How Do AI-Native Law Firms Work?

You’ve probably heard a lot about how AI-native, NewMod law firms are changing the game. But how do they work? How do you make a profit on fixed fees? And what kind of lawyers are they hiring? Artificial Lawyer talks to General Legal’s co-founder JP Mohler about all of this and more.

Lawyers at traditional firms wonder how you make a profit on providing a contract to a client for just $500. The answer is that you can indeed make money this way – if you place AI at the heart of the process, rather than at the side of it – and General Legal estimates it’s hitting around a 40% profit margin on such work. Plus, it is aiming to scale, scale, and then scale some more…!

And you may also ask: so where does the AI fit in and where do the lawyers fit? How does it mix these two key components together? Mohler tells AL plenty about this and underlines that as AI advances they’re expecting the time lawyers spend on each contract to steadily drop.

And then, what type of lawyers are they hiring? Mohler explains that they’re hiring experienced lawyers who are able to make their own judgment calls as to what is ‘good’. Some may have seven or more years of experience, and they may come from inhouse or from Big Law. But, they all have one thing in common: they want to work within a new model legal business.

There is an AI-generated transcription below, but it’s a lot more fun to watch the video. Enjoy. Please press Play to watch here, or you can also go to the AL TV Channel and watch over 200 other videos for free.

–

Note: General Legal earlier this month announced it has hit $11.5m in Seed and pre-Seed funding since launching officially just three months ago. Mohler’s co-founder is Ryan Walker, who previously was CTO at Casetext and then a VP at Thomson Reuters. As explored in the video interview, they started with a focus on contracts that startups and scale-ups need, including serving some of their Y Combinator cohort peers, but are already graduating to serving larger companies and plan to expand their offering. More here about General Legal.

—

A Legal Tech Conference For All of Europe

Legal Innovators Europe – Paris – June 24 and 25.

Look forward to seeing you there!

Richard Tromans, Founder, Artificial Lawyer and Legal Innovators conference Chair.

Note: the conferences are organised by Cosmonauts – please contact them with any queries. 

If you would like to be a speaker at Legal Innovators Europe, especially if you are at a law firm or inhouse legal team in Europe – whether based in France, Belgium, Spain or Germany, or beyond…..then please contact Phoebe at Cosmonauts:  phoebe@cosmonauts.biz

Note: if you are a legal tech company, please contact Robins: robins@cosmonauts.biz or Anjana anjana@cosmonauts.biz

–

And if you’re in the US and looking for the next major event to join after Legal Week, then see you in California this June!

Legal Innovators California, the landmark West Coast legal tech event, will take place on June 10 and 11, in the heart of the Bay Area, the home to many of the world’s leading AI businesses – and plenty of legal tech pioneers as well! More information and tickets here.

—

AL Interview Transcription

Hey everybody, Richard Tromans here again, Artificial Lawyer TV. Great to have you with us, JP. Tell us first a little bit about General Legal.

J.P. Mohler (00:47.726)
Absolutely. So I am the co-founder of General Legal. We are an AI native law firm for commercial contracts. So what that means is that we have built ourselves from the ground up to run MSAs, DPAs, NDAs very, very quickly. We have a team of lawyers who sit on top of a very powerful artificial intelligence stack who can analyze, turn those contracts for our clients much more quickly than you would at a traditional office.

Richard Tromans (01:18.982)
Fantastic, fantastic. And that is of course, as anyone who’s been keeping up to date with artificial lawyer and the market in general will have understood that, know, new model law firms, if you want to call them that, or if you want to call them AI first, AI native, it’s fundamentally this fusion between the AI and the lawyers, but it’s placing, and some people have said to me, I don’t understand the difference. Lots of law firms use AI. What’s the difference? The difference from my perspective is that you’re quite consciously saying no,

The AI comes first. Yes, some lawyers might set up workflows. might develop data and work, playbooks and so forth to get it to work. And then they will come and do quality checking and secondary editing and so forth. But the core engine is the AI. And that’s the fundamental difference.

