Law School Just Got Harder to Shortcut: UChicago's AI Policy and How to Study for It

For a couple of years, the quiet story in law school was that AI could carry some of the load. Feed it a fact pattern, get an outline back. Ask it to summarize the reading. Have it draft the memo and clean up the analysis afterward. Not every student worked that way, but enough did that professors noticed — and the good ones started asking an uncomfortable question. If a tool can produce the work, is the student still learning the thing the work was supposed to teach?

The University of Chicago Law School just answered out loud. On July 9, 2026, it released a strategy statement called Rethinking Legal Education in the AI Era, effective this fall. It is the first time a school of that stature has put a coherent, top-to-bottom AI policy on paper — and the through-line is blunt: the classroom is going to get harder to shortcut on purpose.

This post does two things. First, a straight read of what UChicago actually announced, because it is worth understanding on its own. Then the part that matters for you no matter where you go to school: what an AI-resilient classroom demands of the way you study.

What UChicago Law actually announced

The statement is organized around three themes, and they are the fastest way to understand the whole document:

Notice the shape of it. This is not a school banning AI, and it is not a school waving it through. It is a school drawing a line between the moments where a tool helps you learn and the moments where a tool does the learning for you — and then designing around that line. Here is how it shows up in practice.

Devices are coming out of the 1L classroom. UChicago is piloting a general prohibition on electronic devices in its core first-year classes, with limited exceptions. The reasoning is the one every cognitive-science paper on note-taking has been making for a decade: a room full of open laptops is a room where a lot of people are transcribing instead of thinking — and now, potentially, prompting instead of following.

Core exams go closed-book, no internet, no apps. First-year exams will be taken without internet or application access. Whatever you can do on the day is whatever you actually carry in your head and can reason with under pressure. There is no second screen to lean on.

Legal writing gets a deliberate order of operations. This is the most interesting piece. In Legal Research and Writing, students do the foundational writing without AI first — they learn to construct an argument themselves — and only then is AI-assisted editing layered on top. Write first. Get feedback second. The skill is built before the tool touches it.

Substantial research papers must be defended out loud. Every student has to discuss their major research paper orally, either in class or one-on-one with the professor. You cannot outsource a conversation. Either you can explain why your analysis holds together or you cannot, and the professor finds out in real time.

And — importantly — the clinics lean into AI. This is not technophobia. UChicago's clinics actively use AI tools in real client work: JusticeText in the Criminal and Juvenile Justice Clinic, VisaLaw.AI in the Immigrants' Rights Clinic. The upper-level curriculum adds courses like the AI Lab, Generative AI in Legal Practice, and AI in Appellate Practice and Judicial Decision-Making. The message is consistent with the three themes: learn the fundamentals under your own power, then learn to use the tools like a professional.

Dean Adam Chilton framed the whole thing as continuity rather than reaction: "We've always been willing to innovate with our curriculum," he said. "This moment is no different."

Why this matters even if you're not at UChicago

Schools watch each other, and they especially watch the ones that move first with a clear rationale. A device-free 1L classroom and a closed-book exam are not exotic ideas — they are easy to copy, and the moment one respected school makes the case in writing, the policy stops looking eccentric and starts looking like a standard. Expect versions of this at other schools over the next few cycles.

Which means the practical takeaway is not "here's a UChicago thing." It's this: the era where you could quietly let a tool do the first draft of your understanding is closing. When the exam is closed-book and the paper gets defended out loud, the only thing that helps you on the day is understanding you actually built. So the real question is how you build it — and how you use AI in a way that makes you sharper instead of dependent.

Here's the honest version of the answer, mapped to exactly what the new rules reward.

What an AI-resilient classroom demands of your study

1. When the laptop closes, your notes have to earn their keep

A device-free classroom changes what your notes are for. You're writing less, by hand, capturing the shape of the argument instead of a transcript. That's better for retention — but it puts the real work right after class, when you turn that rough capture into something you can actually study from.

This is where a deliberate system beats good intentions. After class, restructure the day's material into study-ready pieces while it's fresh: the rule, the elements, the case that anchors it, the exception that trips people up. In JD Simplified, that's what Notes and the flashcard system are built for — you turn a messy class capture into structured, retrievable cards. You can build your own decks for free; premium decks come with a plan when you'd rather start from ours. Either way, the point is the same: raw class notes are an input, not a study material. Something has to convert them. (If you've never nailed that conversion, our guide on the mistakes law students make briefing cases is a good place to fix the input side.)

