Appellate Advocacy Blog

Editor: Tessa L. Dysart
The University of Arizona
James E. Rogers College of Law

Sunday, March 8, 2020

Compose: The Robo-Lawyer We've Been Waiting For?

Image result for robot writing

"Select from a library of automations tailored to your motion, jurisdiction, and the side you represent. Each automation includes the arguments, legal standards, and supporting authority you need to craft a winning brief." - 

Folks have been sounding the alarm about robo-lawyers for decades. "Any day now," they say.

"We'll all be replaced."  

Mostly, that hasn't happened. Yes, lawyers struggle to realize the profits they once did. And yes, some of that is because of technology. But no, we don't have robots arguing motions or writing briefs yet.

Or do we? 

Casetext recently launched a product that brings us one step closer: Other than sporting the fancy .law domain, Casetext has done something exceptional here. They've created an automation tool that integrates legal research, writing, and high-level analysis all in one package. Other tools could replicate some of what Compose does. But no one has pulled off such a seamless product that does so much for so little.

If you're interested in looking under the hood or trying Compose for yourself, there's great press coverage here, and Casetext lets everyone try it for free at

I've played around with Compose for a while now, and I'm a big fan. Does it write your brief for you? Not really. It puts together a helpful argument outline, fills in some passable background discussion, and makes it a snap to incorporate research. In short: It looks to be an excellent tool that could offer real efficiency and value. I think you'd be crazy not to try it.  

A few folks have been more skeptical. They worry that a tool like this will short-circuit the all-important critical thinking and judgment that every lawyer must apply to every document they write. And that goes double for law students and young lawyers. 

But there are a few reasons this is probably a worry without a problem. And it mostly boils down to this: Compose is still just a tool. All the major analytical decisions are made by the user. And anything that Compose inserts is obvious. So the user still decides--sentence by sentence--how the final document should come together.

Here's more on why instead of scared, I'm excited. 

First, Compose's decisions are all still user-driven. The user, not Compose, decides which arguments to raise. The user, not Compose, decides which cases to cite. 

Second, the final document that Compose spits out is still quite limited in analysis and legal nuance. For any big-ticket motion or brief, Compose will save you tons of time on the basics (like putting together the background standards), but it isn't going to do all the nuanced analysis that lawyers do. It pre-heats the oven, but lawyers still do the cooking. 

Third, from what I've seen, many of Compose's suggestions are spot on. If a tool will help you produce something helpful to your client, how can that be a bad thing? Our job is to represent clients the best we can. It's not to win awards for our writing acumen. 

Fourth, good lawyers and firms have been trying to replicate this same stuff for ages. Compose is like a world-class knowledge management system integrated with legal research tools. In other words, it's a knowledge management system that actually works and people may actually use. 

Fifth, Compose produces first drafts, not final products. It frees up lawyers to spend more time debating our analytical and persuasive decisions, and less time on rote tasks like finding and describing a general background principle. 

Sixth, the idea that a lot of lawyerly judgment and thinking is needed (or used) on every document is just not true. Big caseloads and cost pressures already force lawyers to rinse-and-repeat prior motions they've written. In some practice areas, you'd be an outlier if you didn't copy and paste your litigation documents. Both as a clerk and in studying legal writing since, I've seen it more times than I can count (the giveaway is that one wrong name that the lawyer missed in their find-and-replace run). With a tool like Compose, these lawyers should put together a better document for their judges. 

Finally, the possibilities for the public excite me. The tool is a bit spendy for them (for lawyers, given the potential time savings, I think it's a steal). But I'd love to see a tool like this available to pro se folks. 

Instead of Compose making our skills atrophy, most lawyers and law students will learn a lot from the platform. I know I did, and I've only tinkered with it a few times. I can't wait until Casetext makes it available to law students and faculty, so I can start incorporating Compose in my classes. 

Joe Regalia is a law professor at the University of Nevada, Las Vegas, William S. Boyd School of Law and regularly leads workshops training legal writing and technology. The views he expresses here are solely his own and not intended to be legal advice. Check out his other articles and writing tips here

| Permalink


Post a comment