J.P. Mohler (02:08.974)
Yes, it is. would say that any law firm that is built from the ground up with AI in mind, I would consider this new mod AI native law firm. Because the technology has moved so fast, most traditional law firms are built with humans in mind, where the human lawyers drive the process. And then they might use AI here and there to speed up particular components of the process.

For us, the entire process is built end to end with AI. So as soon as a document comes in, the first set of eyes on that document is not a lawyer. It’s an AI. The way that the document is routed to one of our lawyers, it’s done with AI understanding our lawyer specializations and understanding what they’re good at. The communications from the client are analyzed with AI so that we actually kind of understand the psychology and the background of

of that client and how they feel about that document and where it should go. Again, the first set of markups, the first set of notes, also the client context and the matter context is all done with AI to prepare that lawyer for the first take. And the idea is that you actually want the lawyer exercising judgment on top of all of the routine work that the AI can do. And I think that’s difficult to…

of construct the architecture of a firm like that if you’ve been practicing as a non-AI firm for a long time. It’s certainly not impossible though and other firms can adapt to kind of a new new mode model like ours. It’s just I don’t actually see it happening in reality.

Richard Tromans (03:53.604)
Yeah, well, it’s going to be very, very interesting. We’ll come back to that a bit in the end. one thing, I mean, thank you for explaining that. And one thing I really want to get into this interview is the economics. As I was just saying off camera, was talking to someone who’s a legal tech expert, used to be a partner in a big law firm. And he said, yes, this looks great. But the problem is, from his perspective, how on earth do you make $500 contracts into a profitable business? I mean, if they’re very, very cheap, bottom of the market, you know, very, very, very consumer level, maybe. But if you’re doing what is gonna be quality work for a big corporate, I’m being slightly devil’s advocate here, but you know, let’s play through this. Let’s play through this, this point. How do you make money with that kind of economics?

J.P. Mohler (04:42.542)
If you look at the average master services agreement in that that’s sent to a mid-law or a big law firm, the lawyer takes between eight and 10 hours on average. And there’s a wide range because, you know, the complexity of the MSA will vary. The degree to which the counterparty wants to mark it up will vary. But on average, it takes about eight to 10 hours. And our attorneys can mark up an MSA.

including all of the turns and include taking into account the occasional necessity for jumping onto a call with a counterparty in about 2.2 hours. So we have gone, we have essentially cut out about 80 % of the human process. And that gets us to actually just today without even making any more improvements to our AI systems to real margin. We’re at 40 or 50 % of margins per contract. We think that

You know, it’s important not just to bet on where the technology is now though and what we can do with the technology, but where the frontier models are going. They are becoming much more capable. And now we might mark up a contract, present it to the lawyer and the lawyer might change half or 25 % of the markups. They might add a few of their own markups. I think over the course of the next six months to a year, we’ll be at the point where the AIs first cut

Richard Tromans (05:47.494)
Yeah.

J.P. Mohler (06:06.466)
becomes closer and closer and closer to exactly what that lawyer would have ultimately done, until it gets to the point where the lawyer really is overviewing. It’s like a partner overviewing work of their best associate. The partner might have a few opinions of what needs to be changed, but they will be really just doing a quality assurance check, doing this very, very high level strategic judgment check.

Richard Tromans (06:12.784)
Yeah.

J.P. Mohler (06:33.484)
And I think in this case, and I think we’re maybe six months away, nine months away from this, given the pace of the frontier models, the lawyer will I think only need to spend a few minutes per turn to ensure high quality output from the AI. And at that point, I think our margins go to more closer to software margins.

Richard Tromans (06:53.212)
So, okay, so you’re getting about, so 40 % profit margin. So if you’re doing $500, you’re collecting, you’re keeping about 300 effectively. Obviously you’ve got other costs that are background there and tax, et cetera. But generally as a sort of top end of a funnel working through the costs, you’re getting 40 % straight away. And also, of course, I two things just bring, know, referring to what you just said.

J.P. Mohler (07:03.566)
Mm-hmm. Yep, absolutely.

