Legal innovation and generative AI: Lawyers emerging as ‘pilots,’ content creators, and legal designers

I’ve been having a lot of conversations with my lawyer friends and colleagues lately over the “rise of the machines”—how generative AI is slated to change, shift, and (gasp) make our line of work obsolete, and generally what the emergence of all this new tech means for the value we bring as lawyers individually and as a profession. As a practicing technology lawyer and as the head of innovation for McKinsey Legal, these questions are both fascinating and tactical for me. In recent months, I’ve been closely reading up on this quickly evolving space and taking some time to form my own perspective.

Generative AI and our opportunity for automation

“The secret of change is to focus all of your energy not on fighting the old, but building on the new.” – Socrates

With gen AI, so much of the lawyering that we’re used to doing manually will be easily automated and readily available for our adoption. For many, that’s a scary thought and leads us to wonder if it makes what we do less important or relevant. My perspective is the exact opposite: I see this as a tremendously exciting time in our profession, because the best use of this newly emerging automation stack can free up real time that lawyers use on heavily time-consuming work with lower payoff. It releases that time to let us do more creative, tailored, and innovative work that will allow us to ramp up our expertise and craft unique value.

Some examples: instead of searching for and aligning on defined terms, we can have gen AI do it, allowing us to work on simplifying the terms themselves and making them clearer. Instead of searching for applicable precedent, we can have gen AI handle it and suggest novel legal theories. We can then draw from those theories, build on them, and come up with an even better argument. Instead of taking a first pass at the issues list, gen AI can do it, and we can inject novelty and creative structuring into the business terms to hit that strategic level up for our client.

As lawyers, we’re constantly editing by using form documents and precedents. But we spend too much time, by necessity, on tailoring the forms. Gen AI could do that for us, taking the first cut, which enables us to be exceptional, thoughtful, engaged, and highly expert editors—letting us put our thinking caps on and have the time to dive into a solid, tailored first draft and position us to make it that much better from there.

Our runway and responsibility with respect to legal tech

With this in mind, I’ve developed what I’ll call my pilot theory in describing the critical posture and expertise of the next-gen lawyer. In a world where many gen AI tools are popping up and serving as copilots (such as Casetext and Harvey, plus, no doubt, others coming to the fore shortly), the real question for today’s and tomorrow’s lawyers is how we’ll play the role of the thoughtful pilot: sitting in the cockpit, using these tools, having our copilot next to us at all times, doing our best to oversee their work and judgment, and figuring out how to best collaborate with them so that we reach an optimal result every step of the way. Among many emerging questions is, how do we avoid automation bias and protect ourselves from deferring to machines, assuming that they’re doing it right?

The shift needed in training and apprenticing for next gen lawyering

“If the only tool you have is a hammer, you tend to see every problem as a nail.” – Abraham Maslow

Traditional lawyering is rooted in an apprenticeship model and Socratic reasoning. In law schools, we’re taught by our professors to derive answers and spot issues by being asked questions in a nuanced and pointed way. Using inductive and deductive reasoning, we learn to read a case or take in a problem or pattern of events and identify the constellation of embedded legal issues. In practice, junior lawyers learn from senior associates and partners by seeing how they work, working closely alongside them and reviewing their redlines with comments, and they learn how to do document review from the senior associates’ instructions.

Gen AI is changing all of that, posing two important questions: first, how will we apprentice machines and train them to be effective copilots for us in the work that we’re doing? What kinds of lessons will we teach them, what kind of data will we feed them, and what kind of instructions will we give them (the important point of prompt engineering, discussed in more detail below)? Additionally, how will we, as teachers and senior lawyers, manage the machines we work with and do our part to be better at our own practice—having the right level of oversight, reaching a balance of trust versus management (and avoiding micromanagement, as so many senior lawyers find tough to do), and using the work product given to us by the junior lawyer (via machine; that is, the copilot) to meaningfully improve the way that we deliver truly distinctive services?

The learning and level up opportunity

“I skate to where the puck is going, not where it has been.” – Wayne Gretzky

In my view, the question we all need to consider is, what does the third-, fourth-, and fifth-year associate look like in 2024, 2025, 2030? What’s their workflow, expertise level, and level of engagement with clients and the real business strategy behind the law? In other words, how do we measure their ability to put it all together and be more of a holistic adviser and true business partner when practicing, rather than focusing on more discrete tasks? With legal tech as the copilot, we have an opportunity for automation and acceleration in the way we work and a real opportunity to reexamine the training, professional development, and real efficiencies across our profession.

For the junior lawyers coming up the ranks, the advent of gen AI allows for automating repetitive, non-substantive tasks and leaves room for substantive learning sooner, empowering junior lawyers to dig into the substance of the law early on and put it all in context. They can learn by application and design, rather than learning nuggets of the trade in bite-size, fragmented pieces and then doing the hard work of putting it all together only at the more senior levels of practice. The value we bring as lawyers is captured in our expertise, judgment, and ability to advise clients in a thoughtful, creative, nuanced, and easily accessible way. These are challenging, multidisciplinary, and demanding skills to learn, and our profession spends too much time at the junior-associate level doing repetitive tasks rather than learning these highly challenging and formative skills in a way that sets each of them up for success. By effectively using increased automation, we can open up the time, space, and aperture on what it means to be an exceptional junior associate and developing lawyer, and from there build upon the wider expertise and set of skills that each growing lawyer has, as they rise up through the ranks.

