’d apply late in the current cycle may be rethinking that strategy and moving to next year. And there are also the changes in federal financial aid availability, which may be affecting potential applicants’ decisions about when to apply. Our research and data teams are taking a close look to see if there’s anything else going on, and if there is, we’ll report on it.
As of this week, 48.7% of individuals who have submitted at least one application for 2026 admission identify as persons of color; last year at this time, that percentage was 48.2%. There are 13% more applicants who identify as persons of color this year as compared to last year. This year, there are 16.4% more individuals identifying as Black or African-American than last year. This compares to the overall increase in the applicant pool, as noted earlier, of 12%. 57.6% of this year’s applicants identify as female, 39.6% as male, 1.9% did not indicate a gender, and just under 1% identify as gender diverse — no change from last month.
Applicants who describe themselves as being first-generation college attendees or graduates represent 26.6% of the almost 70,000 applicants. This is up slightly from a month ago and also up from last year at this time. As was the case last month, applicants from all regions of the U.S. are up: between 10.5%, from the New England region, and 15%, from the Great Lakes region. And applications to schools in all regions of the U.S. are up, too: between 9.3%, at schools in the Northwest, and 18.3%, to schools in New England. An interesting comparison to the applicant increase: Canadian applicants are up 13.4% over last year, and applications to Canadian law schools are up 15.3% over last year. Both of those percentage increases are very close, indeed, slightly higher, than they were a month ago
There was no LSAT® administration in March, so no change in the 16% increase in test takers so far. The next LSAT administration begins on April 9, and registration has, of course, closed for that test. The number of registrants for the April test is virtually identical to the number of registrants for the April test a year ago and about 6% higher than the April 2024 test. Registration for the June LSAT administration — the last one for this testing year, and the last one before we move to predominantly in-center testing — is within 1% of what it was for the June 2025 test. This year, registration for the June test closes on April 21, so we’ll know more next month. Registration for the 2026-27 tests will open in the middle of May.
That’s our report on the March data. Remember that you can stay on top of our volume data by going to our website, LSAC.org, where you’ll find a link on the homepage to the latest volume data, which is updated every night.
Now, we are going to switch gears onto a topic that seems to be top of mind for many and that raises many questions, and that topic is artificial intelligence. AI suddenly, or maybe not so suddenly, seems to be everywhere, and almost daily, we learn about its effect on virtually every industry, including the legal industry. As AI tools become ever more sophisticated and widely available and more widely used, they are beginning to transform not only how legal work is performed, but also, probably necessarily, how the next generation of lawyers is trained. Across law schools, educators and students are grappling with a fundamental question: not just how should legal education adapt, but how must it change in a world where AI can assist with research, analyze vast bodies of case law, draft legal language, and summarize complex information in seconds. These capabilities present remarkable opportunities for improving efficiency and expanding access to information, but they also raise important questions about academic integrity, professional ethics, and the preservation of critical legal and thinking skills.
Law schools are increasingly incorporating AI into their curricula, including developing entire courses on the subject. Schools are wisely, I think, examining both the technology itself and its broader implications for legal practice. Courses and workshops explore topics such as algorithmic bias, the ethical responsibilities of lawyers using AI tools, data privacy, and the regulatory challenges that accompany emerging technologies. Legal educators are considering how to teach students to use these tools responsibly while still developing the analytical reasoning, judgment, and advocacy skills that define the legal profession. When it comes to law school admission, law schools are considering how AI can be a positive part of the process.
Joining me today is Sarah Zearfoss, of the University of Michigan Law School and a member of LSAC’s Board of Trustees — and also the host of the podcast Admissions A2Z with Dean Z. Dean Z has agreed to talk with me about how AI is reshaping legal education, what it means for students preparing to enter the profession, and how institutions are balancing innovation with the enduring principles of legal training.
Sarah, welcome to the Keeping Up to Data podcast. And let me just say that I am very honored to have a podcast superstar like you with me today.
SARAH ZEARFOSS: Yeah. Reports of my fame may have been exaggerated, but I am very thrilled to be here. It’s a very interesting topic, and I’m looking forward to our conversation.
SUSAN: Well, AI is a topic of conversation these days in just about every industry and field. What is your take on how things are changing, not just in the legal system, but on how we should envision the future of legal education moving forward?
