Opportunity costs: Artificial intelligence in the legal industry, with Michaël Hassan

Introductory Speaker: Hello. Hello and welcome to Torys Live podcast. We're thrilled to have our special guest today. His name is Michaël Hassan. He's the Managing Director and General Counsel of Sagard. Michaël is joining us today to discuss artificial intelligence.

So Michaël, thank you so much for being here.

00:37 - 00:38

Michaël Hassan (MH): Thank you. Thank you for having me.

00:38 - 00:49

Introductory Speaker: Of course, I'm going to turn you over now to our host, our esteemed tax partner here from New York, Scott Semer. So, Scott, the floor is yours.

00:49 - 01:13

Scott Semer (SS): Great. Thanks a lot. Thanks very much Michaël for joining us, we really appreciate it, it's great to have you here.

So, I want to talk about AI. And one of the things I wanted to ask you was, as someone who is part of an organization that obviously uses a lot of legal services—and as a lawyer yourself—what makes you most excited about the prospect of AI in the legal industry?

01:13 - 01:46

MH: Okay, well, I have been excited about AI for, for about eight years. Initially... initially it was exciting just to know that you could extract things and do things a little quicker. And I think about two years ago, it really, really became very exciting when you started to see things like ChatGPT and other platforms that could do generative AI.

And what I'm excited about is not so much, you know—or I guess we're going to talk about this—like replacing everybody, Scott, and nobody, nobody can replace you, Scott, that's for sure. It's not about replacing lawyers, it’s about making our lives—your external counsel life and my in-house life—a little easier. Being able to focus on higher value work: work that is really much more important than the clerical things that drag us down, that slow us down and, you know, making us sound, and be smarter. I think that's what it is.

02:30 - 02:31

SS: Yeah, that's super exciting. I guess one question, and that is, you know, what do you think law firms should be doing with respect to AI, especially law firms that maybe aren’t large enough to invest, you know, hundreds of millions of dollars in developing their own AI? Or even if they do, should law firms even be developing their own AI? Or what would you most like to see law firms that are working for you be doing with AI?

02:54 - 03:20

MH: Okay. Well, what I'd like to see is some openness from the law firms to just experiment. Experiment.Talk to me about it. Tell me if you want to collaborate on something. I want you to be open to something that you are uncomfortable with. And that's the reality that I'm facing right now.

I'm speaking to a lot of law firms and working with them and one of the questions that I asked since then, how are you going to use AI in this transaction or in this process? And many of them, say, I'm not. And that's a disappointing answer because I would, I would expect them to start talking to me about what they're doing in the background to make their services more efficient, you know, focus on the quality of the lawyers and their experience rather than just working on clerical matters.

So, I want law firms to start experimenting. I want them to have that open mindset. What I'm discovering is that it's not there yet. You don't need to, to invest a million dollars in AI tools or develop your own AI. You could just simply get (believe it or not) a ChatGPT4.0 license—it costs $20 a month—and just play around with it. You can remove the learning feature and just keep your own little site so that nobody learns from or uses what you put in it, and just play around with it. See how you can become better.

04:40 - 05:01

SS: Yeah, that's interesting. One of the interesting aspects of AI, right, is that hopefully it'll make what lawyers do easier and less time consuming. And that'll [make it] a lot more efficient to do things. But as you know, lawyers have traditionally charged on a cost-plus-hour basis as kind of a proxy for work.

But with AI, there'll be maybe the same amount of work with less hours, or some of those hours will be kind of back office, in terms of training the AI, right, so there’s a lot of energy that goes into training the AI. And there'll be cost of licensing the AI and developing it. What are your thoughts on how can law firms manage some of that and still give you value but maintain a pricing model that kind of makes sense for them?.

