A Client Asks Whether You Used AI on Their Matter

At some point, a client will ask. Maybe they read something. Maybe they heard a story about a firm that fed confidential documents into a public chatbot. Maybe they just want to know. Whatever the reason, the question is coming, and a fumbled answer costs more than the question ever would have.

This article is about preparing a straight answer before you need one. It covers what clients are usually actually asking, where the real trade-offs sit, and what to write down so the conversation is calm when it arrives.

A note before we start: this is about client communication and positioning. Professional obligations, disclosure duties, and conduct rules sit with your own professional body and adviser. This article does not tell you what you are required to do. It helps you think through what you want to say.

What the Client Is Usually Actually Asking

When a client asks "did you use AI on my matter", they are rarely asking about software categories. They are asking two things.

First: did my confidential information go somewhere I did not approve? They want to know whether their details, their documents, their financials, their situation, were fed into a public tool that stores prompts, trains on inputs, or exposes data to third parties. That is a data question, not a technology question.

Second: did a human actually think about this? They want to know whether a qualified person read the work, checked it, and took responsibility for it, or whether a machine produced something and it went out the door. That is an accountability question.

Neither question is unreasonable. Both deserve a direct answer. If you do not have a direct answer ready, the pause itself signals something.

Why the Defensive Answer Costs More

Some professionals hear the AI question and go vague. They deflect, they qualify, they say something like "we use various tools to ensure efficiency." That answer sounds evasive because it is evasive. A client who was simply curious becomes a client who is now suspicious.

The defensive answer also puts you on the back foot for every future conversation. If the client later finds out that AI was involved and you were not clear about it, that is a trust problem that no invoice credit fixes.

Contrast that with a principal who says: "Yes, we used an AI tool for the first-pass research. I reviewed everything, I made the calls, I signed off on the advice. Your documents stayed within our closed environment and were not shared with any external service." That answer is specific. It is calm. It demonstrates that someone thought about this before the client asked.

Specificity is the currency of trust in professional services. Vague reassurance costs nothing to say, so clients discount it. Specific detail costs something, so clients believe it.

When AI Touched the Deliverable Versus the Admin Around It

The answer changes depending on where in the process AI was involved. It is worth being clear in your own mind before the conversation.

AI touched the deliverable. This means AI drafted, summarised, researched, or produced something that ended up in the client's work product. A first draft of a report. A summary of case documents. A set of suggested clauses. Here, the client's question is directly relevant. They deserve to know that a tool was involved and that a person reviewed, edited, and took ownership of what went out.

AI touched the admin around it. This means AI was used for internal processes that did not touch the client's specific matter directly. Scheduling, billing summaries, internal knowledge base queries, practice management. Here, the question is less directly relevant to the specific matter. The client may still care about their data (if any of it passed through an AI-assisted system), but the nature of the conversation is different.

Knowing which situation you are in before the client asks means you can be precise instead of defensive. Precision is not exposure. Precision is confidence.

What to Have Written Down Before You Need It

The professionals who handle the AI question best are not the ones with the best improvised answer. They are the ones who sorted this out before it became urgent.

Three things are worth having in writing.

A use policy that your firm actually follows. This does not need to be long. It needs to say which tools are approved, what data can and cannot be fed into each one, and who is responsible for checking AI-assisted work before it goes to a client. A one-page internal document is enough. If a client ever asks to see it, you can show it. That act alone, being willing to show the document, carries more weight than any verbal reassurance. You can find a worked example of what an AI usage policy covers at getplainstart.com/ai-policy-for-law-firms.

A note in your file when AI was used. If AI contributed to a deliverable, record it in your matter notes. What tool. What task. Who reviewed. When. This is not about liability. It is about being able to answer confidently if the question comes up during the matter or after it closes.

Wording in your engagement letter. This is where a lot of firms are now making a proactive choice. More on that below.

A Worked Script for the Conversation

Here is a version of the answer that covers the two real questions without over-explaining.

"Yes, we used [tool name] for [specific task, e.g., first-pass research on the regulatory history]. Everything it produced was reviewed by me before it shaped any advice. Your documents were processed within our own environment and were not shared with any external AI service. If you want to see our AI use policy, I can send it over."

That is four sentences. It answers the data question (your documents stayed in our environment). It answers the accountability question (I reviewed it, I made the calls). It signals a written policy exists without being defensive about it. And it offers evidence, the policy itself, which most clients will not ask for but are reassured to know about.

Adjust for your actual tools and actual situation. Do not use this script if it does not accurately describe what happened. An inaccurate confident answer is worse than a careful one.

A Worked Example: Raising It Proactively in Engagement Wording

A small advisory firm with four principals decided they did not want to wait for the question. They added a short paragraph to their standard engagement letter. It read something like this:

"We use AI-assisted tools for certain tasks including research, document review, and drafting support. All AI-assisted work is reviewed and approved by a qualified member of our team before it forms part of any advice or deliverable. Client documents are handled only within tools that meet our data handling standards, and are not shared with external AI services without prior agreement. If you have questions about our approach, please ask."

That paragraph does several things. It normalises AI use without making it a selling point. It reassures on the accountability question (qualified review, approval). It reassures on the data question (data handling standards, no external sharing without agreement). And it opens the door for the client to ask, which means the clients who care can ask early, before it becomes a difficult conversation mid-matter.

The firm reported that several clients responded positively, specifically because other firms they worked with had said nothing at all. Saying nothing reads as either evasion or inattention. Neither is a position a professional services firm wants to hold.

The Positioning Question Underneath the Trust Question

Here is the thing about the AI question: the clients who ask it are often the ones who value care and craft. They are not asking because they object to technology. They are asking because they want to know that someone is in charge of the work.

That is the positioning question underneath the trust question. Not "do you use AI" but "are you the kind of firm where a person takes responsibility for the work."

The answer to that question is not a technology policy. It is a posture. A firm that has thought this through, written it down, can show it on request, and volunteers it in their engagement wording is a firm that is demonstrating exactly the kind of care those clients are looking for.

Prepare the answer before the question arrives. The preparation itself is part of the answer.