The real AI risk isn't hallucinations—it's not using AI at all. That's the counterintuitive argument from Ross Guberman, founder and CEO of BriefCatch, who tells hosts Todd Smith and Jody Sanders that lawyers ignoring AI may be approaching malpractice territory. BriefCatch now contains roughly 16,000 editing rules, up from about 10,000 in 2021. Ross introduces two new BriefCatch Next products: RealityCheck, an anti-hallucination tool that uses a red-yellow-green system to verify case citations and quotations against actual opinions; and CiteCheck, which identifies and corrects Bluebook citation errors. He also covers BriefCatch's early 2026 acquisition of WordRake and reveals that courts—despite their sanctions orders against lawyers—are quietly and actively deploying AI themselves.

Connect and Learn More

☑️ Ross Guberman | LinkedIn | X

☑️ BriefCatch | LinkedIn

☑️ Todd Smith | LinkedIn | X

☑️ Jody Sanders | LinkedIn | X

☑️ Texas Appellate Law Podcast on LinkedIn | X | Instagram

☑️ Texas Appellate Counsel PLLC

☑️ Kelly Hart & Hallman, LLP | LinkedIn

☑️ Subscribe Apple Podcasts | Spotify | Amazon Music | YouTube

Produced and Powered by LawPods

Sponsored by Court Surety Bond Agency and Proceed.

Welcome to the Texas

Appellate Law Podcast,

the show that takes you inside the

Texas and federal appellate systems.

Through conversations with judges, court

staff, top trial and appellate lawyers,

academics, and innovators,

we provide practical insights to help

you become a more effective advocate.

Whether you're handling

appeals or preparing for trial,

you'll discover strategies to sharpen

your arguments, innovate your practice,

and stay ahead of the latest

developments. And now here are your hosts,

Todd Smith and Jody Sanders.

Produced and powered by LawPods.

Welcome back to the Texas Appellate

Law Podcast. I'm Todd Smith.

And I'm Jody Sanders.

Our guest today is Ross Guberman.

Ross is the founder and CEO of BriefCatch

and he has been a previous guest with

us before. So welcome back, Ross.

Good to see you guys. I think

it's been about five years.

Yeah, which is wild.

I was checking on that.

It has been since July of 2021 and we

were chatting before we hit record here

about how the world has changed in the

last five years. When you were on before,

you spoke about leveraging

technology to improve legal writing.

You're doing BriefCatch

at the time, I think,

but Legal Writing Pro was still

a part of what you were doing.

Because we spent so much time getting

to know you back five years ago,

we kind of like to focus on

really what's happened since then.

So we will kind of abbreviate

our introduction of you.

A lot of Texas lawyers,

especially appellate lawyers,

will know who you are and what BriefCatch

is because it's a product that both

Jody and I are fans of and we use.

And I think a lot of our fellow lawyers

in our area are the same. So let's just

start where BriefCatch is right now

Yeah. I mean, it's hard to believe it's

been five years. You guys look younger,

by the way.

You're very kind.

Yeah.

So yeah, I mean,

one thing that's changed is I've really

gone kind of full 100% tech founder

running much more of what

feels like a real tech company,

whether I'm well suited for it or

So I'm definitely in the throes of AI

and of all the debates in our profession

adjacent to our profession about

tech, the future of tech with lawyers,

with the courts, with the public, with

pro se litigants, the whole works.

And a lot of what we've done at BriefCatch

has sort of followed all those trends

that you were just referring to.

So we have really moved from what was

like for me in a way more manageable,

fully coded, deterministic

product. That's what they call it,

deterministic of editing rules,

which that was easier for me to understand

as a non-technical person when I

started the company. I mean, now we

have much bigger, broader ambitions.

We're trying to satisfy

all sorts of requests from

different types of clients we

have and we're adopting AI. Absolutely.

We're doing all sorts of

exciting things with it,

but we're still keeping

our core product the same.

We still have our universe of editing

rules and I still have things that I

recognize from the product that

So a lot of what we're doing here is kind

of just a microcosm of what's going on

in the profession at large

and even with tech in general.

So I did go back and check

out our previous episode.

And what I remember from that,

just to kind of recap a few things is

you had a pretty large list of rules that

were encapsulated within

Briefcatch even then.

I think it was about 10,000 or so.

My understanding is that number's

expanded quite significantly.

Yeah. I think we're up to about 16,000.

So that's a lot of things to

change in lawyers writing.

And you were also at the time,

I think maybe you were already

beginning to shift somewhat on this,

but you used to go around giving a

whole lot of workshops as I remember.

And I don't think that's really

part of what you're doing anymore.

Yeah, I've mixed feelings about that.

I actually, around the time of COVID,

so around 2021, I did a count.

I was bored one day as many of us

were and I realized I had done 2,850

workshops around the world. So

it's not like I didn't pay my dues,

but just because of reality, I for

the most part stopped doing that,

although I still speak to

all the new federal judges.

I just spoke to a bunch of female

judges at the National Women's Judges

Conference in Minneapolis. So I

do a litle bit of speaking still.

I do go to some of our bigger briefcatch

clients occasionally at least by Zoom

for a workshop. But yeah,

I'm pretty much out of what was my

livelihood for solid 17 years. And again,

all in tech, sometimes 18, 20,

even 22 hours a day, to be honest.

I won't lie.

I miss it a little bit. I mean, I miss.

I'm very extroverted for an attorney.

I loved crowds. I loved traveling.

I loved meeting lawyers and

judges all over the planet.

So I do have my nostalgic days.

So I think it'd probably be a good idea,

even though we've referred folks

back to that previous episode.

Just give us a thumbnail sketch of what

Briefcatch is and what it's doing right

now.

Yeah.

So the original idea for Briefcatch was

to take a lot of the different lessons

in my books,

also the ones I tried to impart in

my workshops and make them much more

manageable and actionable

twenty four seven.

And that's the problem with a lot of

writing instruction is you can embrace it,

you can internalize it at the time,

like when you're reading

the book or at the workshop.

But as I think you'll agree, it's very,

very hard when you're actual practicing

lawyer and have a thousand different

types of pressures.