Richard Tromans (07:19.548)
the models will get better. Your knowledge, if you’ve got a repeat client, it will get better and better, particularly if they’re sending you the same, know, here’s an NDA again, here’s another NDA, here’s another NDA. You’ll like, yeah, well, we’ve done 40 of these. We know everything about what you do. And then secondly, the key point is that if the lawyer on their side or, you know, a counterparty, whatever it is, wants to spend all day or weekends playing around with it, redrafting it, that’s fine because that’s time on their side….when the contract comes back to you, you zip through it again.

J.P. Mohler (07:53.494)
Yeah, absolutely. And something we’ve actually found is that if there’s been a week long or two week long term before our lawyers put eyes on it and we turn the contract in an hour, we somewhat kind of shame the other side into not taking two weeks. And so that next turn often comes back in a few days because they’re aware that they, and this isn’t in every case and it isn’t for every firm, but if they’re aware that they took two weeks to turn something,

and their counterparty turned it in the same day they got it back to us. There’s some obligation to not consistently be the side that is holding up the negotiation. I sometimes that’s used for leverage and it’s not in every case, but yeah, that’s what we see.

Richard Tromans (08:32.667)
Yeah.

Richard Tromans (08:40.604)
And certainly for things like an NDA or an MSI, you’re like, just guys, let’s just get this done, right? You know, like, let’s use General Legal, let’s use whoever, there’s a whole bunch of them now. There’s 27 at least, as you may have seen today in artificial law, published, Matt Polins, really great legal tech expert, which has published a directory of such companies. But okay, let me ask you another hard question. What kind of salaries are you paying

these junior lawyers, presumably they’re junior lawyers, not equity partner level from Kirkland and Ellis, right? You know, they’re junior lawyers, I presume. What kind of salaries can you pay them and what kind of future can you give them? So let’s say I’m a junior lawyer, or my cousin is or whatever. And I say, hey, let’s go work for General Legal, it’d be great. And I say, yeah, but you know, what’s my future there? Am I just gonna spend six months and then I’m gonna try and get a job at a big law firm or?

I’m going to spend the rest of my life there. mean, how does that work? Because one of the things that law firms, I mean, we can pick apart the big law firm model in part because it’s been so successful. In some ways, great success attracts to some degree an analysis. But one of the good things you could say about big law is that it’s an entire culture. It’s an entire universe. It brings in junior lawyers. It fosters them. It educates them.

Can these new mods, can your new mod provide that kind of framework?

J.P. Mohler (10:10.232)
So that’s a great question. Right now we’re actually not hiring what I would describe as junior lawyers. We’re hiring lawyers either out of Big Law who are fifth to eighth year associates in Big Law who have already really been tripped up.

Richard Tromans (10:23.4)
8th year associates.

J.P. Mohler (10:25.462)
Yeah, yeah. I think that’s our sweet spot. And we’re also hiring folks who have 10, 15 plus years experience in-house. So I think we generally count the years in Big Law as roughly double, particularly if they’re working with commercial technology transactions, which I think is fair given the number of hours that they work and also given the range of matters that you work on in Big Law, right, or relative to in-house. You know, we bring these folks in as subject matter experts and we want folks who kind of on day one are trained up.

J.P. Mohler (10:55.31)
in the area. Now, they’ll also adapt to our system. We think that when we hire a lawyer, we want that lawyer to be there for a very long time. And the idea is that today, with our AI, a lawyer maybe over the course of the year can handle somewhere between 300 and 600 contracts. But we actually think that same lawyer, same expert lawyer, kind of sitting on top of our AI stack will be able to handle multiple thousands of contracts.
….in a short period of time without any sort of decline in quality and while working fewer hours than Big Law. We think it’s a…

Richard Tromans (11:33.372)
But let me ask you about the salary point because lawyers in large commercial firms work incredibly hard and also in-house because they get paid extremely well for it. It’s very demanding work. Are you going to be able to offer the same kind of salaries or maybe you get equity or whatever it is? mean, how is this going to work?

