Leading your law firm into the Gen AI Era
Lawyers are embracing its promise. Clients want to reap its rewards. Here are three ways your firm can respond to the immense disruption and extraordinary opportunity of Generative AI.
Illustration by Midjourney
If a week is a long time in politics, then two weeks was a very long time for me to have taken a vacation, with so many new developments breaking out throughout the legal sector. There’ll be time later next month to talk about what’s happening in legal education, lawyer regulation, and other evolving areas. But right now, everything has to take a back seat to the incredibly rapid ascendance of Generative AI in the law.
“We’ve gone from zero to one million miles an hour in the space of six months” is an apt summary offered by a BigLaw partner last week. Generative AI, notes Alex Su, is “diffusing across law firms much more quickly than expected.” In-house lawyers are feeling real-time pressure to grasp the scope and scale of Gen AI’s risks, because corporate leaders want to use Gen AI today and the legal function has to find secure ways to make that happen or be sidelined altogether.
More surprising to me, however, has been law firms’ willingness — dare I say enthusiasm? — to develop and integrate Gen AI tools, with many firms in the US, the UK, and Canada all pressing forward. Lawyers’ typical response to a novel and disruptive technology is to downplay it, ignore it, or ban it. But even those firms that are forbidding their lawyers from using Gen AI are mostly doing so as a temporary measure until they can get their strategies, data, and analytics in order. Many lawyers seem genuinely excited about the potential of Generative AI. That’s new, and it’s important.
But nothing caught my attention like last week’s three-part series of posts about Generative AI at 3 Geeks and a Law Blog. Ryan McClead did a fine job in the opener of demolishing that facile Goldman Sachs “survey” that alleged 44% of legal work/jobs/tasks would be replaced by Gen AI. Posts 2 and 3, however, should make everyone in the legal sector sit up very straight.
Using anonymized benchmark data from law firm time entries supplied by LexisNexis CounselLink, Toby Brown and Greg Lambert calculated that about 20% of legal “work tasks” could potentially be automated by Generative AI, resulting in “a revenue reduction of about 23.5% to law firms in the very near future.” Reasoning that associate work is more vulnerable to Gen AI than partner work, Toby and Greg went on to estimate that in a conservative scenario (partner billable hours down 5%, non-partner hours by 20%), revenue on a sample commercial matter would fall by 13% and profit by 11% — but in a more dramatic scenario (partner hours down 20%, non-partner hours by 40%), revenue plummets by 30% and profits by 28%.
Now, these results are affected by numerous (reasonable) assumptions around leverage and realization rates, among other things — read the articles for yourself to assess the authors’ methodology and conclusions. But Ryan, Toby, and Greg are serious people with solid track records. And even if you account for variations and allow for some discounting, these numbers are extraordinary, and alarming. They suggest a challenge beyond any the law firm business model has yet faced.
Toby and Greg believe that their conservative 5/20 scenario “will be substantially eclipsed in the long-term, but [is] quite possible, even likely in the next 12-24 months.” That squares with what I’ve been telling law firms, which is that 10% to 15% of the tasks they’re currently billing are already susceptible to Gen AI, and that this technology, still in its infancy, is growing more powerful every day. So I think we can start making some calls about where all this is headed.
Over the next few years, the amount of time and effort required to perform basic and even intermediate legal tasks is going to crash. This will decimate law firm inventory, the vast majority of which is hours billed by lawyers and especially, as Toby and Greg point out, by leveraged associates. Most law firms would struggle with a 10% decrease in revenue and profits; no firm could survive drops of 30% or more. The only thing holding the typical law firm together is the collective confidence of its partners that the firm is strong and its future is bright. If that confidence disappears, so too does the firm — often with astonishing speed. I’ve seen it happen.
Like it or not, the legal sector’s Generative AI Era has begun. Law firms need to recognize this fact, absorb its implications, and deal with its ramifications and opportunities as quickly and decisively as they can.