It's hard to remember a lot of the tips

and tricks and tools, not the tricks.

Those are easy to remember.

So the original idea was a little bit

modest compared to what it is now.

I just wanted to automate a

lot of the tips I would teach.

So looking for wording patterns and then

proposing alternatives. And that's how

it started. And then as you can imagine,

you mentioned the number of rules a

little a minute ago. And as time went on,

I thought of more rules and more edits

and then more exceptions to the edits.

But that was the basic idea is to help

people edit their work when they're under

the gun by automating a lot of the

things that I would try to teach, again,

either through my books or my workshops.

And the way that I receive it,

I think it's the same for everyone

is essentially as a word add in.

So Microsoft Word environment either,

I'm glad to say it is available for Mac,

which wasn't always the case with

every good piece of software out there,

but it has been available

in Mac for quite some time.

I know it obviously was available to

PCs and just the way that it works,

the way that I use it,

and I think this is generally how it's

delivered is just it's an option that

pops up in the toolbar that brings up

a sidebar and you can actually work in

your own document and Word and get the

editorial suggestions and fixes from the

software directly there in Word,

which is one of the things I

think makes it so powerful.

Yeah, one good thing is I did practice

law, although not for very long.

I know lawyers, I know judges,

I know how people like to work.

So we tried to keep it, as you

said, very straightforward,

right inside your document, but

then we did add some other layers.

So for example, sure they like edits,

but they like an explanation like,

why is this a good edit?

That's also, by the way,

a good way to learn better writing

on the job. You get little messages.

And then one thing I added later, I think

one thing I added that's become

incredibly popular were examples.

So would show like, "Hey,

this is Scalia doing what I said or

this is Kagan doing what I said or Paul

Clement." And those also

give people reassurance,

but they can be in their own way a little

bit of quiet tacit writing instruction

as well because you're getting exposed

to some really short bits of really fine

writing.

So that's another thing we tried to add.

Sure.

I want to mention before we go too far

down the road that we recently had Chris

Chandeville on the show and you see Chris

a lot on LinkedIn and your name came

up-.

He's great.

... in that conversation.

I like the way that Chris put it when

we were talking briefly about you and

Briefcatch and he said, "It's like

having Ross as your editor living in.

Word." I hope that was a compliment. Yeah,

for some people that might be a nightmare.

I think it was a compliment.

I think so too.

I think from him, probably was.

I did take him out to lunch one,

so maybe that's why.

Ah, there you go.

He's being kind. He's quite

fantastic, by the way.

I can see why you would have him on

your show. And he's also, I think,

done a lot of consolidating with this,

I think he calls it brief

writing ninja techniques.

He's also taken what can be

very overwhelming for many

people and turned it into

checklists and other things that

I think are really valuable.

Hats off to both of you.

I thought you would appreciate

that reference though.

So then walk us through, I mean,

my recollection is basically the

beginning stages of Briefcatch or someone,

not you because you're not a tech guy,

was hard coding in these edits

or these strategies into the

software. And you mentioned now that

the company's relying more on AI,

but how do you make the shift

from someone who's going around,

gives more than 2000 workshops and

is not tech oriented to becoming

essentially the CEO of a tech company

and delivering a product that is

unquestionably a tech and

AI functioning product?

No, it's a great question.

Most people in my shoes in the legal

tech world actually had a co-founder,

like a technical co-founder with the

kind of background you would expect,

an engineering degree from college, a few

years at Google or Apple and so forth.

So because I didn't know that I was

going to be a tech founder and I just

thought I was creating software,

I'd never thought of such a thing.

So I had to wing it.

One good thing though is even my legal

writing career is not my first career.

Being a lawyer wasn't my first career.

I am kind of lifelong learner type and

I actually have picked up quite a bit

about coding and tech through

osmosis and I enjoy that.

It's very stimulating and

it's fascinating. And the

truth is most lawyers,

if they wanted to learn how to code,

would be good at it and probably

would enjoy it because it is not that

different, right?

There's a lot of logic involved,

intention to language and the

like. So long story short,

I did have to learn a language for

the rules, but it's not really code.

It's sort of like a pre-coded language,

natural language processing

language based on,

I won't bore you with all the details,

part of speech tags and the like.

So I did have to learn something akin

to a code to write the editing rules,

but then as you suggested,

those were converted into XML and other

coding languages that luckily I did not

have to do on my own. So that was part

of it. For a non-technical founder,

I have become quite technical of had to.

But then the other thing is

just as in all businesses,

including running a law firm, you

have to learn how to hire, right?

You have to learn how to hire well and

you have to learn how to hire for things

that you're not good at. So I'm not

good at coding to put it modely,

or at least I certainly wasn't years ago.

So I just had to rely on hunches or got

instinct or try to get other people to

help me hire. But it's just, again, it's

a little bit like any other business.

You have to wear a lot of hats and you

have to sort of know what you're good at

and then you learn very quickly what

things you're not good at and need to

delegate.

I can't imagine what an arc the company

has taken over the last five years.

We've talked a lot on this show about

how AI has kind of taken over and we talk

a lot about AI generally on this show,

even though it's certainly not a

tech focused show specifically,

but it's been interesting to watch.

I mean, any legal conference, yeah,

look at any legal conference right now

on any topic and you look at the actual

panels, everything is AI. It's

all about AI everywhere you go.

Well,

where people seem to get into trouble

is not understanding what it is and what

it does.

And that's so important to at least

educate people even if they don't really

adopt it as part of their legal practice.

Knowing what it is and what

it's not is so critical.

Yeah. There are kind of a lot of myths

out there and a lot of misinformation.

Well.

It's unfortunate that I think the CLE

world has kind of reached the saturation

point on AI. I mean,

I've had people who are

putting on programs who want

to stay away from AI topics

because it's like all you

hear about all the time is AI.

But as we all know, we

have a duty of competence,

technological competence as lawyers.

And so you can make a pretty good case

now that someone who's just completely

oblivious to AI is exposing

themselves to malpractice.

