Most people have just been sort of, 'Why fix what isn't broken?' But this may force the conversation to really ask yourselves, is it in fact not broken?
Welcome back to another episode of Compound Interest. I'm Liz Hoffman, Semafor's business and finance editor. My colleague Rohan Goswami missed today's interview, but we'll catch up with him on the back end. We're talking to Rachel Profit, who's the CEO of Cooley, a big global law firm with its roots in Silicon Valley. And obviously want to talk about AI, which is threatening to completely upend the business model of big law. And Rachel is right at the center, physically out in Silicon Valley and professionally as a CEO of a major law firm, of the changes coming down the pike. A lot of what law firms do is move paper and have an army of young people who do nothing but read that paper, read those contracts, and fight with the other side's lawyers. And it is not that far in the future where that will almost certainly be done by AI. And then the question is, where does that leave law firms? What is the value that they're adding? What are the economics, which right now rely on those young lawyers billing a lot of hours at a thousand, fifteen hundred dollars a piece, and funneling that money up to the partners? As AI starts to be able to replace so much of what humans are doing, what's left? The entire model is under threat. So I really want to get into all of that with Rachel. We have to take a short break, and we'll be right back after this.
Every business starts with an idea. But how can you go from daydreamer to industry leader? Amazon Business accelerates your journey. With smart business buying capabilities, you can get everything you need to grow in one familiar place. From office supplies to IT essentials and maintenance tools, Amazon Business takes the buying experience you know and love from Amazon, plus tools that help you save costs and make insights-based decisions. Ready to bring your vision to life? Learn how at amazonbusiness.com.
Thank you for having me. Nice to be here.
I started my career as a reporter covering law firms 15, 16 years ago. And you know, one thing that I think I've always thought is interesting is that your title there is CEO, but that's a relatively new thing in law firms. You know, I think for the vast majority of their history, they were not run particularly commercially by people who had real commercial vision. They were partnerships that kind of did the work in front of them. And I'm sort of curious what you make of a big law firm in 2026 as a business.
Yeah, I think this is becoming the multi-billion dollar question these days. I think, look, I inherited our management structure, and this has been, I mean, over a decade. I think we evolved this for exactly that reason, that we think that there's both the practicing of law, there's also the business of law. So I am a believer in it. I mean, I think at this stage it's hard to ignore the fact that this is in fact a business that has to be managed, and it's like a really complicated business in the sense that it's a people business, talent management at its core. I think, you know, as I look at the landscape in 2026, you know, I'll be honest and tell you that I don't know that I had thought that this would be like an exciting time to be a law firm leader. I don't know that the words excitement tend to come out of most people's mouths. [laughter] But it is absolutely, I think, an unprecedented time, and I think there's going to be an increasing focus on what it is that we need to do to better manage these businesses. I'm not sure that it's going to be enough to just continue to provide the same good service in the ways that we've been providing it for decades upon decades. And so, you know, I think there is certainly an element of greater management need and responsibility.
Yeah, I mean, we want to get into all of the AI, really, is what you're getting at. But just to start, you know, if you picked up a lawyer from 1975 and dropped them into a big law firm in 2025, right? Let's go back a year maybe. Sure, there's more computers, and tell me if you reject the premise of this question, but that the business of law would be pretty recognizable to them. Sort of still billing in these, you know, I think it used to be 12 or 15 minutes. Now I'm told we're down to 6 minutes at a time. But still basically doing the same work, delivered largely in the same way, but with more email. Do you think that's a fair characterization of 50 years of law firm commerce?
Yeah, as much as I don't love the reductionism of it, I think that's probably fair. I think some core tenets that stick out there, I think it's always been, or has historically really been, an apprenticeship model, right? And so the ways that junior associates become those senior partners is sort of time and grade and sort of watching and learning. But yes, I think, you know, but for the art on the wall and the outfits running around the hallway, I think absolutely. I mean, I think it would look very similar from decades ago.
And I think implicit in both my question and your answer is that it's about to start looking really different. And let's just go straight at AI. You know, these are large language models. Law is a large language, right? It has its own language and its own economy and its own workflows. And just at a 30,000-foot level, are you excited about AI or are you nervous about what it does to the product that you're delivering, how you're delivering it, and the way that you're getting paid for it?
