What to say when a client questions your fee because of AI

When a client questions your fee because AI helped produce the work, the useful response is one you prepared before the meeting, not one you improvise while they wait.

Decide what your fee actually pays for: the judgement to tell sound work from plausible work, the checking that catches an error before it reaches the client, and the accountability when something goes wrong. AI has made the hours shorter, so the hours no longer explain your value on their own.

This piece sets out how to settle that position and say it plainly, before a client settles it for you.

Why clients are starting to ask

The question is fair, and it is coming. Your clients can see that AI drafts a document in the time it used to take to open the file. The management-accounts commentary that took an associate half a day now arrives by mid-morning. The candidate shortlist lands the same afternoon the brief went out. When the work speeds up and the invoice holds still, a sensible client asks why.

The capability behind that question is moving quickly. Ethan Mollick, who studies how people actually work with these tools, describes a shift from chatting with a model to handing it whole tasks. He notes that "a few months later, you can get sixteen hours or more of work from a single prompt". A tool that can do most of a day's drafting in one pass changes what an hour of billed time looks like from the outside.

Clients have noticed, and most firms have not caught up with the conversation. Thomson Reuters, in its 2026 survey of professional-services firms, found that around three-quarters of clients and firms agree the firm should lead the discussion about how AI is used on their work. Those discussions have mostly not happened yet. The same survey found 40 per cent of firms are getting contradictory instructions, with some clients asking them to use AI and others asking them not to. Into that silence, a client writes their own story about what your work is now worth. It is usually cheaper than yours.

What your fee actually buys

Here is the part worth being clear about, first to yourself and then to the client. When AI drafts the work, what is left is the part that was always the point: knowing whether the draft is right.

Picture the three scenes. The accountant reads the AI-drafted commentary and it is clean, except for one figure. That figure is materially wrong, and without the accountant it would have gone to the board unquestioned. Catching that is the job. The recruiter gets ten well-summarised candidates and knows which one will fit the team and stay, which the summary cannot tell you. The agency lead reads campaign copy that is fluent and sits on the wrong side of a brand line the client would not forgive. In each case the machine produced the draft in minutes, and the expertise that made it safe to send took years to build.

That expertise is the work itself. The client is paying for the person who can tell a sound draft from a plausible one. Mollick's research points the same way. In a study of professionals using AI, the best results came from those with the deepest domain experience, whatever their technical background. Experience is what lets a person see when the output is wrong. The client is also buying the accountability behind it. Your name goes on the advice. You carry the responsibility if it fails. A person stands behind the result, rather than a model that will cheerfully produce a confident mistake.

State it in those terms and the fee stops looking like a charge for keystrokes. It looks like a charge for judgement and for carrying the risk.

Decide your position before the client sets it

Work out your answer in advance and write it down. The alternative is deciding it under pressure in a meeting, and that is how firms end up discounting on the spot and teaching clients to expect it next time.

There are two honest routes, and many firms use a mix. You can keep charging by the hour and be plain that the rate buys expertise and responsibility, while billing the time you genuinely spend. Or you can move suitable work to fixed or outcome-based fees, so the price reflects the result the client gets rather than the hours it happened to take. The second route fits AI-assisted work better over time, because it stops tying your income to a measure the tools are actively shrinking.

The number itself is a separate question, and we have written about the mechanics of it in what to charge when AI cuts your hours. This piece is about the position underneath the number. Two things make it up: what you have decided your firm sells, and how you answer a client who asks why AI-assisted work costs what it does. If you cannot say those sentences plainly, the pricing maths will not save you, because the client hears an unsure supplier and pushes.

Be honest about the hours you did not spend

One line has to stay on the right side of trust. If you bill by the hour, bill the hours you actually worked. When AI turns a day's task into an hour, charging for the day is a problem waiting to surface. Clients are better than ever at sensing it.

The honest response is to change what you charge for, not to keep the clock running on time you did not spend. Charge for the judgement, the verification and the outcome, and let the hours fall where they fall. A firm that does this can look a client in the eye and explain the bill. A firm that pads its time to protect old revenue is one procurement review away from an awkward conversation. Trust, once it goes, tends to take the client with it.

This is where setting a position pays off. Deciding in advance what your firm sells, and being able to explain it, is exactly the kind of judgement call our AI Lessons for Leaders coaching helps a leader think through. That means what to keep in human hands, how to describe it to a client, and how to price work as the tools keep changing under you. If a fee conversation is coming and you would rather not improvise it, book a discovery call and we will work through your real client situations.

Common questions

Should clients pay for AI-assisted work?

Yes, when the fee reflects the judgement, checking and accountability the firm adds, rather than the raw time the tools saved. The client is paying for a sound, defensible result they can rely on, and for the person who stands behind it. What changes is how you justify the fee, and sometimes how you structure it.

How does a firm charge for work that AI made faster?

There are two honest routes, often combined. Keep hourly billing and be clear the rate buys expertise and responsibility, while charging the time you actually spend. Or move suitable work to fixed or outcome-based fees so the price tracks the value delivered rather than the hours. The right split depends on the type of work and what your clients value most.

What do you say when a client asks for a discount because AI did the work?

Acknowledge that it is a fair question, then explain what the fee covers: the errors you catch, the risk you carry, the accountability if the work goes wrong. Decide that answer before the meeting, so you are explaining your value rather than defending it, and so you are not discounting under pressure.