Right. I mean, it's interesting because

if you just look at the legal headlines,

you would think the number one issue

is avoiding hallucinations and avoiding

sanctions over hallucinations.

And I have a lot to say about that

knee deep in it if you're interested.

But I think what you're talking about is

the stealth issue that's actually much

more important and it's

kind of the opposite, right?

Not can you get in trouble using AI,

but can you get in trouble not using AI?

And I think we're probably

at an inflection point now

where you're talking about

it's going to become a much bigger story

than the story about the pitfalls of

using AI.

Well,

you gave us the perfect opportunity so

I'm going to lob that softball about your

thoughts on hallucination.

Go for it.

Yeah. Well, I mean, for a

long time, this has been,

I think even just empirically,

the number one topic in the legal press

when it comes to tech in general and

GenAI in particular is this

whole hallucination problem.

So I've spent about the last six

months of my life and I mean my life,

not my work life,

but like 24 hours a day it seems

sometimes dealing with confronting the

challenge of hallucinations in our

profession. So one thing I'll say,

we're talking about kind of myths. So

obviously when I say hallucinations,

most people think correctly of the ones

you hear about the actual hallucinated

case filled briefs that make the news.

And it's the same story over and over

and over again. Just the names change,

the details change, the level of

angry rhetoric from the court changes.

But as we all know, if you use one of

the standard LLMs, Claude, ChatGPT,

Grock, Gemini, what have you,

there's a chance that it will,

if you ask it to generate

a legal document,

that you will get very serious somber

looking citations and information or

quotations from the cases the citations

refer to and then all that's basically

BS, it doesn't exist,

but it has the veneer of formality and

people fall for it and keep it in their

briefs.

And let's be honest,

some judges have gotten caught too

recently keep it in their opinions.

So that's the sort of like knee-jerk

hallucination problem people are very

familiar with, but actually there

are other more interesting layers.

So very often nowadays, by the way,

there's a much lower chance right now of

getting a purely hallucinated case than

there was a year ago because again,

anthropic OpenAI are onto the problem

and they've fixed it to some extent.

So what's more common now is a correct

case that does exist with the right

reporter number and the like,

but then there's either something like

a quotation from the case that's not in

the case at all or there's a statement

of what the case means or what the

holding was and those are inaccurate.

So that's something else we've tackled.

And then the third thing I like

to point out is that this is,

you said I've been doing this for a

while and that's true. The problem of

inaccurate discussion of

case laws as old as time,

it was a huge problem long before

we ever heard the term GenAI. Yes,

it's true that not that many lawyers

literally cited completely made up cases

before GenAI,

but plenty of lawyers and frankly judges

said all sorts of things about cases

intentionally or not that just weren't

true and that has nothing to do with

GenAI and it has nothing to do with

sloppiness. I mean, I guess it's sloppy,

but it's human. These are human

errors. So although Reality Check,

which is our anti-hallucination product,

it was sort of conjured up

to fight hallucinations.

GenAI produced hallucinations. Luckily,

I think for us and for

hopefully the profession,

it also catches all sorts of other

errors people make in discussing case law

that have nothing to do with GenAI.

That gives us an opening then to move

straight into reality check because that

is a new announcement,

a product announcement that you

guys have made pretty recently.

I know you've been speaking on it and

educating lawyers about it and you kind of

covered the basics of what it does,

which I think is super helpful to reach

that next level of hallucination besides

just making up cases. What's

the status of the rollout?

Is that widely available

now or where does it stand?

Your timing is great for this episode

because we launched it and we had more

interest than expected and we were not

able to scale it at the level we would've

had to. So we're

relaunching it so to speak.

We have a shocking number of courts

clamoring for it, courts at all levels.

So they're kind of next

probably next week.

And we are also starting to make

it available to our customers,

but we are getting feedback from the

courts first because can't predict the

future,

but it may be the case one day that every

lawyer who files a brief in America,

at least in the federal courts,

the first thing that's going to happen

is the product's going to be run through

reality check, the court level.

So we want to make sure that is

seamless so that lawyers, of course,

ideally and the perfect lawyers run it

through it themselves and the courts have

nothing to detect.

So that's where we are.

That's another reason I haven't

been sleeping very much lately.

We are very excited. Again, demand

is outpaced our expectations,

which is a good problem to have. And

we are rolling it out as we speak,

literally right now, probably

someone's getting it.

I saw an update in Briefcatch

on my computer and I was like,

"Maybe this is the day.".

Did you? I'm glad to hear that.

You're teaching me things.

Yeah. There was an update a few days ago,

but I'm not sure that I actually

have the full capability yet.

Not yet. Just a teaser.

Yeah. And also there for

a while, I was unsure.

I've moved back into solo practice and

I wasn't real sure if it was going to be

part of my level of

service or not. At first,

I think the rumor was it was

going to be enterprise only,

but now I've since been disabused of that.

I think maybe I DM'd you about it or

messaged you on LinkedIn about it at one

point. My understanding is

it's going to be available.

You weren't the only one, by the way.

I'm sure. So it's going to be available

to any level of practice because I mean,

solos do great work.

Don't get me wrong, Ross,

but we need these tools as

much as the big guys do.

100% Or more so, right?

Because it's a huge burden when you

don't have that infrastructure. Yeah.

I actually learned something.

There are 410,000 solo

practitioners in our country.

Wow.

That number actually shocked me.

That kind of shocks me

too. That's amazing.

Yeah.

But you're going to give it to the courts

first so they're going to be able to

check our papers before we can using your

tools. Lesson to be learned out there.

Maybe I'm good at marketing. I don't.

Know. Sidecheck your cases. The courts

may find them even if you don't.

I may not be that bad at business for a

lawyer might know what I'm doing here.

I mean,

you've been public about really getting

involved in making the product available

to courts all over the country.

And I think it speaks well of the company

and the product that the courts are

not only expressing interest,

but really want to get it or giving you

a level of demand from that source you

didn't necessarily expect to get.

Yeah. It makes me nervous too.

I want everything to be perfect now

that we're talking about the courts.

But in all seriousness, I am very,

very blessed to have the

trust of the nation's courts.