Where I sit, this is probably one of the most intriguing conversations, but I would be remiss if I didn't note that I'm sure there's a bit of nervousness. I think skepticism is probably moving out of the field of view, but I think there's a nervousness and a lot of looking around. Like I see that a lot in the industry right now, looking around to see what everybody else is doing to make sure that you're not left out and to make sure that you're not missing that opportunity of like, well, gosh, they bought a thing, so we should buy a thing. They deployed a thing, we should deploy a thing. And it's happening so quickly that I think it's not as easy to pause and figure out, 'Okay, how do we reset the model?' And so it's going to be sort of the building while flying metaphor that's going to come to bear. From where I sit, I think there's a real opportunity to realign incentives. And when I say that, I mean something that I talk about as sort of value-based pricing. And you've seen some of that creep into the ecosystem to date where clients want to talk about alternative fee arrangements, where they kind of want to divorce themselves from the billable hour model. Like everyone loves to hate it. There's articles all the time every year about how much this model has to die.
The death of the billable hour has been predicted many, many times.
Exactly. And yet it persists, and I find it fascinating actually that sometimes even when presented with alternative fee arrangements, clients still revert to the norm, right? Which is, 'Ah, gosh, maybe we should do time and materials and keep focusing forward.' But I do think that there's a real opportunity for firms to lean into, again, incentive alignment and value-based pricing. Not dissimilar to how banks price work, right? Which is, if we're going to help you get through this transformative transaction, let us share in the upside of that with you in some meaningful way. And similarly, and that's the harder part, when we do all this work and it doesn't materialize, maybe we aren't recouping for all the time.
Yeah, and by the way, the dirty secret of the legal world is that Wachtell has billed like this forever, right? They're for services rendered with a very large number, and I guess you can quibble with it if you want, but you will not be a Wachtell client going forward. So I always wonder why that didn't catch on more broadly.
That is the lore. Yes, you can only get to ask about it once. I think it is a little bit a product of the type of work and how long cases or transactions persist, right? So, you know, maybe it's not so interesting to have a 10-year litigation say, 'Well, just talk about it in 10 years,' because, you know, all law firms have to empty their coffers at the end of the year and start again. So that might be one of the issues. And then, I think, back to your point about the way that technology is necessarily going to change what we're doing here, there's been a slow and steady disintermediation of the types of work that firms do over time. And that hasn't necessarily been AI, that's just been legal tech and other innovation, or other companies that are kind of coming in to take over some of that work. And that's the stuff that might be more productized. I think that trend is going to continue. And so really what you're talking about is that big, meaty transactional work or litigation work. And I think most people have just been sort of, 'Why fix what isn't broken?' But this may force the conversation to really ask yourselves, is it in fact not broken?
That's why I wanted to talk to you about this because, you know, Cooley is a Silicon Valley firm that's obviously global, but that's where its roots are. A lot of tech clients. So you are probably hearing from this sort of first from your clients who I assume are thinking about this harder than perhaps the sort of more blue-chip kind of East Coast Fortune 500 GCs might be thinking about it. Like, do you get the sense that they are like, 'Why am I paying for all these associate hours on this bill?'
I'd argue that some version of the statement you just made has been in the conversation for years, right? Which is, 'I'd rather just talk to the most sophisticated partner on the account for less time and be done with it.' So I don't know that this in and of itself is that. I think more fundamentally, what's happening also within our clients is that let's not forget that there's lawyers who work there, too. And so they're also feeling the pressure of, 'Gosh, I need to figure out how AI is going to change what I do internally. I need to figure out how my effective cost center is going to continue to relate to the business.' And so they're getting pressure from a lot of sources. And they absolutely, and this is interesting and we've spent some time talking to some of our largest clients, they actually have told us like, we are keen to have our lawyers use AI more. We are keen for it to help affect outcome and whether that's speed, efficiency, greater value driving, but we're not quite sure what that looks like. And so in some respects, I think there's going to be a bit of a partnership. I hope there's a bit of a partnership as this continues to evolve. But again, I think unless we want the whole business model to — and 'fail' is too strong of a word — but unless we want to upend it overnight, I think it is going to have to turn not just like, okay, we can do this faster, therefore it's cheaper, but we're going to be able to do it faster, maybe better, and there's going to be a premium attached to that sort of greater certainty for execution and outcome.