J.P. Mohler (11:53.976)
Sure, yeah, yeah, there’s an equity component in the C-corp. There is also a quality of life increase coming from Big Law. So the attorneys that we bring in are not working Big Law hours. They’re also not expected to be on the clock 24-7. And the reason for that is that we are able to effectively route work across the firm to folks who are on or offline in the same contract.

can actually be marked up on different turns by different attorneys while ensuring that the context of what’s going on in that negotiation isn’t lost from attorney to attorney. And that’s something that AI can do very well, which means that you’re off the clock at 5 PM. You can go off and enjoy your life and feel like you’re a real person. Having been in big law, expectation is generally that if something comes in at 11 PM at night and it’s important to the client, you might have to turn it.

in real time, then you might be back online at 7 a.m. the next morning. We don’t have that same set of expectations. folks do coming in, as you know, like the salary structure, coming in from a fifth, sixth year associate, you are taking a lower salary than you have at Big Law. It’s a similar step down that you actually see from Big Law to an in-house, like a junior.

kind of a first in-house position, the commercial counsel, say at like a series C, series D tech company, where the salary is solid, but it’s less than you make in big law, you’re making a quality of life trade-off. And then you’re also doing something that interests you. Our lawyers are really into AI. Like law attracts this very intellectually curious archetype of person. And there’s a subset of that person that right now in 2026 is obsessed with AI.

and is very eager to kind of be part of a system that streamlines their talents more efficiently, kind of to the maximal degree. And we get those people coming in who have this anxiety about the future of the big law model. They might be able to look back and say, in the last 20 years, it was very stable. But what about the next 20 years? What if I’m a fifth year associate? The path to partner is 11 years.

J.P. Mohler (14:09.894)
How long is it going to take artificial intelligence to obfuscate this role? And if that happens before I hit the end of the 11 years, what are the partners going to do? Well, I think a lot of folks have this fear that the partners are going to say, well, thank you so much for supporting my practice and my client list for the last eight years. But AI can support it from here. So we hire only partners. We only hire folks who we think are at the level of being able to handle their own work and give the final sign-off.

And think that’s, you know, that kind of along with the other advantages I mentioned is what compels.

Richard Tromans (14:45.552)
That’s point. mean, there are probably, I mean, in very particular work streams, there are very junior lawyers who effectively, you might say, are partners in the sense that they are able to take proper responsibility for that work stream and sign off on it. Obviously, they’re not partners in the sense of ownership and so forth and seniority, but in the sense that they can take personal mature responsibility for that thing. And that’s one of the other issues as well, is how many associates actually want to become partner. then second, on top of that,

How many associates you want to become partner actually even have a chance given that your chances of making partner are probably about 5%. And even in these days, back in the day, 20 years ago, once you became partner, that was it. Unless you did something nuts like rode a horse through a lobby, you pretty much were guaranteed to stay as a partner. Now, not even that is guaranteed.

J.P. Mohler (15:38.978)
Right. Yeah, absolutely. think with the rise of non-equity partners, I think also you see this significant divide between relation, what they call inside firms, relationship partners and service partners. And at the end of the day, I think that if AI continues to kind of increase in its capability at the rate that I see it increasing, you’ll have a lot of service partners who really will previously be doing a lot of the blocking and tackling for relationship partners.

J.P. Mohler (16:07.64)
who aren’t necessarily needed anymore. So I don’t think it’s as safe of a route within a big law firm as it was even just five or 10 years ago.

Richard Tromans (16:17.084)
no, definitely. Definitely. mean, yeah, I totally agree. mean, the future is going to definitely be much, much more folk in a funny way. It’s going to be back to the future because the partners who will matter will be the ones who have the client relationship with people who have very, very, very good, either very super incredible technical skills, which is so good that they just eclipse everybody else, or they’ve got very good personal skills and they’re brilliant at maintaining a client relationship. But yeah, the idea of just like, you know, here’s Bob, he maintains X contract because

he’s 40 years old and therefore he deserves to look after that contract. I imagine those dates will be over fairly soon. There’s no economic sense in it. okay, moving on to another tough question, and you’ve answered the last couple very well. Where do you go from here? So you’ve got some funding, obviously it’s nascent, you’re a startup, you have to take that into consideration, but where do you go from here?