SARAH: With specific regard to artificial intelligence, you mean? I mean, it’s certainly cropping up everywhere. There’s this huge amount of pressure within law firms to start using it. Law schools are responding in kind, as you mentioned, offering coursework both about how to use AI in your legal studies, but also how to use it responsibly, how to think about the, as you mentioned at the outset, the ethics of it. And we’ve come a long way in a very short amount of time. As you know, the legal industry, and legal education in particular, tends to be slow-moving and "small-C conservative" about change, and I’ve seen more evolution in this space than I can think of in any other time in my many years working in law schools.
SUSAN: Am I remembering correctly that a few years ago, the University of Michigan Law School required applicants to certify that they had not used AI in their applications?
SARAH: That’s correct. As soon as ChatGPT became a thing, it was a matter of concern to me. Obviously, this was going to be a tool that people could use to draft essays. So I immediately thought, as I was revising my application, as I do every year, I thought, “Better address this in our certification.” It seemed to me like an obvious thing to do. We had to make [it] very clear. We already had said, “You’re not supposed to be using outside assistance, apart from basic feedback and that sort of thing.” And I thought, “Well, I should make it clear and include generative AI in that.” I was very surprised at the amount of ... Not pushback. I guess I was surprised at the amount of surprise that that move generated. So, a local TV station wanted to interview me about it. I and the dean got calls from places wanting to talk about it. People were very surprised at that step, where to me, it seemed, at that time, obvious and necessary.
SUSAN: Well, speaking of changes, last fall, your school rolled out an optional prompt for applicants, which requires the use of AI to answer the question, “How much do you use generative AI tools, such as ChatGPT, right now?” How did that decision come about, and why that particular prompt?
SARAH: Again, we revise our applications every summer, and I was at the LSAC annual conference, and there were a couple of sessions on AI, and it started me thinking about, “Well, how could I use it?” It’s definitely going to be a tool that is very important in some legal settings. A lot of what we’re doing in our application is trying to see: Does the person have the tools to become not just a law student, but a lawyer? So I wanted to see, how can I examine this?
And I went to one of our professors who was teaching a course on AI and writing, and we went back and forth in different topics we could ask, and this was actually his idea. His name’s Patrick Barry. He’s really an amazing writing professor, but also very smart about AI. And he thought this would be an interesting topic. So, it’s: How much do you use it now, how much do you think you’ll use it at the end of law school, and why? So, it wasn’t really my idea. I know a good idea when I hear one, and that’s almost as good as having them yourself.
SUSAN: You can tell me if this is wrong, but according to a news report, only about 5% of Michigan applicants opted to use that AI-generated essay or to respond to the prompt, while at the University of Miami, we understand that about 45% did so, albeit to a different prompt.
SARAH: Yeah.
SUSAN: Any thoughts on why that difference? And do you think that’s going to change your prompt next year?
SARAH: First of all, I’ll say that that story came from September, and I was guessing based on about one month worth of data. But to my surprise, we just finished our application season, and it looks like it held steady — it’s about 5%, maybe 6%, but not much more than that. So, the 5% was a good guess, even after only one month of data. Now, why we would have a smaller percentage answering than Miami is hard to say. They’re very different schools; they’re very different pool sizes, I think; and we offer 10 different topics to write optional essays about. And so, I assume that’s part of why that’s going to obviously depress the number that choose any one topic. So, I assume that’s why.
But I will also say I was very surprised, as I was reading throughout the year, I would read applications from people who had sophisticated backgrounds in AI or in computers, but specifically in AI, and they didn’t choose to answer that prompt. I’m not sure why. If they enroll, if they’re people I admitted and they enroll, I do intend to ask them about that. And I was also surprised by the number of people who did choose to answer and would say in part of their answer, “I don’t use it at all” or “This is the very first time I have used it.” And that’s a surprising choice, I think, to choose to experiment with it for the first time in your law school essay. I kind of admired their boldness and sense of adventure, though. But a lot of the people who said they don’t use it cited environmental reasons, and their concern about the effect on the environment, for choosing not to use it.
SUSAN: Do you think you’ll offer a different prompt next year, or still thinking about that?
SARAH: I’m still thinking about that. So, I’m in the midst of collecting, trying to print out all those particular essays and collect them in one place. And then, I and a couple of other people at the law school are going to review them and think about: Do we want to keep having it, do we want to change it, do we want to stick with it? I’m inclined to stick with it, because I do think I’m more interested in the quality of the answers than I am in the percentage of people who chose to answer it, and I do feel like I got useful information about people’s AI skill set. There was quite a range.