But just to say the obvious thing out loud, AI can do a lot of what you used to do as a baby lawyer with your colleagues on the transactional side. I just don't see a world where — now maybe Cooley is bigger in 5 years, but that's because you've done something right and taken share from someone. I just don't see a world where we don't have a lot fewer junior lawyers.
So I don't — it's hard for me to disagree with that. And look, I would pay good money for a crystal ball right now because it is hard to know when the change is really upon us. And like I said, that's why I go back to my metaphor of what's this going to be, the build while flying exercise. I think that it is likely the case that maybe the quantity of lawyers changes. I certainly think that what they're doing is going to change pretty significantly and we can already see that as folks are experimenting with workflows and experimenting with helping automate some of the things that they've been doing, you know, basically interacting with these large language models to think deeper, construct arguments, have it kind of play back to you the way that you're thinking about something. So I think it's absolutely the case that we are going to see evolution in that regard. And then I think the inevitable question comes, okay, so if there are fewer of the younger lawyers doing the thing, which again, I'm not entirely sure I see an absolute drastic decline, but I absolutely accept the premise that it is going to change the type of work that folks are doing and again, potentially the volume of people who are needed to do that work. How do they become the seasoned lawyers of tomorrow? And to me, I actually think there's going to be an even greater emphasis placed on what I think is one of the keys to even success today, which is the relational element and the trust element, right? And I think it is most of the time the case that folks make decisions about who to hire for their bet-the-company litigation, for their high-stakes transactional work, for those seminal moments in a corporate life cycle, or again in a big dispute, they go on brand and they make decisions based on trust. And so I think there's going to be a greater emphasis on that in the next X number of years, 3, 5, 10, doesn't matter. And I think that you're going to see a greater consolidation in firms themselves because of the fact that it's going to be my original point that this is a talent management business. I think there's going to continue to be talent aggregation and in some respects talent consolidation. And so ultimately, when you get to the places of greater consolidation and maybe a fewer number of firms who are kind of doing that bet-the-company elite level work, it probably stands to reason that there may be a relatively smaller workforce that's supporting it. But again, unless I want to go to a total dystopian state, I can't say that goes to zero because we're never going to get to a place of only robots because I think that's where the trust and relationship element breaks down.
Right. Though the sort of cauldron that has produced those real rainmakers, I believe is what they're called in the industry, has been this kind of filtering up, right? Where you have relationship people at the top and a bunch of execution people whose jobs seem very vulnerable to me. But also in the process of sorting them, is that just going to happen sooner where you're going to realize when an associate is in their second year what it might have taken you till year seven or eight to realize, which is like this person is never going to bring in a serious piece of business and we should cut them loose. Does that filtering happen earlier if really softer skills are what you're sorting for?
I actually love the question because it goes to something I've been thinking a lot about, which is the fact that this might enable us to get to a more merit-based sorting function, if you will, because I don't know how much time you spent studying this and I'm happy for you that you've moved on from this part of your life. But you know, we have to be — spot for the legal industry. I really do. Well, there is a reality that like most of the time in most firms, it's just kind of a deemed year advancement, right? It's almost the lockstep of the old partner models applied to associates. And I do think that if you start to think about the way that this could evolve, the way that firms relate to associates, it would in fact enable you to do exactly that, but in both directions, right? So you'd be able to identify those superstars who come to this business, to this profession with a variety of skill sets, right? They're super technology fluent, they are extraordinarily sticky and have high EQ, they are well networked, like whatever the variables are that will end up making a really successful practitioner. And absolutely, I think it could be the case that you could start making those types of determinations earlier.
Maybe asked slightly differently, do you think that law firms will have less use going forward for perhaps low EQ, but highly competent, highly efficient, can take a transaction from A to B person who just does not score off the charts in the charisma department?
I hate to single it out to charisma, but I think they will need to be an increasingly distinct element to the value add for the business, right? And I think your parallel to what's happening in the software world and the coding engineer world is similar, right? You need big thinkers, you need strategic thinkers, you need people who can help design strategy and not just do the execution piece. And so I do think that there's a possibility that the laboring oars in the most traditional sense, that there ends up being less of those, and unless folks can retool to become, again, more technology fluent or to refine some of those networking and business development and strategic skills, that it'll be possibly another place they end up landing long term.
You've got what, 350, 400 partners around the world? Just over 400. And how many associates?