J.P. Mohler (17:07.276)
Yeah, so one thing that we’re doing right away is we’re expanding to other categories of commercial counseling. So we’re not just providing markups on MSAs anymore. We’re saying, hey, any commercial agreement will be able to mark up. And our AI and our lawyers can adapt to it. Because again, we have good lawyers and we have very kind of adaptable AI processes in the background. We’re also saying, well, what do these exceptional commercial attorneys already do very well?

They do things like product counseling, because almost all of our clients are B2B SaaS companies. So we bring in folks who can understand, can actually take a quick kind non-AI driven 15 minute call with the client before we mark up any contract. We can then ingest all of this information and kind of background about what the client is doing, what they care about, and integrate that into the process of marking up every single document we see. Kind of if anything about that client that we now know.

in the background kind of comes into play, we can proactively provide that. There are other areas that we’re asked to do a lot. So think the two major categories aside from commercial contracting, or guess adjacent to commercial contracting, are corporate and employment. We get a lot of requests in those areas. So we’ll see ourselves going deeper in commercial. We’ll also see ourselves becoming more of a full stack.

kind of horizontally balanced firm that can now we refer out some of that work to traditional law firms, to other AI law firms. That’s not going to be necessary in the long term because we think our model actually applies to a range of practice areas.

Richard Tromans (18:51.436)
Yeah, yeah, I’ve always felt this. mean, it’s such early days, but in four or five years, I I imagine that the vast majority of these new model law firms will be doing multiple things in the same way that big law firms have multiple practice groups and you’ll have multiple offices and so forth. Just on that one point, you mentioned that you do refer some work out to other AI first law firms. Who are you sending other work to?

J.P. Mohler (19:15.106)
So I won’t say the names of any firms that we send work out to, but I would say this. would say this. I think we have a lot in common cause with any firm that says, we want the business model, we want the business structure to be aligned with, we want it to be aligned with efficiency and productivity.

We want our legal services to be provided in as fast and inexpensive way as possible. And that does require a actually a different incorporation structure than the standard partnership-only law firm. And so these firms, while we do want to be the best of these firms, we also think it’s important to have good relationships with them.

Richard Tromans (20:07.696)
Yeah. It’s still human to hear at the end of the day, it’s still human to human companies are owned by people. They’re managed by, by people and so on and so that, that won’t change. But exactly. goes back to the original point. The question is, as I’ve always called it, the means of legal production, you know, what is the real engine, the, the, the, the bulk of the legwork, who does it, what does it, how is it done? Where does the data come from? How to play books fit in and all these things. And of course, until very, very recently,

It was manual. No, it doesn’t have to be. So very, very, very, very interesting. Thank you. And just to remind people, how long has General Legal been going?

J.P. Mohler (20:48.078)
Absolutely. So we started, our firm was founded on January 7th of this year. So I think that’s two months and 23 days. yeah, and we’ve got about, I think, a hundred, a little over 110, 115 clients now and hit a million in annualized revenue run rate a few weeks ago.

Richard Tromans (21:08.636)
Which is pretty incredible considering that obviously at the beginning, your marketing environment was relatively limited, but now it’s going to expand and just, you know, pie in the sky. very, I mean, I’m not going to, no one’s going to hold you to this, but just very roughly, where would you like to get to in the next few years?

J.P. Mohler (21:25.966)
I think in a year we want to be at 10 million in annualized revenue. I think we’d want to be at a place where we have 30 or 40 attorneys across a few practice areas. I think because of the efficiency, in theory, think any one AI law firm could actually have a much more significant portion of the market share.

in any area of law, particularly in these run the company kind of day to day areas of law that we’re focusing on than existing kind of law firms today. So I think, you know, I think probably the more important statistic would be how much of the markets that we participate in have we captured. And, you know, I think we want to be, the first, you know, the we have more market share than any other AI law firm. But but also, you know, I think that market share needs to be in the 30 to 50 percent range across any industry.

Richard Tromans (22:17.946)
Yeah, I believe without a doubt that it might take a decade, but I imagine the most successful new models will be approaching a billion revenue in a decade. And they’ll be right up there in the Amlog 100 if they can scale. But there’s no reason why they shouldn’t.