SUSAN: You, like I, have been in and around legal education for some time — we won’t say for how much time — and you and I have seen trends come and go. Do you think AI is a trend, or is it here to stay?
SARAH: “Some time,” meaning a lot of time, yes. I think it’s here to stay. This question reminds me of CDs. When CDs first came on the scene, I was like, “I am not buying these. I don’t think that’s real.” And of course, they now have come and gone again. So, I don’t know what comes after AI, but I think AI will be with us for quite a while. Even if it never got any better than it is right now, at least in the legal profession, it’s still able to do things that are of great utility and efficiency for law firms. And I say law firms, in particular, as opposed to legal employers, because I don’t think all legal employers will be using AI to the same degree.
SUSAN: Do you think AI will have an increasing role in law school admission, and if so, how?
SARAH: Well, let me ask you this. Do you mean an increasing role in terms of how admissions offices evaluate applications, or do you mean as I, with my prompt, was interested in seeing how good people are at using AI?
SUSAN: I probably meant the former, although I will happily hear about both.
SARAH: Yes. I do think that it will be important for schools to try to start assessing people’s ability to use AI. That doesn’t mean everybody has to be able to use AI to get into law school. But you reviewed applications for many years before your role at LSAC, so you know that when someone says, “I’m interested in being an international lawyer,” and they have done nothing international in their lives, you question what they’re thinking when they say that and how sophisticated their understanding of what international law actually is, and it has an effect on your evaluation of their application materials. So, by the same token, if someone says they want to go into a big law practice of M&A or deal making, and they say they don’t use AI at all and they don’t want to use it, I would say that’s kind of a mismatch. Your goals and your skills are not matching up here. And I think all law schools are going to have to think about that.
In terms of what I think the role of AI will be in actually conducting admissions, I do think there will be some use for it in terms of perhaps using it to test your consistency of decision-making, as an anti-bias move, or perhaps to evaluate parts of the application. One part of the application that I think is quite useful, that I think we underutilize, is the LSAC Argumentative Writing™ sample. We look at it; it’s more of a very quick evaluation of your writing skills than anything else. So, I can imagine using AI just to assess that one little piece of the application, for example.
I think that the day when AI is used to fully evaluate an application and render a decision is very far in the future, if ever. I think humans are very uncomfortable with that idea. I am very uncomfortable with that idea, and I don’t think it is just my age that makes me uncomfortable with that change. I think that young people are also very uncomfortable with that idea, as I think they should be. Not to wax too philosophical here, but there’s been a lot of discussion in the news about Anthropic and the use of its tools with the Department of War, and whether they can set any limitations on the way it’s used, specifically for surveilling American citizens. And I feel like there’s this natural reaction to that of, like, “I want a human surveilling me; I don’t want a machine.” I feel like we should be limited in the state surveillance of people by what humans can actually do, because there’s just a terrifying amount of power in these tools, and I think we should use them, but keep them in check.
So, that’s a long answer of saying I just don’t see admissions officers being fully replaced by AI anytime soon.
SUSAN: Should law schools be thinking about ways to use AI productively in other areas?
SARAH: One of the worst jobs, I think, for law school employees is putting together the class schedule. You have 80 faculty members who have differing preferences for time of teaching, day of teaching, room for teaching — as in the particular room that they are in — courses that they want to teach. And so, there’s tons of preferences by faculty, who are very valued employees. And then, there are, of course, institutional needs — like, if nobody wants to teach comm law, well, we still need to have it taught, right? And this is an incredible logic games puzzle of putting all these pieces together. I think AI could really help with that, and that isn’t going to put anybody out of a job. That’s going to keep our beloved registrar from pulling all her hair out. She has lots of other things that she should be doing, too.
You could also use AI to, again, maybe check for consistency in grading. I could imagine that. I don’t know how excited faculty members would be for that, but perhaps they could make that choice, whether they wanted to use the tool. I think there’s myriad uses for it just in running a law school, let alone in the classroom.
SUSAN: Thank you so much for taking time out of your busy schedule to share your thoughts. I am sure there will be a lot more to talk about on this topic in the coming months. Again, thank you.
SARAH: Absolutely my pleasure, Susan.
SUSAN: Thank you for joining us at Keeping Up to Data. We look forward to your joining our next episode. Until next time, stay well.
Keeping Up to DataSM is a production of LSAC. If you want to learn more about the current law school admission cycle and the latest trends and news, visit us at LSAC.org.