Kirkland Signals Meaningful but Not Transformational AI Future

Kirkland & Ellis’ decision to invest half a billion dollars in artificial intelligence is less interesting than its decision to announce that intention. The announcement is important, as it was intended to be. Kirkland is, by many measures, the world’s most successful law firm. When it makes a public announcement of this magnitude, it is sending a signal: Visible commitment to AI matters.

There is much to unpack in what Kirkland is signaling—to the market, to clients, to its own lawyers. But the question that came my way first, and then frequently, is what the announcement does, and doesn’t, mean for clients asking, “Where is my AI discount?”

Several firms I confer with have already received client emails linking to the announcement and expressing some variant of: “Please tell us how the efficiencies these technologies deliver are translating into reduced matter costs.” Present tense.

Clients may well have valid reasons to think they should be able to spend less with their law firms, but the Kirkland announcement does little to bolster their argument.

Money matters. But the headline $500 million figure isn’t likely to translate to a fundamental change in the Kirkland business model—or in the Big Law model generally—that will lead to immediate discounts for clients.

Part of the reason for this is that half a billion dollars sounds so enormous that it induces a sort of denominator blindness: $500 million with a denominator of what?

In 2025, Kirkland became the first law firm to surpass $10 billion in annual revenue, up more than 20%, or $1.75 billion, year-over-year. That means the $100 million it revealed it will spend this year—and maintain over the announcement’s four-year horizon—represents less than 1% of revenue.

For reference, companies plan to spend about 1.7% of revenue on AI this year, roughly double 2025 levels, Boston Consulting Group’s January 2026 AI Radar found. Viewed against that benchmark, Kirkland’s announcement appears less dramatic. The firm is committing capital at roughly the level corporate America was investing a year ago and about half the level its corporate clients now expect to invest.

Viewed another way: If large firms spend 5% to 7% of revenue on technology, the announcement implies Kirkland intends to raise total technology spending by, at most, 1% of revenue, while directing less than 20% of its existing technology budget toward AI. The investment is meaningful but not transformational relative to the enterprise making it. This is simply not evidence that the economic model is being rapidly reinvented.

Turning back to the client’s query about when they will see their AI discount, I am sympathetic. But having reviewed responses from more than 180 firms in connection with the intelligence platform I operate, I’m also keenly aware that cheaper continues to crowd out better and faster, let alone different, in the client/firm AI dialogue.

Our data shows law firms perceive clients to be almost twice as interested in savings as in improved quality or speed. More tellingly, when asked to rate 12 distinct threats to their business model—homogenization, client insourcing, client willingness to move work—only one registers as genuinely material for a majority of firms: clients demanding to capture AI-enabled savings.

This focus is both comprehensible and concerning. Comprehensible because the client logic seems facially sound: We pay for time; AI reduces time; therefore, we should pay less. Concerning because over-indexing on savings relegates AI to just another lever in a reductive pricing conversation that hit diminishing returns long ago and could keep tomorrow looking much like yesterday.

Speaking of yesterday, in 2023, the Wall Street Journal ran a new varietal of an evergreen headline: “End of the Billable Hour? Law Firms Get On Board with Artificial Intelligence.” In that article, Kirkland allowed itself to be named as part of the beta group testing legal technology built atop the then-state-of-the-art GPT-4.

Three years later, Kirkland’s big AI announcement remains forward-looking. It tells us about Kirkland’s intentions, not its results—despite the firm being among the earliest adopters and the one with the deepest pockets. That Kirkland is still talking about tomorrow rather than today tells us how early today still is.

AI is too often regarded as free productivity. It isn’t. It‘s a capital allocation decision. If the costs of implementation, integration, governance, and verification exceed the value created, adoption does not pencil—until it does, when economic pressure leaves firms no choice.

Kirkland’s announcement suggests we are in the no-choice era—but one that is more incremental than exponential. Deliberate beats indifferent. But fast doesn’t necessarily beat slow.

Slow is steady. And law remains very much a credence good; maintaining credibility is paramount. Client exhortations to deliver AI’s benefits are invariably paired with admonitions to avoid its pitfalls—a tension far easier to articulate than to resolve.

A rational case can still be made that the reputational risks of moving too fast outweigh the commercial penalties of moving too slow, especially because the fixation on discounts is double-edged.

If a firm isn’t using AI, clients believe they should pay less because the firm should be using AI. If a firm is using AI, clients believe they should pay less because the firm is using AI. All of that is to say clients believe, as they long have, that they should pay firms less—period.

The attraction of the AI discount is undeniable. It is so very simple—everything stays the same, we just pay less. But there is very little automatic about enterprise AI; it doesn’t just happen, it must be done. In the near term, it usually means more work, not less.

This inconvenient reality extends to law departments seeking to capture AI’s upside. Fundamentally changing the economics of law firm relationships will take real work. Simply demanding that firms immediately pass through savings skips over the part where the savings must first be created—and doing so must be properly incentivized.

Incentives have shifted to the point where the world’s first $10 billion law firm believes it necessary to invest heavily and signal seriousness about AI to win more business. Kirkland is betting there‘s more work to be won. More, not less.

The Kirkland announcement suggests things have changed, just not all that much. It seems predicated on tomorrow looking much like today, which isn’t all that different from yesterday.

An immaterial amount of this content was drafted by generative artificial intelligence.

This article does not necessarily reflect the opinion of Bloomberg Industry Group, Inc., the publisher of Bloomberg Law, Bloomberg Tax, and Bloomberg Government, or its owners.

Author Information

Casey Flaherty is partner of Baretz + Brunelle, a growth advisory firm to the legal industry, and co-founder of LexFusion, a legal tech accelerator. Previously, flaherty served as director of legal project management at Baker McKenzie. He also established the legal operations function within a Fortune 5 company.

Interested in writing? Review our author guidelines, and submit pitches to Insights@bloombergindustry.com.