J.P. Mohler (22:39.606)
Yeah, yeah, I agree with you. I agree with you on that. And I think you will see this kind of suite, as you noted, like Matt Paulins has that great chart. Like there are a lot of AI law firms kind of emerging now. And I think kind of a handful of us will, for each of the markets that we’re oriented toward, we’ll really be able to kind of get to those AMLAW and hopefully even beyond those AMLAW numbers eventually within the next decade.

Richard Tromans (23:05.82)
Yeah, it’s entirely possible. Just very last thing as well, because a lot of the new ModSpark startups are focused equally on other startups. And to some degree, that has made a lot of the big law partners relax. And they said, oh, OK, well, that’s good, because I don’t deal with little startups anyway. I don’t deal with garage bands, right? Doesn’t matter. But presumably, you’re not going to stay there, are you? You were talking about practice verticals, but presumably, you’re going to go up the value in terms of size.

J.P. Mohler (23:27.936)
No, that’s, that’s, that’s exactly right. So we started, you know, I like, I like to say our first 30 clients were all YC startups that were either seed or series a, our last 30 clients were only about half YC startups. And those that weren’t, and even a few of them that were about half of those were series B, series C, series D or later companies. So we’re actually representing, I think at least 10 companies.

that have raised over $100 million in venture capital. These are serious clients, some of which have built out in-house teams. where we fit in within there as essentially another commercial counsel within their in-house team to either supplement what they have to do the day-to-day work, or in some cases, of conversely, to do the more complex work where nobody in-house has the expertise, but one of our attorneys does. So we are in that way operating very much like

kind of an outside council that can sit there and take work that gets sent out off of the plate of the in-house teams in some cases. And again, it’s still only a small portion of our clients, but we are representing some huge companies now.

Richard Tromans (24:43.516)
Of course, not all of them, but some of them will grow to become the next Microsoft or Apple, whatever we shall say. Very, very, very last question just to end. I know on your website, it says something along the lines of $500 for a standard contract review. Is that $500 like a sort of like emblematic price that you’re gonna stick to it forever and ever? You know, like $1 tacos, right? It’s always gonna be $500 or is that just gonna be flexible and maybe in a couple of years, it’ll be $1,000, but frankly,

If you’re a successful company, a thousand dollars for a review is no big deal anyway.

J.P. Mohler (25:17.934)
Absolutely. So that’s a great question. think that the $500 that there will always be some form of review that’s at that price point. I think that there are going to be two trends. So one, as we take on more complex work, I do think the price point, for example, we’ve done a few, just a few asset purchase agreements for very, very small transactions. That asset person’s agreement for a transaction for all of your assets as a startup, we didn’t charge $500 for that. So we’re adaptable as

the contracts that we’re willing to support become more complex. There will always be some contracts that fit into that $500 bucket, I would imagine. On the other side of the equation, as we really do drive toward this point where the AI is able to really perfectly match what the lawyer or nearly perfectly match what the real expert, know, Atheer Associate lawyer, that partner at our firm would have done. As we run to that point,

There will be some contracts for which that price comes down. We already do NDAs for $250, but we are going to, and we also do three page or fewer agreements for $250. But I would predict that certain MSAs also over the course of the next 18 months come down in price as the human touch goes from 2.2 hours to 30 minutes, or 40 minutes. We’ll always need to make a healthy margin on our end.

but we’ll be able to pass on that savings as well. And that’s what we really believe in. Like any, I think, good venture-backed company, we want to be at scale. We want to run a huge quantity, and we want to have the best quality for that category of agreement. But we don’t necessarily need to make $1,000 an agreement. so it will be priced in some part according to whatever the cutting edge of AI is and what the complexity of the agreement is.

Richard Tromans (27:09.158)
Fantastic. Thanks JP, real pleasure. And we all look forward to seeing how General Legal evolves. Thank you.

J.P. Mohler (27:15.18)
Thank you so much for having me, Richard. I’m a big fan.

Richard Tromans (27:18.278)
Thank you.