Clients expect AI-enabled work from advisers. What should a professional-services firm publish?

How wide is the gap between what clients expect and what firms deliver?

Wide, on the best recent data. In the Thomson Reuters Institute's Future of Professionals 2026 survey of 1,816 professionals in 62 countries (March to April 2026), 78% of corporate clients said AI-enabled quality improvements from the firms they work with are very important or essential.¹ Just 6% said most or all of their providers deliver them.¹

The same survey puts a commercial edge on the gap. 32% of corporate clients said they had already reconsidered relationships with firms they feel are falling behind, or planned to within 12 months.¹ The survey covers law, tax, audit, accounting, compliance, risk and trade, so it applies well beyond law firms.

Our view: much of that gap is a communication failure as much as a delivery one. Many firms already use AI in research, drafting or review, and say nothing about it in public. A client who cannot see the method assumes there is none, and so does an AI engine asked to compare advisers.

Adoption is high; measured benefit is not

Firms are adopting AI faster than they can show results from it. Deltek's 2026 Clarity study of 896 architecture and engineering firms in the US and Canada (published May 2026; fieldwork not stated) found AI adoption had risen to 70%.² In the same study, 38% reported a measurable positive business impact from AI, and 39% said it was too early to see benefits or that they were still piloting (fieldwork not stated).² Deltek sells project and ERP software to these firms.

That pattern matters for what you publish. A claim such as "we use AI across our practice" is easy to write and hard to believe. A page that says which tasks use AI, what a person checks and what changed for clients, such as turnaround on a first draft, gives a buyer something to test in a pitch meeting.

Professional services and AI, 2026

Clients want it; few see it, and AI rarely links to firms

78%of corporate clients call AI-enabled quality very important or essentialThomson Reuters Institute, Mar–Apr 2026 ¹

6%say most or all providers deliver itThomson Reuters Institute, Mar–Apr 2026 ¹

Under 4%of US ChatGPT answers in professional services carry a citationSimilarweb, May 2026 ³

Why AI answers rarely send clients to a firm's site

AI engines answer professional-services questions mostly without links, so the firm's name has to be in the answer itself. Similarweb's data on US ChatGPT prompts shows citation presence rising from about 1.6% in June 2025 to about 6.8% by May 2026.³ In the same data, professional services sat under 4%, against about 23% in travel.³

Business buyers ask these engines for recommendations anyway. In G2's 2026 Buyer Behavior Report of more than 1,000 software buyers (published July 2026; data period and region not stated), 82% had sourced software recommendations from an AI chatbot in the previous 24 months.⁴ That survey covers software, not advisers, but the habit carries over to any business purchase.

So the test for a firm in the United States is simple. Ask ChatGPT, Gemini or Perplexity "Which mid-sized accounting firms in the United States use AI responsibly on audit work?" and read what comes back. Our note on citation rates by sector explains why mentions, not clicks, are the metric to watch in this sector.

What should the firm publish?

Publish four things on crawlable pages: an AI method page, credentials, sector experience and dated case facts. Each gives a buyer something to verify and gives an AI engine a sentence to repeat.

An AI method page

Say where AI is used on client work, where it is never used, who reviews the output and how client data is handled. Name the review step: "Every AI-assisted draft is reviewed and signed by a qualified professional before it reaches a client." State whether client documents are used to train any model. Keep the page dated and update it when tools change.

Credentials and people

List licences, professional memberships and certifications on each person's page, in text. Engines repeat what they can read. A partner's bio that says, for example, "licensed CPA who leads audits of manufacturing companies" answers a buyer prompt; a headshot and a quote do not.

Sector experience and case facts

Write one page per sector you serve, with the matter types, typical client size and the questions clients ask. Add case facts a client has approved: the kind of problem, the work done and the timeline, dated. Our guide for accounting firms and our note on how people choose lawyers show the same approach for two professions. Where conduct rules restrict superlatives, our note on Australian law firms covers what a firm can publish safely.

What a firm can publish this quarter

1. AI method page

   Where AI is used, where it is not, who signs off, how data is handled.

2. People pages in text

   Licences, memberships and sector experience a crawler can read.

3. Sector pages

   Matter types, client profile and the questions clients ask.

4. Dated case facts

   Approved by the client, specific about the work and the timeline.

5. Check the answers

   Run buyer prompts in the main engines and log what they say.

An intake triage agent shows AI-enabled service in practice

An intake triage agent is the safest place to show clients AI-enabled service, because it sorts enquiries and never gives advice. It reads each new enquiry from the web form or inbox, tags the matter type and urgency, checks for conflicts against the client list and drafts a routing note for the right partner.

The controls are simple. The agent drafts and a person routes. Anything that looks like a deadline, a regulator or a dispute goes straight to a named person. The agent never replies to the prospect with substance, only with an acknowledgement and a realistic response time.

Our view: a firm that answers every enquiry within a working day, with the right partner named, shows its AI use better than any brochure. Our AI automation page describes how we build agents like this on the CRM or practice-management system a firm already runs.