A $500 million headline draws attention like a thunderclap. So, when Kirkland & Ellis, the highest-grossing law firm in the world, announced it is building its own AI platform, the legal press called it audacious, a watershed, and the opening salvo in a new arms race. Measured against the firm's own revenue, however, it is a careful, even conservative bet — and that gap holds the real lesson for the rest of us
Key insights:
-
-
-
The headline obscures the math — Kirkland’s roughly $100 million a year is about 1% of its $10.6 billion in annual revenue, a level most research-driven industries would consider a maintenance budget.
-
Build versus buy is the real signal — Choosing to own a proprietary platform rather than license the same tools that competitors can buy reflects a belief about where advantage now accumulates.
-
R&D is a habit before it is a budget — The discipline of continuous reinvestment, not the size of the check, is what compounds investment into greater capability.
-
-
Welcome back to The AI Law Professor. Last month, I examined the jagged fit problem: Why the same AI tool produces uneven results from one lawyer to the next, and why matching capability to task matters more than the brand on the box. This month, I want to widen the lens from the individual lawyer to the law firm itself and ask what it actually means for a firm to invest in its own future.
On May 28, Kirkland & Ellis claimed it will spend $500 million over the next three to four years building its own AI platform, starting with a roughly $100 million investment this year. The firm will fund the work from its annual revenue, which reached about $10.6 billion last year.
Outside technology companies are helping to build the system, but they will not be permitted to resell it to competitors. Some 250 Kirkland lawyers, including 100 partners, have already contributed detailed accounts of how they work so the platform can be tuned to the firm’s own methods. The ambition is end-to-end. They’re creating a system that can carry a complex mandate from initial scoping through execution, rather than a scattered collection of point tools for document review, due diligence, and drafting.
A half-billion dollars, in context
Obviously, $500 million dollars is an enormous sum in absolute terms — but, as a share of Kirkland’s revenue, it is far more modest. The roughly $100 million the firm expects to spend this year sits close to 1% of annual revenue, a level that firm chair Jon Ballis has framed as the firm’s appetite for taking “big swings.”
Place that 1% against how other industries fund their own future. Software and internet companies reinvest an average of about 13% of revenue into research and development. The United States pharmaceutical industry routinely spends north of 20%. Defense contractors land somewhere between 10% and 15%; consumer electronics makers, 8% to 12%. By those benchmarks, a 1% commitment reads less like a moonshot and more like minimal maintenance.
For most of its history, the legal industry has carried no R&D line item at all. This means that there are two things that are true at once: Kirkland is leading the profession, and the profession as a whole still invests a fraction of what the most innovative industries treat as the cost of staying alive.
The deeper signal is ownership
The dollar figure is the headline, but the strategy underneath it matters more. Kirkland concluded that if every firm can license the same AI from the same vendors, that AI stops conferring any advantage at all. So, the firm chose to own its platform rather than rent it, and to bar any outside builders from selling the result to rivals.
This instinct is not new for Kirkland. In 2017, the firm built CTRAN (Corporate Transactions Database), a proprietary database of past M&A transactions that let its lawyers spot patterns in deal terms that its competitors could not see. That data advantage proved difficult to replicate, and it helped carry the firm to the top of the global revenue tables. The firm’s planned AI platform is the same instinct on a vastly greater scale: Treat institutional knowledge as an asset to be compounded, not a byproduct to be discarded.
That is the signal worth absorbing. The strategic question is no longer only which tool to license; rather, it’s what you are building that a competitor cannot simply purchase for itself.
Three ways to invest in R&D without a big budget
Certainly, most law firms do not have $500 million, or even $5 million, to commit. Yet, they may not need it. R&D is a discipline before it is a budget, and the discipline can be scaled down.
First, make R&D a standing commitment rather than an occasional impulse. Set aside a fixed share of revenue, even 1% to 3%, and a fixed block of protected, non-billable time each month for experimentation. Name someone to own it and stick to it.
Second, turn what you already own into a proprietary knowledge asset. You do not need to train a model to draw an edge from your own data. Your closed matters, briefs, clause libraries, and playbooks can be organized into a structured, searchable knowledge base, then connected to a retrieval system. This gives you the small-firm version of CTRAN, which can compound with every matter you handle.
Third, run experiments with a clear measure of success. Pick one workflow, define a baseline for time, cost, or error rate, pilot a tool against it for a fixed period, then decide deliberately whether to keep it or kill it. And write down what you learn.
The strategic habit is the asset
The temptation when reading about Kirkland’s half-billion-dollar bet is to conclude that R&D belongs only to firms with billions to spend. Actually, the opposite is closer to the truth. Kirkland’s real advantage is not the size of its check; rather, it’s the decision to treat building as a permanent part of its long-term strategy and how it operates. That decision is available to a solo practitioner with a free weekend and a stack of old briefs just as surely as it is to the largest law firm in the world.
The half-billion-dollar figure will make the headlines, but it’s the strategic habit that compounds in value.
Tom Martin is CEO & Founder of LawDroid, Adjunct Professor at Suffolk University Law School, and author of the forthcoming AI with Purpose: A Strategic Blueprint for Legal Transformation (Globe Law and Business), where he shares exactly how you can build your own strategic habits and assets for your own law firm.