Okay. So, even with my journalism math, you've got actually a fairly low leverage model for the industry. I guess the question is, law firms are a classic pyramid business. And I guess my instinct is that that pyramid is going to get steeper with a smaller base at the bottom and sort of actually maybe bulge out in the middle a little bit. When you look out and you envision the law firm in 5 or 10 years, like what are those numbers? Is it you have more partners, fewer associates? Is there some middle thing that we haven't thought about that's actually a new job?
Yeah, it's so interesting. And I'll use that as a jumping-off point to another comment I'll make. That part of it is a crystal ball bit, which is I have no idea because I've heard thematically people saying like this is just going to become almost like the audit firms of the world, right? Where there's a few of them, really, really big. Or there's just going to be a bunch more smaller shops that are much more specialized. And again, not because it's easy, but I still see there to be a space in between. I think part of the answer to your question about what the workforce looks like and how the pyramid is shaped is going to be a little bit dictated by, and we've spent some time talking with schools about this, how are we training the talent that's making a decision that even in this day and age, I want to still go be a lawyer. Whatever that means.
If you were designing a law firm from scratch today, and let's set aside the ownership rules about who can and can't share in a legal fee, because I actually think that is helping the industry for now, but can't last forever. How would you design it today? What practice lines would you be in? Pick your five co-founders. What would it look like from a blank piece of paper?
To me, single sheet of paper designing from the beginning, I think that you have to aim to put your firm into the echelon of bet-the-company cases or most impactful transactions where when you're sitting in that boardroom, there's only a couple firms who are possibilities to hire. And that's the standard. And so when you start setting that kind of standard and say, 'Okay, I want to be number one and two in any industry in which we're competing, whether that's M&A, whether that's commercial litigation, go down the list.' That's the standard. And I don't think you can actually be everything to everyone. And so I think that goes back to recruiting the talent that enables you to have that brand profile and frankly to have that expertise. And when I say expertise, which is the ability to execute. It's wonderful to have a brand, but if you can't back it up with outcome, that doesn't really do much for you. And then, again, I think acknowledging that if you do want to hold on to some of the stuff that may not be that highest stakes work, that most elite work, that you are going to have to embrace technology. Because the thing that I think people forget is that that early work, the stuff that may not seem as consequential as some of the biggest ticket work that's being done, that is where the relationships are forged, that is where the trust is built, that is where people make those repeat buy decisions. And so I don't think you can let go of that work, but I think it would have to be highly technology leveraged today.
You do a lot of fundraisings for startups because A, it's a decent business, but also because some of them are going to turn into the next school and you're going to take them public and it's going to be huge.
Like going forward, right?
Is there a world in which there is a sort of lower tier that is very transparently AI produced and then you can pay more to talk to a human, essentially? Is that where this stuff is headed?
Maybe. Maybe or accepting that — I mean, I think probably in the ideal world, it is a firm-branded interface, right? You can imagine the portal at which you come and talk to the bot that is attached to the firm. Like that's probably the ideal world just because again, there's like a brand recognition and a connection there. And then there's also the reality that like some of that work is just going to be outsourced to maybe a native AI firm or another smaller firm that just doesn't have the interest in competing at that high end that I was mentioning. And I think that both of those things or some combination of them is probably going to be true in the near and medium term.
Time to take a quick break and we'll be back with more from Rachel.
Welcome to What's Working, the show within a show featuring Compound and our season partner Amazon Business. I'm your host, Semafor Chief Revenue Officer Rachel Oppenheim, joined as always by Doug Gray. Doug, I feel like you must have a very unique vantage point as Amazon Business's VP of Technology. Looking ahead to the next era of business buying, what do you think is going to separate the companies that pull ahead? Is it lower cost, faster speeds, better data, or is it something else completely?
Well, when I think about it, I think it starts with a simple truth. You don't get the world you want, you get the world as it is. And our job is to get from the world as it is to the one we want. Constantly reassessing and modifying our processes because the things that worked yesterday have to change for tomorrow. And that's what we're building for our customers at Amazon Business. We're helping customers adapt fast without adding risk. Priorities flip overnight. Maybe you planned ahead, but suddenly you need 20 pallets tomorrow and you're willing to pay more to get them fast. Our AI and automation can react with you, making those pivots easier. And we meet customers where they are with a foundation that integrates to their existing systems while giving buyers flexibility and procurement the controls and visibility they need.
