By their own best lawyers' account, most law firms already have an AI strategy. What they are short on is the unglamorous part: changing how the work gets staffed, priced and sold so the strategy earns something. That gap is predictable, because the lawyers who shape those firms will tell you which leadership skills they value, and the one that closes it sits near the bottom of their list.
Ask the lawyers clients single out as their stand-outs what matters most in a firm's leaders, and strategic vision wins by a wide margin, according to the Thomson Reuters Institute's 2026 survey of client-nominated stand-out lawyers. Change management and technology leadership rank near the bottom. In the same survey, more than three-quarters say their firm has an AI strategy, and only a quarter strongly agree it has a plan to make money from that AI use.
Read those two findings together. A firm that prizes vision and rates change management low will write a strategy and then wait for it to happen on its own.
An AI strategy stays a slide until someone changes the work
A strategy says where the firm wants to go. Making it pay means somebody decides what is different on a Tuesday: which work gets done another way, who does it, how it is priced, and what the client is told. That is change management, and it is one of the skills these leaders rank near the bottom.
My standing view is that projects fail on ownership, not technology. I spent two years building and deploying AI inside real businesses, and I said it plainly in a working session this summer:
We know how to do almost anything now. … We just underestimated how hard it was to get people out of their bad habits or what they're comfortable with.
Knowing a tool exists does not mean anyone uses it. As I put it in the same session, that's like saying someone who knows how to use a calculator always uses Excel, and plenty of people refuse to use Excel.
The payoff shows up in pricing and staffing, not in the tool
If AI lets a team finish the same work in less time, someone has to decide who keeps the time saved. The firm can bill less for the same matter, take on more matters with the same people, or move people to work clients pay more for. The question is already on the table. The report notes that rates at the Am Law 100, the largest U.S. firms by revenue, have hit year-over-year highs, and that clients are asking for lower fees.
Each answer changes how juniors are trained, how partners are paid and what the client conversation sounds like. None of those is a technology decision, and none of them gets made by the person who bought the software. A firm that treats AI only as a way to get the same work done faster can miss every one of those decisions.
The pattern is not special to law. Accounting firms, agencies and consultancies run on the same model of experts selling their time, and each of them faces the same decision about who keeps the time AI saves on billed work.
Give the change an owner the partners will follow
If I were sitting with the leadership of an expert-led firm, I would start here:
- Name one person who owns making the AI strategy pay. It has to be someone the partners listen to. A technology hire off to the side will run pilots that never reach the pricing sheet.
- Pick the commercial number that tells you it worked before you buy anything else, such as margin per matter or turnaround time on a type of work clients already pay for.
- Bring the people who do the work into deciding how it changes. AI cannot be a separate journey that one team takes while everyone else watches.
- Decide what you will tell clients, and tell them early. A client who hears about your AI from you is easier to price with than one who reads about it and asks for a discount.
The vision your partners prize is still the starting point. What they need beside them is a person who turns it into a map with dates and owners, then walks it with them and the whole firm.
If your firm has not put AI into client work yet, start with one workflow
Firms still deciding whether to use AI at all do not need a change program. They need one workflow, one owner and one measure, run long enough to see whether the work gets better and whether anyone keeps using it. Build the change plan once you have something worth changing for.
Ask your partners one question this month
Ask each partner, separately: what will we charge, staff or sell differently a year from now because of AI? If the answers match, you have a plan. If they do not, you have a strategy and no owner, which is the more common place to be.
I have written about why AI pays when you score it on growth and about getting a decision through managers to the work. You can see how I work with leadership teams on AI and growth. If your firm has the strategy and is waiting on the payoff, I am glad to talk it through.
John Wyatt · John Wyatt's page · All posts · RSS