It's very hard to earn that and I don't

want to squander it. So yes, I mean,

it did help all those years.

Good lesson for people with an

entrepreneurial bent out there.

It takes a long time to build

up that kind of credibility,

not as much time to lose it, I'm

told. So yes, I do have a very,

very good relationship

with the nation's courts.

Without getting too technical,

can you kind of walk us through how

RealityCheck corrects the two different

types of hallucinations

that you brought up?

Oh, absolutely. So I'll just try

to explain it in normal terms.

So imagine you have a

brief, whether filed or not.

The first thing that happens is we are

extracting every single case that you

cite and then we're also linking to

that what you say about the case.

So that could be a quotation,

could be a parenthetical,

obviously your own words,

you're paraphrasing the holding,

talking about what facts the court

relied on, any combination thereof.

So that's the first thing that happens.

Now we have our own proprietary system

for tagging all that and for classifying

it and what type of legal

reasoning it is and the like,

but that's the first thing that happens.

So the next thing that happens is we

take every single case you cited and we

actually find it. We're not using

AI. We're not asking ChatGPT like,

what does the case say? We're finding

the actual case. We use midpage.

For the most part, we

use midpage right now.

Midpage for various reasons

is the best database for us.

And then once we pull the case, we

are doing the same kind of thing.

So if you quotations, we're

pulling the quotation.

If you say the Fifth Circuit held X, Y,

Z, we are often either, if we're lucky,

we're copying the holding,

often you have to infer it,

especially from trial courts.

We are then in any event,

we're doing the analog of

whatever you did in the brief.

We're finding that in the opinions and

then we're matching them up and then

we're verifying or not. So if

quotation's a little bit off,

if what you say the holding is exaggerated

or just way off, we note all that.

And then what happens is it's

all very straightforward.

I'd like to think at least then you

see it's just like a red, yellow,

green system. You want lots of greens.

Green means whatever you did is great.

Case exists. The quotations are accurate.

What you said about the

case is good enough.

Yellow means exactly what you'd

expect. You better double check this.

It's like a little bit of a stretch.

The holding might not have gone quite

that far or the quotation's a little bit

off. It's a real quotation,

but it seems like you left out a word or

you're changing words without following

the blue book rules for brackets.

And then red would be exactly again

what you would think. Bad news,

this case doesn't actually exist or what

you're saying is very, very far afield.

So you can see given that why it

helps you beyond GenAI produce legal

writing, right?

Because all those things that GenAI

does also happen in normal writing just

because we're human. Now,

sometimes it's because people are trying

to play fast and loose with the case

law. And one of the challenges to be

frank in creating this product is where

should we draw the line? You

have to draw a line somewhere.

Is this zealous advocacy or are

you actually lying about the case?

And not everyone's going to agree

with where I'm drawing the line in the

product,

but at the very least we're making your

job much easier on whether you're a

judge or a lawyer because we're verifying

a lot of things so you can exhale and

then we're limiting the number of things

that you really want to spend time

double checking or triple checking.

I've heard of mid-page and I've

heard good things about it.

I don't use it personally.

I'm curious though,

will we get down into say secondary

sources or is it limited to case law or

statutes?

Yeah. So we are adding

statutes and regulations next.

And then after that, that's by

far the next most requested thing,

statutes and regulations. Incidentally,

people generally are accurate.

I've learned in what

they say about statutes,

generally they're

copying from the statute.

People don't paraphrase

statutory language.

And then depending on what people want,

we have other techniques we've already

figured out for checking things like

restatements, law review

articles and the like.

But given what courts tell me and

tell us and what lawyers tell us,

99% of what people care about is

case law. That's what makes the news.

Well, and there's the most room

for interpretation and error there.

Exactly. Right?

So if you think about what are people

really duking it out over with statutes,

it's the interpretation of the statutes,

which is either in case law or

in dictionaries or elsewhere.

It's not really the

statutory language itself.

It's not the actual wording of the

statute. It's what the wording means.

And just a quick follow up on that.

We have a lot of case law that's released

here in our state that is outside of

the official reporters that's

still binding authority.

How does midpage handle that?

We identify unpublished cases and we

let you know whether we are or are not

able to verify them. So again,

we're greatly reducing the

number of things that you check,

but we will never ever say that we verify

a case unless we can actually find it.

What you're talking about, of

course, exists in other states.

It also exists at the federal level like

Westlaw and Lexus unpublished cases.

It's tricky. It's not just a question

of databases. It's a question of like,

is it really precedential?

What's its status? Another thing,

these are all things you

learn. I didn't know.

There's a day when it's like a

kid graduating from preschool.

There's a day when the unpublished case

becomes a real case and then you need to

have a correspondence

between the two citations.

So what we're doing now is

erring on the side of caution.

We are telling you that something is an

unpublished case and you want to make

sure you verify it yourself unless we

are confident that we can verify it for

you. It's just not possible the rate

that these cases are issued and the form

they're issued in to give

you full confidence all the

time in unpublished cases.

Well, the universe of unpublished cases

are just so vast. I don't know how ...

I mean, it's mind boggling how

many unpublished cases there are.

So I know that would

have to be a challenge.

It seems like the right approach to

me to flag it if you can't say with

confidence and I suppose in the

flag, does it come back and say,

"You need to check this case

manually or something like that? ".

Yeah, it's very clear that we

are not willing to verify it.

We let you know you want

to check this manually.

We'll say not verifiable is the actual

wording and it will explain it's not

verifiable in the sense that you

should track it down yourself for sure.

In most briefs,

what that ends up meaning is that we're

taking care of 90 or 95% of the things

you would be doing painstakingly with

the risk of human error yourself and so

about 5% will remain.

What about people who are concerned

about confidentiality, work product,

that kind of thing in the idea that

you're extracting information out of work

product and putting into

a third party database?

Right. So we followed the most

stringent security protections,

the same ones that we follow

with briefcatch in general,

which means we don't see a word.

We at briefcatch don't even see a word

that you're writing. So we have, again,

you said, you told me not

to make a technical ...

