The Client AI Communication Gap: What 2026 Data Shows Firms Are Getting Wrong

Published: May 29, 2026 | By: The Crossing Report

---

On a Tuesday morning in March, a Chicago CPA got two client emails before 9 a.m.

The first was from a mid-size manufacturing company: "Going forward, we expect your team to leverage AI tools on our engagement. Our board has asked vendors to confirm modern AI practices."

The second was from a closely held family business she'd served for eleven years: "We've heard firms are using AI on client files. We need to know you're not doing this with our data."

Same firm. Same morning. Two directly opposite instructions — and no policy for either one.

That scenario is not hypothetical. Thomson Reuters asked thousands of firm professionals about client AI expectations in their 2026 AI in Professional Services Report, and 40% reported receiving contradictory client instructions on AI use. Some clients demanding it. Others prohibiting it. Many in the same week.

The data also showed that despite all this client urgency — real or anticipated — 50% of firm professionals have had zero conversations with clients about AI use. Not one. The silence is almost total.

Here's what that costs, and how to fix it in a week.

---

The Numbers: Where Clients and Firms Are Out of Alignment

The 2026 Thomson Reuters AI in Professional Services Report is the most detailed look yet at the client-side of the AI adoption story. The headlines are striking.

More than half of corporate clients want their outside law and accounting firms using AI on their matters. They see it as a sign of competence — not a risk. Clients who work with large companies have watched their in-house teams adopt AI tools. They expect the same from outside firms.

Fewer than one-third of those clients know whether their firms actually do so.

The disconnect is not because clients are uninformed. It's because firms aren't saying anything. According to the Thomson Reuters AI Guidance Gap research (2026), the majority of clients have never received a communication from their professional services firm specifically about AI use — not an email, not a policy document, not a sentence in an engagement letter.

85% of clients say firms should disclose when AI is used on their matters. Three-quarters of both clients and firm professionals agree that firms should be leading these conversations. They're not.

For a firm owner, the summary is uncomfortable: your clients want you using AI, most don't know if you are, and they believe you should be telling them. The conversation gap isn't a technology problem. It's a communication failure — and one that's entirely within your control to fix.

---

Why 40% of Firms Are Getting Conflicting Instructions

The contradictory instructions problem has a clear cause: clients have no AI policy of their own.

When a manufacturing company's procurement team says "confirm AI practices," they're often responding to a board directive or a vendor questionnaire someone read. When a family business owner says "don't use AI on our data," they're responding to a fear — usually about confidentiality, accuracy, or losing the personal touch they pay for.

Neither client has thought through the implications carefully. Neither has a written policy. They're reacting.

The firm that steps into that vacuum with a clear, written, proactive AI communication wins the trust advantage. Not because clients always prefer AI — they don't, and some genuinely don't want it. But because the communication itself demonstrates competence and respect. You thought about this. You have a position. You're telling them before they have to ask.

The alternative is the "don't ask, don't tell" approach that currently describes most small firms: use AI tools, say nothing, hope no one asks. That strategy has two failure modes.

Failure mode one: A client discovers you're using AI and experiences it as a surprise. Not because AI use is wrong — it isn't — but because you withheld it. The relationship moves from trust to suspicion in one conversation.

Failure mode two: A client gives you a blanket "no AI" instruction because you gave them no framework to think about it differently. You had a chance to have the conversation. You didn't. Now you've locked yourself out of tools that would benefit their work.

The conversation doesn't have to be complicated. It has to happen.

---

Three Conversations Your Firm Should Start This Month

The Thomson Reuters data shows three-quarters of both clients and firm professionals agree that firms should lead the AI conversation. Here's how to do it in three distinct contexts.

The Intake Conversation: AI Disclosure at Matter Opening

The most efficient place to discuss AI use is at the start of a new engagement — before the work begins, when clients are forming their expectations.

Add one sentence to your standard intake intake process: "Before we begin, I want to share how we use AI tools in our practice and get your input on your preferences for this engagement."

This is not a legal disclaimer. It's a professional conversation. It has two effects: it establishes transparency before there's anything to disclose defensively, and it gives the client a structured moment to share concerns — which surfaces the "no AI" objection before you've done six weeks of work assuming otherwise.

For law firms, this also begins to satisfy the professional supervision obligations that ABA Formal Opinion 512 describes: the competent, actively supervised use of AI on client matters.

The Policy Brief: One-Page AI Use Document for Clients

A short written document sent to all active clients is more durable than a verbal conversation. It answers the question before clients form it. It works for existing clients (send it once as an update) and for new clients (include with engagement letter materials).

The document doesn't have to be long. One page. Cover four points:

1. Which AI tools your firm uses and for what kinds of work
2. How AI output is reviewed before it reaches clients (human oversight process)
3. How client data is protected within those tools
4. How clients can tell you their preferences, and that you'll honor them

A firm that sends this document positions itself as thoughtful, organized, and transparent — the same qualities clients value in professional services generally. The firms that wait for a client to ask will always be responding defensively. The firms that send the document first will be cited as examples.

See the one-page AI policy template for small professional services firms for a downloadable starting point.

The Check-In: Quarterly Update for Active Clients

AI tools change quickly. A firm that added one AI tool in 2024 may be running four integrated tools in late 2026. A quarterly one-paragraph email to active clients noting any material changes to your AI practices keeps the disclosure current without requiring formal process.