05:31 - 05:57

MH: Well, that's the million-dollar question, or the trillion-dollar question, depending on how you see the legal market. For sure, you know, this is a bit of a challenge because we've never managed as an industry to get away from the hourly rate model. There were spurts and waves, and there continue to be attempts at seeing if you can use AFAs—alternative fee arrangements—in the legal profession and it hasn't worked really well. Lawyers are fantastic at, you know, doing budgets. They're not great at respecting them and they're not so great at keeping you … like they want to do a really good job, and it doesn't really matter sometimes—I'm not talking about Torys, I’m just talking generally—or in my experience, they really want to make sure that they cover every single angle. However, that's not necessarily what clients want. And I think AI is a bit of a solution to that and you don't need to check every box, and it can it can help.

Now, how do you price this out? I think firms are going to have to review their model. They're going to have to review their headcount. They're going to have to review how they train young lawyers and what these young lawyers generate in terms of hourly, you know, billable hours and how much revenue they generate. I think that's going to be the big challenge for the industry.

In-house, it's a similar challenge, right. We are looking at this… typically, the typical structure of a of, you know, an in-house legal department, you'll have a general counsel and then you have a couple of VPs, and then you go like some sort of pyramid down. And with AI, you might not need that type of pyramid. You could be, “Okay, I think I'm safe, so the general counsel is safe”, and then you've got a pool of people that can use AI to supplement and be able to be a lot more cross-functional thanks to the AI tools to supplement the experience that they don't necessarily have.

But it is the big question right now, I don't have the answer for you, Scott, for Torys and other law firms. But they have to think about it, and part of it is, you know, experimenting and speaking to clients as to what they think it should be, right? But it's definitely not the hourly rate model. I think if you tell me that you can go faster, I will be willing to pay a premium for the work that you do.

08:33 - 08:52

SS: Yeah, it's interesting, right, because there are two kind of models we both kind of proposed. One is maybe you end up with just sort of, you know, only top-level legal talent, as you described. And you have, instead of a team of lawyers charging each, you know, $1,000 an hour, you have one lawyer who uses AI and bills at $10,000 an hour. And another approach is, you just sort of agree on a fixed fee and then the law firm kind of manages that as best as they can.

I’m curious if you have a view? I mean, the traditional benefit of the hour is that it provided a kind of transparency to you, right? You can see how many hours people worked and are you really paying for value? If someone's just charging you a fixed fee a) they may not, you know, not do as much work as is needed, right, because as they start to increase the amount of work they've done and get close to the cost exceeding that fixed fee, there may be an incentive not to work as hard as you would like them to and, and make sure they're doing an outstanding job. And it also gives you a little bit of transparency. So those two models though, which one do you think has the most appeal to you as an outside general counsel?

09:43 - 10:06

MH: I think the certainty on fees—so having a fixed fee—is definitely something that would be desirable. And then, you know, we agree on what the cost of it is and then how you get to, I don't know, like $100,000 is your problem. And you're able to use AI to build and create efficiencies and create a work product that is that is acceptable.

But you could also look at this AI model, a little bit differently. You know, for example, if you've got extensive due diligence to do on a large portfolio of real estate or towers, et cetera, you wouldn't be using a law firm to do the due diligence on 8,000 tower leases, right? You would spot-check it and you would scope it out and say, okay, I'm going to review, you know, 10% of the revenues and I'll review the biggest leases on top, and then I'll just make some assumptions that, you know, the rest should be okay because it's not driving. If I find a pattern, then I'll do a little bit of a deeper dive. With AI, you have an opportunity to do the review of the 8,000 leases. And I think using the fixed fee approach, and also being able to scope out exactly what you want as a deliverable and what you want in terms of level of due diligence and risk appetite, is the way to go.

But for sure, my business folks are looking for cost certainty. And historically, law firms have been poor at providing that cost certainty, with the use of budgets or the hourly rate. So, there's definitely something there and law firms have to become a lot better at, you know, creating the efficiencies and building them out themselves. And then they can compete as between themselves: like, okay, who's the better law firm in the in the grand scheme of things, who has the better talent? And how well you use these tools to generate those great deliverables.