I can't make a technical because I

don't know all the terminology myself,

but we do all the things that

everybody wants us to do with

encryption, with lack of data

retention, data storage and the like.

So we're simply processing just to verify

and then there's nobody seeing a word

that you've written and there's nothing

stored and nothing is sent to Anthropic

in our case because we use Claude.

So it's comparable to using Westlaw,

Lexus or any of those in

that regard. Security,

I have an informal PhD in security in

the legal profession I was not seeking.

So it's interesting with the

courts, if you think it through,

what are the courts using it for to

process briefs? There's no security issue,

right? They've already been filed.

Now when the courts use briefcatch itself

on opinions, that's a different story.

So there's like a higher security

need than there would be.

Now it's different for

lawyers using reality check.

They're often using it

on drafts of briefs.

So it's work product and they want the

highest security protection and we give

it to them.

Okay. So that covers how you're

dealing with hallucinations.

I really like what you're saying on

that. I think that's a great service,

especially in a very timely service.

You also are offering part

of the briefcatch next

suite is something you call

site check. How does that

differ and what does it do?

So SiteCheck was a very challenging

project, I will tell you,

to fix Blue Booking errors. So again,

it's like a big time saver,

saves people some embarrassment.

So we site check again,

same extraction procedure except it only

extracts your citation because we're

not interested in what you say

about the case for that product,

only the case citation itself.

And then it identifies and

fixes Blue Book errors.

But that goes not just

for cases. We also fix ...

There's so many mistakes I've learned

people make in citing statutes,

like whether they have the section number,

whether they have multiple sections.

So we correct Blue Book

errors, books, periodicals,

legislative materials of the whole

works. That was a tough product.

That's one of those things

you just think, "Oh,

it can't be that hard." Then

you ask yourself, "Well,

how come no one's done it?

" That usually means it's very hard.

Yeah.

How far down-.

The blue book itself,

I think most people would agree is

not the simplest thing in the legal

profession.

Do you incorporate in that kind of

some of the local preference? I mean,

for example,

Texas has the green book for legal

citation on top of the blue book.

Yeah. So it's funny you say that.

So now that can come up for error and

I'm not drowning in reality check we are

adding Texas, California and

New York local citation rules.

The good news is we already suffered

through the sort of apparatus and

scaffolding and

infrastructure for citations.

So this is going to be relatively

speaking a piece of cake compared to the

original Blue Book product would not

want to live through again, to be honest.

This is a little bit of an inside joke,

but I need to run brief

catch and sidecheck on one

of my briefs citing a 15th

court of appeals case

and see how it comes out.

Oh, there you go. That

was a clean inside joke.

It was a clean inside joke, yes. Yeah,

there's been some debate

on how that should go.

And I think the Texas Law Review folks

now have announced the correct way of

citing the 15th Court of Appeals,

which is a new court that

we have in our state, Ross,

that I wouldn't expect you to

be completely in the loop on.

I actually do know.

I am embarrassed to say that I do

know what you're talking about,

but it is funny that-.

Oh, I'm impressed.

Everything in America's controversial,

even the 15th Court of

Appeals and how to cite to it.

One thing I've noticed is you have a

choice when you're using site check to

ignore the recommendation.

I've done that a few times because I

guess maybe this is in the new Blue Book,

but it wants me to allow it to abbreviate

the city names in our courts of

appeals citations. And

I don't know what it is.

Maybe I'm old school and I'm just that

guy that doesn't want to pick up the blue

book anymore.

Yeah, it's funny because we get aggregate.

We never ever see anything

at the user level,

but we get aggregate stats on what Blue

Book rules people are accepting and

rejecting.

So another one that people don't like is

having an end dash instead of a hyphen

between page numbers. I think

they feel like life is short.

I just don't give a shit.

I'm not going to like-.

This is a clean podcast, but okay.

I don't give a darn. No worries.

On that. No, it's not going

to be this one. I'm a tech.

Founder. I'm allowed to

talk like that now. Yeah.

That's right. Exactly.

I'm at WeWork. It's very crass here.

And who doesn't want to let one fly when

you're talking about some of the things

that we're talking about here? Yeah. Yeah.

I have noticed that it's suggested

endashes to me for page ranges,

which as we talked with Chris Chandeville,

I have a hard time accepting that,

but I'm trying to be flexible in the way

that I look at this and maybe I can be

persuaded that I'm wrong. So

I'm keeping an open mind there,

but I've noticed that. By the way,

just kind of to wrap up on that point,

if you don't accept,

is there a way to pre-program preferences

within the app so that I don't get

asked 20,000 times whether to

abbreviate Houston in my sites?

Yeah. So we have feature called Ignore

All and it will ignore it forever.

Maybe you'll miss it one day.

Maybe I should actually open up the

Blue Book and make sure that I'm not off

base before I reject any suggestions.

You can also have a litle bit of autonomy

and tell the Blue Book to go take a

hike. I mean, other than

in law review articles,

maybe in Supreme Court opinions,

I rarely see and Anybody following

all of the book rules. In other words,

our product is never a waste of

money. It always flags something.

I mean,

the sense I get is I just remember going

back to my big firm days as a young

lawyer,

we had a paralegal in our group whose

sole job in working with the appellate

lawyers was to site check our briefs.

And it seems like what the product is

doing now is it's getting awfully close to

replacing that paralegal. Now

as a solo, I've got a paralegal,

but I don't require her to site

check my briefs. Like the old days,

pulling books off the shelf, I

take responsibility for that,

but your product is great for that.

But it does seem like it's approaching

that level of capability where it does,

I mean, this is an overused phrase,

where it does sort of level the playing

field between different practice types

and sizes and so forth.

Yeah.

I prefer not to think about replacing

people because that's depressing and it's

not our goal and I prefer to

think of it the way you are. Well,

I'd say two things.

So one is that these legal tech products

are absolutely leveling the playing

field and making it easier for smaller

entities or less well-funded entities

to compete with the big ones,

which I think we'll all

agree is a good development.

And then the other thing is for

any one lawyer, any one firm,

it's in its best form and

that's a high standard.