For more senior and seasoned lawyers, the legal tech copilot can be a true creative partner—maybe for a really tough and novel case, it examines a set of precedential decisions and comes up with several legal theories. The learned lawyer can use that initial analysis to then level up and string together a perspective and theory that’s significantly more novel, because the first cut of the creative process was already shared—like working with your really smart, thoughtful, and insightful law school study group buddy and kicking around thoughtful and outlandish ideas, getting closer to that genius point.

If we get it right, the impact on our profession can be enormous and our own growth potential as lawyers, individually and collectively, can skyrocket. We’re at a point in time where we’re shifting from the abacus to the voice-activated calculator and it’s a beautiful thing. The question is whether we have the vision, aptitude for adoption, innovation readiness, and change management strategy to get there thoughtfully, responsibly, and efficiently from where we are today.

The power of lawyers as creators and designers

“There’s a way to do it better. Find it.” – Thomas Edison

But that’s not all. Legal tech as a copilot is only a piece of the legal-innovation puzzle. The elemental and crucial shift for us as lawyers—especially amid our own tech revolution—is to position ourselves and see ourselves as content creators and legal designers who are constantly generating content (because we are, down to the written words in our emails), but doing it in novel, crisp, and engaging ways. As lawyers, we traditionally see ourselves more as technical doers and processors—we apply legal reasoning to problem statements and give advice to clients.

But in 2023 and beyond, our profession is demanding so much more from us. We ought to be proactively designing strategies to partner with our business teams rather than simply responding to discrete questions, and the way in which we respond ought to be engaging, innovative, and easily accessible (for example, using interactive data visualization to answer a question, instead of a dry and verbose email). Let’s design, create, and set ourselves apart. The operating system of the legal profession has been a set of outdated terms for far too long, and we’re tech savvy enough to understand that each operating system benefits from the occasional restart and refresh.

Like the emerging adage that every company is now in some way a tech company because of the prevalence of technology, all lawyers are now in some way asked to be tech lawyers. But tech law is new, quickly evolving, and largely interpretive—there’s not a treatise for it in the way that we have a treatise for property law or civil procedure. So much of it, too, is simply emerging and around the corner—and if there’s one constant in technology law, it’s that it always trails the development of the technology itself. And so tech lawyers, by default, are a cross-disciplinary compilation of many things: we draw from litigation trends; ground ourselves in ethics; work closely with our tech experts, data scientists, and engineers; and forecast trends. We look at industry norms and emerging trends, form a perspective, and thoughtfully string guidance together. The best tech lawyers—and more broadly lawyers, really—are the ones who can see the bigger picture, frame the problem statement in the right way, and put it all in context. That’s a pretty time consuming and tough gig, and legal tech as a copilot can help us get there, freeing up time from menial tasks to instead keep us reading about the direction of the law, collaboratively engaging in thought leadership and discussion, working directly with our clients and business colleagues, and figuring out the best steps to getting toward a more ethical, thoughtful approach to how we practice. Essentially broadening the application of the law school study group or just going down the hall to ideate with your really smart associate friend—we can now do that in broader circles, because we might just have more time.

The importance of design thinking as a core legal tool and skillset

“Innovation is taking two things that exist and putting them together in a new way.” – Tom Freston (cofounder of MTV)

But as we figure all of that out, we should also keep in mind the importance of equipping ourselves with the right tool kit. Traditional issue spotting and inductive reasoning isn’t enough anymore for us to do our best work, and we need a reexamination of the legal tooling we’ve been trained on. We ought to figure out a succinct way to frame a problem statement and anchor it in humanity; to put the client at the center of what we’re doing and frame the problem statement from a place of empathy, rather than a solution orientation anchored in getting toward the answer that’s technically optimal or “correct”—because that’s not the answer that may resonate best with the client, and that means we may have framed the problem statement all wrong.

Human-centered design thinking is a way of approaching problems that focuses on the person and teaches us to frame the problem statement in a way that tailors the question with the right mix of empathy and specificity to get the best answers to that specific question. When it comes to piloting and legal tech as copilots, so much of how we interact with gen AI tooling will be rooted in prompt engineering and prompt design—this is, asking the tech the question in the right way, so that you get the best answer to resolve the problem statement that you’re examining.

Application of design thinking in prompt engineering will become (and already is) a critical way of designing the right question to get a better answer. Prompt engineers and designers are in high demand now across industries, and the ones who seem to do it best are the folks with human-centered design expertise, who know how to frame a question with the right level of empathy, tailored architecture, and clarity. The design decisions are based in their craftsmanship and training and lead to better outputs. As lawyers increasingly interacting with gen AI, all of us need to develop the skills of becoming truly skilled and expert prompt designers, so that we can work with our copilots most effectively. And approaching that from a design-thinking lens, as legal designers who are looking to create innovative content with gen AI as creative partner, may just be the way to get there.

One of my favorite quotes is from Maya Angelou: “You can’t use up creativity. The more you use, the more you have.” And I truly find that to be the cornerstone of the legal profession. We’re not becoming obsolete or getting replaced; we’re just now given the opportunity and the time to be more creative, and with that we can be so much better at the profession we initially stepped into.