As the foregoing links demonstrate, a pleasantly surprising number of firms have gotten off to a fast start. These firms are encouraging their lawyers to experiment with Gen AI, brainstorm use cases with clients, and think about ways in which they can use this technology to create new and greater value. All these activities are sensible and useful. But I feel like they’re also table stakes, the minimum required to prepare for the transition coming our way.
At a higher level, law firm leaders must understand the scale of the strategic vulnerability Gen AI has created. They have to take steps not just to minimize that vulnerability, but also to begin turning this development into a unprecedented opportunity to build competitive advantage. Three fundamental aspects of law firm structure and culture lie at the core of this vulnerability/opportunity. Firms need to address each of them, roughly in the following order.
Move fast to implement project and client pricing. If you keep on selling hours, sooner or later you’ll go out of business, because billable hours will continuously decline until they reach a baseline well below your level of survivability. Start now to develop pricing structures based on individual matters (an insolvency, an injury claim, a trademark defence, a shareholder dispute) — or, for particularly reliable and high-volume clients, a structure that covers all work of a certain type on a quarterly or annual basis, in each case with options, off-ramps, and trapdoors triggered by previously specified elements or occurrences. Pricing consultants (I’m not one) can get you started down this road.
Prepare to hire fewer associates and to rethink partnership. A 20% to 40% decrease in the number of non-partner billable hours wouldn’t automatically translate to 20-40% fewer non-partners. I’ve written previously that some associates’ time will get “levelled up” to higher-value engagements. But barring a true AI-driven lawyer-value renaissance, you should count on at least 15% to 20% fewer hours for associates in a couple of years’ time, and hire (or don’t hire) accordingly. As Gen AI changes our understanding of leverage, you should also re-examine what “associates” and “partners” even mean anymore, and get ready to restructure your workforce into “core” and “ancillary” categories.
Establish a fresh approach to developing future law firm leaders. Law firms have long assigned new lawyers low-level tasks, not just to generate revenue but also to train them in the structure and operations of deals and cases. Many of those tasks will soon be automated, by Gen AI or otherwise. New lawyers also learned the technical and cultural ropes by immersion in the daily life of the law office, and that’s going by the wayside too. You’ll need a new approach to developing the firm’s future workhorses, stars, and leaders — starting with a recognition that early-stage lawyers on the leadership track will no longer be revenue generators the way they once were.
I’ll explore each of these aspects in greater detail in the next three editions of the newsletter. My hope is to sketch out a map with which law firms can set out to manage each of these very challenging transitions. But to be very clear, I don’t have all the answers here — nobody does. We’re all figuring this out as we go, sailing into an unexpected storm in uncharted waters, and every ship will have to plot its own course.
To that end, I want to leave you with one request: If your firm hasn’t yet formed an AI Action Group to grapple with these issues, please do so now. And if you do have such a group, ensure that it’s multi-disciplinary (not just lawyers), appoint key clients as advisors, and give its members the bandwidth and resources they need to assess options, choose strategies, and deliver actionable advice to your leadership team. But also — and this really matters — give them permission to be positive!
I’ve framed this Gen AI Era as both a vulnerability and an opportunity for law firms. Whereas the first aspect should inform a sense of urgency, the second should encourage real feelings of excitement. These are extraordinary times! It’s an awesome moment to be a lawyer and a golden opportunity to guide a law firm, grappling with the potential of truly world-changing technology. Yes, the challenges ahead are real and substantial. But keeping an open and curious mind, as well as a confident and enthusiastic heart, will help you overcome them. This is going to be an amazing ride.
It's really difficult to predict the future given the pace of change here. Microsoft, Netdocuments and other vendors are embedding AI into their core offerings. So if you are employed to do document review, comparisons or clause identification, then I hope you have interesting hobbies and a big fat bank account. Tech helps you do dumb things, faster. Some lawyers will embrace AI to iterate faster.
Two thoughts on this excellent piece.
1) Since Generative AI has shown its power in many different ways, firms have seen that they need to start investing in it and using it or risk being seen as not as "up to date" as other firms.
2) AI, like any technology created by imperfect humans, is imperfect. It can learn and get better faster than a human though, right the right data, the right amount of time, and the right focus.