It will allow you to spend more time

on satisfying tasks and less time on

tedious ones. And again,

I think that's good for business because

most clients really don't like spending

X dollars an hour for people

to be checking whether

commas are italicized after

EG and things like that.

So people have very strong emotions

about AI often like one extreme or the

other. All those things are true,

right? Yes, I'm not going to lie.

It is going to replace some

people in the profession.

I hope they find better jobs, more

interesting ones, but that's just reality.

It's already happening as you

know, but it has a lot of very,

very important benefits as well.

Oh, I love being able to run a kind of

final draft through briefcatch and catch

those little last minute things that

used to take a long time and a magnifying

glass.

And now it's going to tell me if my comma

is italicized instead of me having to

put stronger readers on.

Right.

But it is true. I mean,

that type of stuff that does save the

tedious task of going through, okay, well,

did I italicize this?

Did I miss an Oxford comma in here that

should have been in this or did I not?

It makes a difference.

And I think it's also great for emerging

legal writers because for me it's kind

of a finishing tool and a polishing

tool, but for other people,

it's just a good, "Hey, here's your

first draft. What can I do better?

What can I condense? How do I cut

legalese?" Those kind of thing.

Yeah, I appreciate it.

And this is a time a really

common complaint among

young lawyers is they're not

getting feedback and mentoring anymore.

And that's for many, many reasons,

cultural, societal, technological.

It's above my pay grade

to diagnose the causes.

But that's the other thing

I think is important.

A very good legal tech product can

provide training, can provide development.

The concern, of course,

is often the opposite that it's going to

make everybody dumb and lazy and we're

going to turn our brains

off. But I'd like to think,

and I know many users report that

this is true, that as you say,

it does help you get

your work done faster,

but it does also teach you and

inspire you at the same time.

We're both obvious fans of the product

and there's never a time that I

run brief catch on a brief that I'm

trying to get out the door that it doesn't

flag something that I either

haven't thought of or just missed.

And kind of to something

you brought up early on,

to hold up as examples of good

legal writing, Scalia opinions,

other wonderful legal writer opinions,

that's very valuable to me even at this

stage of my career to be able to see

that and think, "Oh, you know what?

That's right." And if that person is

okay with saying whatever the phrase is

that way, I should be okay with it too,

because that's the top echelon of legal

writing. So that's a huge benefit to me.

I think that's really important.

And not to get too broad here,

but I think it's one reason a lot of

legal writing instruction fails starting

from the first year course.

People just declare these things like

don't use nominalizations or use short

words.

And there's never any proof offered other

than just the authors or the speaker's

opinion.

So I learned very early on in my legal

writing career that it's your duty if

you're the one running your mouth and

telling people how to write better,

you better show that admired legal

writers actually do what you say.

And my first kind of big break was this

article I wrote about John Roberts.

Now this is many years ago now because

it was when Bush pointed him to be chief

justice.

And all I did was break it down

into five specific techniques,

one brief that he wrote when he was a

Supreme Court advocate and it was like

nowadays we'd say the article went viral

and that's because no one had ever done

it before.

You should have to prove

that legal writing advice

is actually followed and not

just that it's a good

idea. And my view has been,

if I think it's a great idea but there's

no evidence that admired writers do it

more often than the rest of us, then I

stop giving the advice or in briefcatch,

I'm often sad, but I'll do it anyway.

I'll delete the rule, right? Well.

It seems like conventions and

legal writing have changed

over the last couple of

decades too.

People are kind of cutting out those

old default rules in favor of a more

conversational, perhaps less

formal tone, which I find with.

And it's nice to see when court opinions

do it because then I think, okay, well,

if they're going to do it, I get to

do it too and they can't tell me no.

Yeah, totally.

It's funny.

There's one thing that hasn't been true

for and I finally gave up about a month

ago and it was like a funeral.

I had to delete all the rules and

I was wearing blacket in morning.

So for whatever reason,

people refuse in briefs.

If they write five and they write it

out twice in FIVE and then five in

digits,

I've been trying since the very first

edition of Briefcatch to get them to

change it to just one,

but they won't do it.

I'm in your camp on that one.

It's sad. It's sad.

Yeah.

No, I'm with you though. We

generally don't need it twice.

I think until you get checks completely

out of people's financial instruments

and you have to write both the words

and the numbers until that changes,

I think you're never

going to get rid of it.

Maybe the generation after us.

I don't think they'll even know

what a check is that generation.

I hope they'll see it in museums.

It won't be long before the

lawyers practicing don't

even know what a check is,

much less have ever written one.

It's on the bar exam, if I remember

correctly, negotiable instruments.

It used to be. I don't know.

It's been so long since Todd and I

have taken the bar. We can't tell you.

It was on the bar exam.

And I actually remember I never took

that particular class or secured

transactions and I learned so much

studying for the bar. I was like, "Really?

This is really interesting."

But that's an aside.

You probably learned so much, but it was

probably enough. I doubt you were like,

"Wow, I wish I could do this for a whole.

Semester." You're right.

I didn't regret it.

There was one other thing that I know has

gone on recently with the company that

I wanted to ask you about and that is

BriefCatch's acquisition of WordRake.

I was also a WordRake subscriber

and a big fan of that product too.

And I mostly was happy when BriefCatch

acquired WordRake because I thought,

well, okay, now I'll just have to pay one

subscription, right? But just tell us,

just kind of explain generally what

WordRake is because your company just

acquired it,

but what was the overall strategy

and the acquisition and is WordRake,

is that functionality going to

be unfolded into BriefCatch?

Yeah. So WordRake was actually the first

tech that I ever used or legal tech.

I mean, I used other tech as we all did.

It was the first legal tech I

ever used that I found exciting.

And this is many years ago.

So WordRake has been around

a lot longer than Briefcatch.

So it is sort of poetic that ended up

acquiring the company, as you said,

not very long ago, so at the

So WordRake was quite

innovative in two ways.

So on was that kind of as I tried

to do later with Briefcatch,

it tried to take legal writing and

turn it into algorithms and rules.