11:52 - 12:29

SS: Interesting … I'm curious what your thoughts are on, you know, if you imagine a world where it turns out that lawyers just aren't very good at managing AI and that AI is kind of ubiquitous, in the sense that there's not really much difference between one AI diligence program and another, and there's no reason why each law firm should have it and you just kind of split off a lot of the services that law firms used to do, and you have maybe a couple of very large AI firms that specialize in those types of services, and you're hiring the lawyers just to provide the actual advice and not to run those processes, and so law firms actually get a lot smaller and a lot of the work the law firms are currently doing aren’t done it all by lawyers?.

12:36 - 13:03

MH: There's definitely something there. I think that's likely to happen. But, I see it as not much different from the waves of outsourcing, you know, to, on-shoring or running some due diligence from Latin America or India. I don't see it is very much different from that. You use a tool, and it just accelerates, and you're still interpreting a deliverable. Right? What you're trying to use AI right now for is to accelerate the review of massive quantities of either unstructured data or structured data and to, you know, distill it into something that needs interpretation. And that's where somebody like you, Scott, with a wealth of experience in tax, et cetera, is able to look at it and provide value, and just provide something and a perspective and tell me “this is low risk, you don't have to worry about that, I see a pattern here”. I need lawyers to kind of focus on those types of, those types of—that's the value that you bring, or that a law firm can bring to me.

It's insight. So, how do you take data and turn it into real insight? It has been a slow process, historically, for law firms, and now you suddenly have the ability to accelerate that. You could also mean that law firms are able to do many, many, many more transactions than they have been doing in the past, so that you have a certain capacity for it, and you can do it consistently and just quicker, right? So, you might need the same number of lawyers, but you'll be scaling. So, instead of having like a Torys that has maybe 500 lawyers, you will be able to do, 10,000 transactions instead of doing, I don't know, 600 or 1,000. I think that's how I see the value of AI, in just being able to scale your business so much more than you have in the past.

What does that mean, when you scale? Well, it might mean that, you know, you don't cost out, the 10,000 transactions the same way you costed out the 500 or 600 transactions you’ve done before.

15:13 - 15:33

SS: Yeah. It's very interesting. Right? It's one of the conundrums that have often faced in law firms. You mentioned an interesting point, which is, you know, one of the values that lawyers provide, even in the world—or especially in the world—of AI, is the insight and the experience to know, what does this data actually mean, in terms of the transaction? And having obviously been a young lawyer yourself, you know the traditional way we train young lawyers is, we kind of bury them in work, right? And you kind of sink to the bottom. And the ones that manage to flow back to the top end up becoming senior lawyers who have the experience to understand the data. Where you have AI maybe doing a lot of the stuff that younger lawyers would do, how do you think we're going to manage to keep giving that training that allows people to eventually gather that insight and that knowledge and experience?

16:06 - 16:39

MH: That's a great question. I don't have all the answers on it, but I'm going to give you what I've seen around in the business. That is a big challenge. It's really tough to be a young lawyer and being trained right now, because you can see that—there are actually studies, Scott, that show that, with respect to due diligence, AI is better than, or faster and better and more precise, than a senior associate in the law firm.

That's pretty scary information. So how the hell do you train these young lawyers—young associates and interns—and make them the next Scott Semer? Not that there were many Scott Semers, I think when they made you, Scott, they broke the mold. But it is a difficult task.

What I've seen some law firms do, especially in Silicon Valley, is that they're teaching young lawyers how to prompt. So, prompting is now part of the curriculum in training lawyers. So they teach them how to prompt AI: how to interpret the results that come out of AI, how to question it. I think one of the skills that might be missing, especially in the young lawyers today—and we're not necessarily testing for it because we're testing for high grades, we're testing for things that are more classic, classic grading systems, et cetera—but what you really need is a person that has critical thinking. And once you have critical thinking, you can look at a deliverable coming from AI and say, “That doesn't make sense. I need to change the prompt. I need to help it and feed it something else, or another data point, in order to get to the right answer.” I think a critical thinking is part of it. Learning how to ask the right questions is another part of the answer. And yeah, law firms are definitely going to have to change how they are training. You're no longer going to be able to review 10,000 leases and you're suddenly going to become a leasing expert. That's not going to be it.