2. Closed-book exams are a retrieval test — so practice retrieval

An exam with no internet and no apps measures one thing: what you can pull up and reason with under time pressure, unaided. Re-reading your outline until it feels familiar does not train that. Recall does. The research term is retrieval practice — repeatedly pulling information out of memory rather than pushing it back in — and it is the single most reliable way to make knowledge available on a closed-book day.

That's the whole idea behind spaced-repetition study mode: you're quizzed on the rule before you can look it up, at intervals timed to the edge of forgetting, so the act of remembering is the reps. Full study mode is free, so there's no reason to wait to start.

The companion problem is your outline. A closed-book exam punishes the topics you thought you covered but didn't. This is precisely the question Outline IQ is built to answer — is your outline actually complete, or does it just feel complete? — by checking your outline against a subject's full topic map and showing you what's missing before the exam does. (For the craft of building the outline in the first place, see the complete guide to mastering law school outlines.)

3. If you have to defend it out loud, practice saying it out loud

The oral-defense requirement is the sleeper. Most students have never once explained their analysis to another human before the moment it counts — they've only ever written it. But explaining a rule and its application in your own words, unaided, is a different skill from recognizing it on a page, and it exposes the difference between "I've seen this" and "I understand this" instantly.

You can rehearse that. Ask JDS — the AI chat built into JD Simplified — works well as a Socratic sparring partner: state your analysis, let it push back, and find out where your explanation falls apart. (Free accounts get 30 chat messages a month, which is enough to pressure-test the arguments that matter most.) Issue Spotter drills the upstream skill — reading a fact pattern and naming what's actually at stake — and Essay Coach makes you commit an analysis to the page and then tells you where it holds and where it doesn't. All three train the same underlying muscle: producing legal reasoning, not just recognizing it.

4. The strongest signal: write first, get AI feedback second

Look again at UChicago's writing sequence — foundational writing without AI, then AI-assisted editing layered on afterward. Read that carefully, because it is not a compromise. It is a precise statement about which kind of AI belongs in learning: the kind that reacts to work you did yourself, not the kind that does the work for you. Feedback AI, not ghostwriting AI.

That distinction is the exact design of Essay Coach. You write the essay. Then the AI grades it against a rubric and tells you what your analysis missed, where the structure wandered, which issues you underdeveloped. It never hands you the answer, because a graded read of your reasoning is the thing that actually makes you better — and a ghostwritten essay makes you worse while feeling like progress. When UChicago says "write without AI first, then use AI to edit," it is describing the shape of tool JD Simplified already built. The schools are now sanctioning a specific way to use AI in learning; it happens to be the way these tools work.

The new rule	What it rewards	How you train it
Devices out of 1L classes	Notes converted into real study material	Notes + your own flashcard decks (free)
Closed-book, no-internet exams	Unaided recall + a complete outline	Spaced-repetition study mode; Outline IQ
Oral defense of your paper	Explaining analysis out loud, unaided	Ask JDS (Socratic), Issue Spotter, Essay Coach
Write without AI, then edit with it	Reasoning you produced, then critiqued	Essay Coach — feedback, not ghostwriting

The direction of travel

None of this is unique to one school for long. The bar exam is moving the same way — the July 2026 NextGen UBE shifts the whole test toward applying law you're given rather than reciting law you memorized, which we broke down in why the bar exam was built for Mike Ross and isn't anymore. Legal education and legal licensing are converging on the same conclusion from both ends: the point was never how much you could carry in your head or how well a tool could fake it. The point is whether you can think.

UChicago is first. It will not be last. The good news is that studying for the AI-resilient classroom is not mysterious — it is the study advice that was always correct, finally made mandatory by the rules. Build the understanding yourself. Practice pulling it back out. Say it out loud until it holds. Use AI to critique your work, never to replace it. Do that, and the closed-book exam and the oral defense stop being threats and start being the format you've been rehearsing all along.

The study work you do this way carries forward, too — the understanding you build for a 1L closed-book exam is the same understanding the bar will test years later. You're not studying for one test. You're building the thing every test is trying to measure.