But the other is back then,

so you got to realize this is many years

ago is more impressive then it made the

changes in track changes at a time when

that just wasn't really a thing. So

that's what kind of turned me

onto WordRake in the first place.

Tech is very competitive.

There's a lot of consolidation.

So when the opportunity

arose to acquire WordRake,

I was very interested obviously and we

did acquire the company including its

patents. It also has a number

of interesting patents.

So there's a lot of overlap in what

the two products were trying to do.

I'm very happy that we have this fantastic

Wordrake employee named Danielle who

is kind of the guru and genius behind,

you probably noticed the product

improved a lot over the years.

The last year or two of WordRake where

they really improved their product a lot,

I noticed that was all her and she

now works for us and she's a fantastic

addition to a briefcatch in many ways.

So what we've done is

with the Wordrake users,

we found out what it was they liked

about Wordrake and a lot of it was just

certain rules that we hadn't

really prioritized here

and we now offer all those

rules. We didn't make a big

song and dance about it.

We just simply created

them in briefcat two.

And then the other thing that we're

pointing with is for some rules to allow

people to accept them all at

once, which I've always resisted,

but I'm coming around. So a good

example is, this is not a Wordrake rule,

by the way, it just made me think of it.

So a lot of people when they write

briefs and they're copying and pasting,

they'll have a mix of straight up and

down quotes and curly quotes or smart

quotes. So we're very good at flagging

that, but it's irritating, I think.

It makes people depressed because they

have to accept each one and they realize

how many times they messed up.

So we're making it easier to accept all

different changes at once. Otherwise,

there was a lot of overlap again in the

two products. And one thing that we have

that WordRake didn't have,

we have choices, right? So

if we tell you to change,

demonstrate, we don't just give you

show. It can also be like signal, say,

state. We have a lot of choices.

So even with those new WordRake

rules that we have integrated,

we offer more choices.

So you have more control than was the

case in the original WordRake product.

And are the changes showing in track

changes now in Briefcatch because you are

living in the word environment.

Yeah. So we have a new modality

called Briefcatch Express. I mean,

what it does is for people who

are in a rush, for example,

or who just completely trust us

and don't want to make choices,

it automatically makes what would

normally be the first suggestion for

everything. In track changes, you just

push one button and it's all there.

So that's an option.

And then we have the traditional

briefcatch where you go

card by card and decide

what you like and don't like as

well. So we actually offer both.

I haven't tried that one yet.

That is one thing like you I really

liked about WordRake was seeing the track

changes and being able to visualize it

there in my document and like any other

track changes is being able to accept

or reject and then move on to the next.

The cards has taken a little getting

used to, but I think I'm there now.

You mentioned overlap.

It would make sense to me if the

essential functionality of WordRake after

purchasing the company was

basically folded into Briefcatch.

Sounds like it mostly is.

Yeah. I mean, we're going to have

one product not two very soon,

but people need a little

time to get used to change.

There are still people

using Wordrake. I mean,

they're paying us now because

WordRake doesn't exist as a company,

but they get it, right?

Our salespeople just need three minutes

to show them that we do everything

WordRake did plus a lot more

and then they're comfortable.

But it's a known phenomenon in business

and in software that people don't like

change. They don't like being told,

"Well, the company doesn't exist,

but now you buy your suits or ties or

whatever from a new company." People need

assurance that the new product is safe.

Well, I know another topic that ...

I mean, we've covered this somewhat,

but another topic that you have some

strong opinions on is the use of AI and

legal writing. And so I'd like to

explore that with you a little. I mean,

obviously AI is affecting legal

writing at so many levels. I mean,

we wouldn't have all the hallucination

cases that we see if lawyers weren't

using it or misusing it is probably

the better way of describing it.

But let me ask you this

way, what do you think,

considering everything you've seen

over the last five years and then some,

as someone who gave legal writing

instruction before there was AI,

what is the proper use of

AI in today's legal writing?

Yeah. I mean, believe it or not,

even though I'm in the legal writing

AI business, I do have concerns.

I do have a heart and a

soul and I do have concerns.

I have a kid about to start law school.

I wonder what that's going to be like.

So I absolutely think people should.

They should worry about education,

about law school, about learning,

they should worry about learning

writing skills. All that is valid,

but it doesn't really matter

because there's no way this

training is going to be

going in reverse. I know I saw Berkeley

just tried to ban AI for their students.

It's not going to work. It's

like teenagers and phones,

unless you're in Australia or the UK

where I believe they have banned social

media. So I like to kind

of focus on reality, right?

We are not going to stop AI.

So my take right now as we speak,

it is still not a great idea to use

AI to generate documents from scratch.

Obviously, I think it's a bad idea

if you're talking about skills,

but even if you're talking

about very seasoned attorneys,

there are too many things that can go

wrong and there are too many ways that the

finished product can be very kind of

easily identifiable as AI. And by the way,

when I say that, I don't believe

in shaming people like, "Oh,

you used AI." I just mean there's a

certain kind of AI style. It's not,

by the way, just about M dashes or

not only, but also it's more subtle,

but it's all very real.

So I don't think people should use

AI for generation of documents.

Now on the other hand, I'm not naive.

I know there are people who can't afford

lawyers. I know that some documents are

very straightforward. Now,

if you use AI to generate a

set of interrogatory requests,

it's probably going to be okay.

But any kind of really important

strategic document, we're not yet there.

On the other hand, there are a lot,

lot of great ways to use AI

for things like fact sections,

building fact sections. But again,

you've got to decide what

facts do you want to emphasize?

Do you want to go chronological or do you

want to have some other organizational

device for building things, no doubt

about it for building tables, images,

for comparing things like witness accounts

or competing accounts in depositions

or you guys are appellate types.

If you want to compare the trial

transcript against some other source,

it's great for that. And within reason,

it's very, very good for editing.

