The AI premium now belongs to lawyers who can prove control
The market is rewarding firms that can show they govern AI, not merely use it—and the courts are rewarding the opposite with fines, referrals, and embarrassment.
An AI-assisted editorial, reviewed by a human before publishing. It reasons over our own tracker data (and cited context) — a point of view, not legal advice.
Here is the working-lawyer version of the AI story: the upside is real, but the premium now goes to lawyers who can show control. Not enthusiasm. Control. The firms making the loudest moves are not buying novelty; they are building systems, training people, and putting lawyers back in the loop where liability actually lives.
Kirkland & Ellis has earmarked $500 million of its revenues to develop its own artificial intelligence platform. Cleary Gottlieb bought legal-tech firm Springbok AI and brought a team of AI engineers in-house. Tarter Krinsky & Drogin launched a firmwide Office of AI and Innovation to embed responsible AI, technology, and process innovation across operations. That is not a hobbyist phase. That is a profession deciding that AI belongs inside the firm, not just on the browser bar.
If you cannot explain how the output was checked, you are not using AI; you are renting trouble.
The market is already writing the memo
The strongest signal is not what vendors promise. It is what large legal departments are willing to buy. Microsoft’s Corporate, External, and Legal Affairs organization, about 2,000 lawyers and compliance professionals, will use Harvey for legal and compliance work. ADP named Helena Almeida chief AI legal officer on July 1 and expanded her role to product compliance and responsible AI, while she emphasized that CFOs want human in the loop to mean something real, not a box checked for morale.
That is where the pressure lands on practicing lawyers. Clients are not asking whether AI exists. They are asking whether the people billing them can use it without turning judgment into guesswork. Gartner says legal departments need to reinvent talent, data, and workflow strategies in response to AI, flat budgets, and rising workloads. It also says companies may double legal technology budgets by 2028. In plain English: the client-side bar is getting more selective, not more sentimental.
Courts are setting the price of sloppiness
If the business case sounds abstract, the sanctions docket does not. In the Coomer v. Lindell matter, defense counsel filed a brief with nearly 30 defective citations and cases that do not exist, then was sanctioned after admitting unverified AI use. In Ader v. JS Property Holdings LLC, Michael Fourte filed briefs with AI-hallucinated citations and quotations, and Justice Joel M. Cohen awarded fees and referred the matter to ethics committees. In a New York City matter involving Innocent Chinweze, a lawyer filed a brief riddled with fake cases after using Microsoft Copilot, and the judge fined him and referred him for discipline.
And the courts are not reserving embarrassment for high drama. In a California custody fight over a dog, counsel filed briefs with fabricated cases and was fined $5,000. Oregon declined formal sanctions in a Buchalter matter after AI-created fake case citations appeared in a filing, which is almost worse in its own way: even when a judge shows mercy, the file still bears your name. The lesson is not mystical. If AI writes your brief, you own the hallucinations, the correction, and the bill.
This is the new competence test
The profession likes to talk about AI as if it were a single decision: adopt or do not. That is too crude. The real question is whether a lawyer can use these tools without surrendering responsibility. The tools themselves are increasingly capable. CoCounsel Legal, Harvey, Lex Machina, and RelativityOne are each scored 90 out of 100 in the catalog. Lexis+ with Protégé and Gavel are close behind at 89. The point is not that every lawyer must become a fan. The point is that the market has already moved past “is this real?”
What matters now is whether the lawyer can distinguish between speed and proof. A platform can help with research, analysis, drafting, document review, litigation intelligence, or e-discovery. Fine. But a brief full of fake cases does not become less fake because the draft looked efficient on a screen. The lawyers who will thrive are the ones who can use AI to compress routine work and still verify the record, the citations, and the judgment. The ones who cannot will discover that the cheapest part of AI is the software.
The bottom line for practicing lawyers
The AI shift is not a referendum on whether lawyers are obsolete. It is a referendum on whether they can still be trusted to supervise their own work. Firms are committing serious money. Clients are reorganizing around AI. Courts are imposing fines, fee awards, and referrals when lawyers let the machine write checks their name cannot cash.
So the opportunity is straightforward: use AI to do more useful work, faster, and reserve lawyer time for the parts that actually require a lawyer. The risk is equally straightforward: if you cannot explain how the output was checked, you are not using AI; you are renting trouble. That is not innovation. That is a malpractice fact pattern with a nicer interface.