18:40 - 19:04

SS: It's interesting. One of the things I was thinking when you were talking about that is, in a way, the future of law firms that you're describing (and lawyers) is to become more like you are in terms of as a general counsel. You're not reviewing all the leases, you're not doing all the due diligence, but you're getting tons of information fed to you by various law firms and various advisors, right? By more information than you could even really conceive of. And your job, in part, is to apply critical thinking to that information to say, “Okay, which of this is actually relevant for us and for the deal team? And what and why is it irrelevant?” And because you have access to all these external legal providers doing a lot of heavy lifting, and now a lot of lawyers will have access in the same sense that you have, because AI will serve the role of every lawyer as kind of outside counsel. And so, the AI will be doing a lot of that work and the question will be, how do you apply that? So, I'm curious about your thoughts on, you know, are we moving to a world where lawyers become more like general counsels?

19:54 - 20:18

MH: I think there's a little bit of that. Clearly, you know, everything that you're going to get is going to be fairly generic, right? And so, you're going to have a fairly generic answers. What I'm able to bring, or another general counsel or outside counsel is able to bring, is perspective. Critical thinking and perspective. So, what do I bring when you give me a work product, when Torys gives me a work product? I challenge it with my perspective. This is the risk appetite that I'm prepared … this is the risk filter that I'm prepared to give it. And this is how much risk that I'm ready to give this particular situation. AI can't do that—can't do that yet. I can ask it to adjust it, but ultimately the risk appetite of a firm is going to be different client-to-client.

So, if you spend a lot of time with Sagard, Scott you’re going to get to know me, you’re going to get to know the firm, you’re going to get to know a lot of the principles, like how they operate, how they think, et cetera. And that typically will feed how you are going to adjust the work product that you prepare for us, right? I think AI still needs that. And it will still require perspective and those inputs that are not necessarily—I’m going to coin a word—AI-able. How’s that?

21:37 - 22:00

SS: Yeah, that’s fascinating. Well, it’s been a fantastic discussion. I want to end on kind of an interesting question, focusing on our young lawyers and the future. If someone was—let’s say it was a third year in college—came to you and said, “My parents were lawyers” or “I’m thinking of going to law school,” you think that’s a good career choice? Or do you think they should choose plumbing, or something where they’re relatively protected from AI?

22:04 - 22:44

MH: I think there’s always going to be room for lawyers. They’ll never be able to kill all the lawyers. So, I think the way —and the way I would present is in the legal profession—is that it's here to stay. We are not going to be replaced by AI. The only thing that AI will do is it will help us make better decisions—decisions that are informed by data, that are informed by real insights, by experience— and that is probably the most amazing thing that this all brings, right? It's going to be readily available to a young lawyer, to an older lawyer, to an experienced lawyer in one sector that doesn't have experience in another. It levels the playing field in a way that you didn't see before. So, it's really exciting to see what's happening.

It's also going to help young lawyers that have problems with drafting, that are not so great with drafting. It's going to help them. It's going to help people prepare work products that are higher quality. It's just going to help you get to a level of maturity that you didn't necessarily have at that particular point in time. It accelerates everything and that is what's exciting to me about AI. I use AI every day.

23:48 - 24:08

SS: Yeah, it's fascinating. That's a great message of hopefulness and excitement, and it really emphasizes that there is a tremendous potential for AI and really a lot to look forward to. And so, I really appreciate you come in to talk to us today. It's been a great discussion and I look forward to catching up later and seeing where AI and everything else takes us.

24:08 - 24:11

MH: Thank you so much, Scott. Always a pleasure speaking to you. Take care.