I think it's generally a bad idea to just

take a brief and put it in ChatGPT and

say,

"Make this better." I think that is

almost as risky as having it generated in

the first place. But if you have some

sort of intermediary tool or some way to

ensure higher quality and

exhaustive coverage, it can be very,

very good for editing. For those who

are going to do it, matter what I say,

I'll give a couple of

tips. So at the very least,

if you put a draft in and you just say,

"Make this better," and I know probably

right now 100,000 lawyers are doing

that, whether I think it's good

or not, ask it to at the end,

let it do its thing you want it

to do and rewrite your document,

but ask it to enumerate at the end every

single change it made and then explain

why it made that change.

And even if you don't care and you're

not going to read it, it will actually,

this has been proven in many studies,

but you don't have to read them.

You can just trust me.

When you ask AI to justify

what it does and enumerate it,

you will get better output. I don't

know. I can't explain. It's over my head.

You have to ask the computer

science PhDs why that's the case,

although I have an intuitive sense,

but at the very least you want to force

it to list and then explain every edit

it makes. Yeah.

That makes a lot of sense.

I'm curious about what you said about

legal education and some schools actually

banning students from using AI.

I know that there are some schools that

actually require it as part of their

curriculum.

And so it's going to be interesting to

see in the next few years how that plays

out and who's going to be at an advantage

in the marketplace when these young

lawyers get out.

I think another related thing I'm seeing,

which I approve of is even

offering certifications.

Now it's kind of a thing I think in

law school. It's almost like a major,

like international law or I

did trial ad. So now they have,

I believe certain law schools are letting

you take a number of classes and maybe

even have some practical experience.

I know because we get a

lot of law students wanting

to work for us and this is

why they get credit for it. I think that's

a very good idea as well. And again,

like if you go to Vegas,

if you like betting,

who's probably going to end up

being right? The ones who said like,

"We're going to ban it, don't use

it, you're not allowed to use it.

" All we care about is that it's not

fair and grading or the ones who are

innovative. I think if you look

at the history of legal education,

generally the ones who are a

little bit braver, let's say,

or a little bit more innovative

end up being proved right.

It does seem a little naive to think that

you're going to educate a law student

and send them out into the world where

those first few years of practice are so

critical.

And we're seeing the trend now where

firms are hiring fewer new lawyers

and the investment of time it

takes to train them and everything,

the industry's changing and I certainly

hope that the law schools are sort of

forecasting that and building

that into their curriculum.

Yes, but I also understand

the other argument.

I think that it's in good faith and

it's well intended and the idea is yes,

you need those skills, but you still need

all the regular lawyering skills too.

And if you go through law school never

really having to work and never having to

struggle and having to write all your

papers for you and even having it like

distill all the cases so you don't have

to read them and that way doing our

students a disservice too. So

what I would guess, and again,

I did teach law school as an adjunct,

but I don't anymore and I'm frankly

glad I'm not teaching right now.

I think it's a tough time. But

what I would guess is the really,

really smart law schools will figure out

how to give the students all of those

skills and opportunities. And then

for things like grading and papers,

they will figure out some way.

I mean, they're very smart people,

these professors and deans, they'll

figure it out, I don't know,

close book tests or make them turn off

the internet and make them write papers

without GenAI. I don't know, but I can

guarantee you there's a way to do both.

Well, and you just ultimately,

you can't remove the human judgment aspect

and replace it with AI and the places

that balance that are going to

have the better rounded students.

I think that's right.

I think maybe law school doesn't

focus on judgment and strategy enough.

And if you remove some of the sort of

antiquated elements of law school and

turn some of them over to tech,

then maybe there is more room for some

of these higher level or more interesting

skills and topics too.

Well, Ross,

we appreciate this so much and hopefully

we can have you back and not wait five

years for the next update.

But as we close out,

our tradition is always to

ask for a tip or war story.

I don't know if you've got one or

both that you want to share with us.

We'd love to have it.

Okay. So a tip would be

you're kind of an AI,

legal AI skeptic.

A tip would be to give it something that

you've written that's persuasive and

say, if you are opposed to me,

what parts of this would make

you squirm because they're right.

And on the other hand, what

would be your best con arguments?

That's my go- to when people

say, "I don't believe in AI.

I think it's terrible or evil

or a joke." I'm like, "Well,

just try what I just said." 99%

of the time they're like, "Wow,

this was actually really,

really exciting." The feedback I got and

it also makes me realize the benefits.

And I'll think it's a little bit

less evil than I did yesterday.

So war story off the bat,

this isn't really a war story,

but I think one of the most under-reported

developments right now in the whole

kind of debate over legal

tech and the profession,

I don't think people realize how much

AI the courts are actually using as we

speak. They're not supposed

to publicize their processes.

A lot of people are really

misunderstanding these

sanctions orders and thinking

that they reflect a general

aversion to AI. It is not true.

And just one example, I'm not

telling tales out of school here,

this is also public.

There's been an influx of pro se filings

overwhelming most of the nation's

courts, appellate

especially, but also trial.

And courts actually do

take these seriously,

but they were overwhelmed even

before GenAI with pro se filings.

Now they're getting 50 page briefs that

are clearly written by ChatGPT or Claude

all day long.

So they are absolutely desperate

for help and they are using many,

many AI tools and AI

prompting techniques in many,

many ways that just

people don't understand.

So that's what I think that's one of

the biggest myths out there is that the

courts don't want the profession using AI.

That's not true and proof of that is

they're using AI themselves.

Well, Ross, this has been great. Thanks

for coming back. We look forward to,

as Jodi said, maybe not five

years before you come back again,

but we'd love to check in with you a bit

and just kind of see what's going on in

your world and how the

industry is changing and we

appreciate your perspective

on all of this.

Thank you and try to stay

cool in Texas this summer.

We'll do it.

No kidding.

Okay. Great to be with you guys.

Thanks for listening to the

Texas Appellate Law Podcast.

If you enjoyed this episode,

please share it with your colleagues

and rate and review the show on your

favorite podcast platform.

To connect with us,

suggest a topic or inquire

about being a guest,

visit TexAppLawpod.com or

find us on LinkedIn and X

@texapplawpod. Produced

and powered by LawPods.

The views expressed by the participants

on this podcast are their own and not

those of their law firm's

courts or employers.

Nothing you hear on this show establishes

an attorney-client relationship or is

legal advice.