What does the client-side data actually say?
That a meaningful minority of corporate clients are already treating AI capability as a selection criterion. Thomson Reuters’ Future of Professionals report of 22 June 2026 found 32% of corporate clients have reconsidered or plan to reconsider firms that lag on AI.
The same report found 74% of professionals using AI multiple times weekly and 44% daily, while 91% said their organisation falls short of AI’s potential value. Where a named strategy existed, 66% reported realising value against 22% without one — which suggests the differentiator is organisational rather than technological.
Why is the gap between the two numbers the real story?
Because it is a mispricing of risk, and mispricings are actionable. Roughly 32% of in-house professionals are reconsidering firms over AI while roughly 11% of lawyers believe they could lose clients over it. Three times as many clients are contemplating the move as there are lawyers who think the move is possible.
A firm does not need to believe AI transforms legal work to act on that. It only needs to believe its clients are having a conversation it is not part of.
What are clients asking for that they say they are not getting?
Quality improvement, specifically, and the delivery gap is severe: 77% of clients say AI-enabled quality improvements matter to them, and 5% say they actually receive them. That is a 72-point gap between stated demand and reported delivery.
Note what is being asked for. The demand is not for lower prices or faster turnaround in the abstract — it is for better work. A firm positioning AI purely as a cost story is answering a question its clients are not asking.
Are clients uniformly enthusiastic about their lawyers using AI?
No, and firms reading only the pressure data will get this wrong in the other direction. Clio’s 2025 Legal Trends Report found 36% of clients would trust a lawyer less for using AI generally, that 78% want AI use disclosed, and that 35% of lawyers rarely or never disclose it.
A 2026 UK survey of 642 participants found 45% completely or slightly comfortable with AI supporting legal service delivery, 32% uncomfortable and 22% unsure, with comfort dropping to 25% among those aged 75 and over.
The two datasets are not in conflict. Sophisticated corporate buyers are pushing for AI capability while individual clients are more cautious and want to be told. A firm serving both is managing two different conversations.
Does adoption actually convert into revenue?
Unevenly, and firm size is the strongest predictor. At solo and small firms, adoption runs at 71 to 75% while only about 32% saw a revenue increase; enterprise firms convert at around 59%. That gap between adopting and earning is the efficiency paradox in one line.
Measurement is part of why. Only 18% of professional organisations track AI return on investment at all, and 40% do not know whether it is measured. A firm that cannot tell whether the investment paid will keep making the same investment decision on feel.
What can a firm defensibly claim about its AI use?
Less than most marketing departments assume, and the constraint is advertising rules rather than taste. Florida Bar Ethics Opinion 24-1, issued 19 January 2024, is the most operationally specific guidance here: lawyers may advertise their use of generative AI but cannot claim their generative AI is superior to that used by other lawyers or firms unless the claims are objectively verifiable. The same opinion requires informing prospective clients when they are communicating with an AI program rather than a lawyer or firm employee.
Model Rule 7.1 governs every pitch, matter description, directory entry and post regardless of jurisdiction, and conduct rules are adopted state by state and diverge. The defensible claim is a specific description of what the firm does and how it is supervised. The indefensible one is a comparative superiority claim nobody can verify.
The Institute’s Marketing & Business Development area covers positioning within those constraints.