This is also good relationship management. Clients who work with forward-looking firms want to see that the firm is staying current. "Here's what we added and why, and here's how it affects your work" is the kind of communication that strengthens retention — not because AI is impressive, but because the proactive communication is.

---

What to Actually Say: Language Templates

The biggest obstacle to starting these conversations is not knowing what to say. Here's language that works for the three firm types most affected.

Law Firm Intake Disclosure

"Our firm uses AI-assisted tools for legal research, document drafting, and matter analysis. Every AI-generated work product is reviewed by a licensed attorney before it reaches you. If you have questions about which tools we use or how client data is handled, I'm happy to discuss. And if you have preferences about AI use on your matters, please let me know — we accommodate client requests on a matter-by-matter basis."

Accounting Firm Intake Disclosure

"We use AI tools in our practice for tax research, document review, and workflow management. All AI-assisted analysis is reviewed by a licensed CPA before delivery. Your client data is processed only through tools with business-tier privacy terms — not consumer accounts. If you have questions or preferences about AI use on your engagement, I'd like to hear them before we begin."

Response When a Client Asks "Are You Using AI on My Work?"

"Yes — I want to be straightforward with you. We use [tool name(s)] for [specific task: research/drafting/analysis]. Every output goes through my review before it reaches you. What I can tell you is that it makes my work more thorough and helps me catch things faster — which benefits your matter. I'm also happy to discuss any specific concerns you have about the process."

These aren't scripts to memorize. They're frameworks. The tone that works for a 25-year client relationship will differ from a new institutional client. What they have in common: specific, direct, and confident. The firm that hedges — "we might use some AI sometimes" — signals uncertainty. The firm that is clear signals competence.

---

The Small-Firm Advantage

Here's the structural reality that the Thomson Reuters data points toward: large firms are paralyzed by this question.

Enterprise law firms and national accounting firms have risk management committees, ethics officers, legal counsel, and brand teams involved in every AI communication decision. Some have been drafting client-facing AI disclosure language for 18 months without sending it.

A 12-person accounting firm can write and send a client AI communication policy in a week. The managing partner can decide the position, draft the document, review it, and send it to active clients — all before a BigLaw committee has its first meeting on the topic.

The window to be the proactive, transparent firm in your market is open right now. It will close as the larger firms eventually standardize their disclosures and the conversation becomes industry-standard. Firms that get there first own the trust position; firms that follow are just complying with expectations.

The client AI communication gap is not a technology problem. It's a communication habit. The data from Thomson Reuters is clear: your clients want the conversation. They believe you should be leading it. And the firms that do it first will be the ones clients recommend when other business owners ask "how do you handle the AI question with your firm?"

This week: Draft one paragraph describing how your firm uses AI and how you protect client data. Send it to your five most active clients before the end of the month. That's the whole first step. The conversation takes care of itself from there.

---

For more on building the internal side of this policy first: Your Staff Is Using AI on Client Work Right Now — and Your Firm Has No Policy | AI Governance Framework for Professional Services Firms | How Accounting Firms Are Using AI in 2026

---

Frequently Asked Questions

Should I tell my clients I'm using AI tools?

Yes, and proactively — not just when asked. Thomson Reuters 2026 data shows 85% of clients believe firms should disclose AI use on their matters, yet the majority of firms haven't done so. The firms that disclose first build the trust baseline. Those that wait until a client asks — or discovers it — operate from a defensive position. A one-page AI use policy sent to active clients costs 30 minutes to write and eliminates the ambiguity.

What do I do if a client says "don't use AI on my work"?

Honor it, document it in the matter file, and use it as an opening to understand their concern. The most common client objections are confidentiality, accuracy, and billing transparency. Each has a factual response. Clients who say "no AI" often mean "I don't trust AI I can't see" — not "I want you to work inefficiently on my behalf." A 15-minute conversation resolves most of it.

What if one client requires AI use and another says to avoid it?

40% of professional services professionals report receiving exactly this conflict, according to Thomson Reuters 2026 research. The solution is a matter-level AI use disclosure in your engagement letter — not a firm-wide policy that has to please everyone. Each matter gets a one-line disclosure: "We may use AI-assisted tools for [document drafting/research/analysis] on this matter unless you instruct otherwise." That puts the client in control and eliminates the contradiction at the source. See our guide on AI disclosure in engagement letters for specific language.

How do I handle billing transparency when using AI?

Separate the efficiency benefit from the fee. The value of a legal memo or tax analysis is the judgment, not the hours. If AI tools helped you deliver a result in 2 hours instead of 8, the client's outcome didn't change — your margin did. Be transparent about using AI; don't automatically reduce fees for speed unless speed was your stated value proposition before. Clients who understand this distinction generally respect it. Clients who push back on it are often raising a different concern — usually about whether the work received sufficient attention — that deserves a direct answer.

What's the minimum AI communication policy a small firm should have?

Three elements: (1) A one-page AI use statement sent to clients at matter opening — naming the tools, describing the oversight process, confirming you review all AI output before delivery. (2) A matter-level notation in your file recording each client's AI preferences. (3) A quarterly update to active clients noting any material changes to your AI practices. This takes about two hours to set up and prevents the majority of client trust incidents that arise from silence.