Not automatically. If you bill by the hour, you can only charge for time actually worked, so faster work means a smaller bill unless you change how you price. If you charge a fixed fee for an outcome, you can usually keep the price, sharing some of the gain where clients would expect it, and being open about using AI.
The better question is what the client is paying for: your time, or the result and the judgement behind it. What follows: what professional guidance already says, how the answer changes with each pricing model, a worked example with numbers, and wording for telling clients about a change.
It depends on how you price today
| How you charge | What faster work does | What's fair |
|---|---|---|
| By the hour | The bill falls, if hours are recorded honestly | Bill actual time. If that hurts, change the pricing model for that kind of work, openly |
| Fixed fee per job | Your margin rises | Keeping the price is usually fair if the result is the same. Revisit it if the fee was quoted on an estimate of hours, or if the work has become routine |
| Per appointment or per item | AI trims the admin around the appointment, not the appointment | No change needed. Don't shorten the service to match |
| Retainer or subscription | Frees capacity within the same fee | Use it to give clients more, and review at renewal |
Most small businesses in health and care, such as dental practices, osteopaths and pharmacies, charge per appointment or per item. For them, AI mostly shortens the admin behind the scenes, and the price question barely arises. The difficult cases are hourly work and fixed fees that were built on hours.
Retainers sit in between. A marketing consultant on a $1,500 monthly retainer, scoped as "up to 20 hours of content and campaign support", might find AI drafting cuts the same output to 14 hours. The bill doesn't change, but the scope wording now overstates the time. The honest options are to fill the six hours with something the client can see, such as a monthly one-page review of competitors' campaigns, or to rewrite the scope around deliverables (four articles, two email campaigns, one report) at renewal. Leaving "20 hours" in a contract you know you no longer work is the retainer version of padding.
What professional guidance already says
Lawyers already have formal guidance on this. In July 2024, a large bar association published Formal Opinion 512 on generative AI tools, and its section on fees sets out principles that apply to anyone who bills for professional time:
- Hourly billing means actual time. Lawyers who bill hourly must bill for the time actually spent, not for the time the work would have taken without AI.
- Flat fees should account for efficiency. The opinion says lawyers should take efficiencies into account when charging flat fees, too.
- Learning time isn't billable. Time spent learning a tool you'll use across your practice can't be charged to clients, unless a client specifically asks you to use a tool you don't know.
- Normal software is overhead. AI built into standard tools, such as a word processor, shouldn't be charged to clients. A standalone tool used for a particular client's work may be a reasonable expense.
It's written for lawyers, not for your profession. If you belong to a professional body, check whether it has issued its own guidance on AI and fees, and follow that. Where it hasn't, these four principles are a sound default.
Worked example: a podiatrist's medico-legal reports
Say a podiatrist writes expert reports for personal injury cases, billed at $150 an hour, about six reports a month. Rates and hours are invented for the example. A typical report takes seven hours: two and a half reading medical records, one and a half examining, three writing.
With a business AI plan, the podiatrist uses AI to build a chronology from the records, which they check against the originals, and to draft the factual sections. The examination and the opinion stay entirely theirs. Reading time falls to one hour and writing to two, so a report now takes four and a half hours. Business plans such as Claude Team Standard or ChatGPT Business cost $25 a seat a month billed monthly, with a two-seat minimum, so about $50 a month for a sole practitioner. That's overhead, not a client expense.
| Option | Client pays per report | Podiatrist's effective hourly rate | Verdict |
|---|---|---|---|
| A. Keep hourly, record honestly | $675 (4.5 hours) | $150 | Fair; the client gets the whole saving |
| B. Move standard reports to a fixed fee of $900 | $900 | $200 | Fair if agreed up front; both sides gain |
| C. Keep billing seven hours | $1,050 | Not applicable | Dishonest: billing time that wasn't worked |
Option A is simplest but hands the whole gain to clients, and at six reports a month income drops from $6,300 to $4,050, unless the freed 15 hours are filled with more work. Option B is the usual answer. Clients pay $150 less than before, the podiatrist earns more per hour, and the price no longer depends on how the report was written. The freed hours can go on more reports or on work AI can't help with.
Two cautions for this kind of work. Expert evidence carries its own duties, and the opinion must be genuinely the expert's own, so check any rules that apply to expert reports and tell the instructing solicitor how AI is used. And medical records are sensitive: they only go into a tool on a business plan with terms you've checked. The time savings themselves need measuring honestly, too; measuring time saved after rolling out AI shows how.
Four fair ways to price AI-assisted work
- Fixed fees for defined work. Describe the deliverable and what's included, and price it. This is the cleanest answer for anything repeatable.
- Tiers. A standard version at a lower fixed price and an extended version with more of your time, such as a longer review or a follow-up call. The client chooses how much of you they're buying.
- Same price, more value. Keep the fee and give more: faster turnaround, an extra check, a clearer summary. This only works if the extra is something the client actually values and can see.
- Share the gain openly. A modest reduction for work types where AI clearly cut the effort, explained as such. It builds trust with long-standing clients and costs you less than losing one.
Tiers suit work where clients' needs vary. An HR consultant who writes staff handbooks might offer a standard handbook at $1,200, drafted with AI from a policy library the consultant maintains, checked against the client's contracts, with one review call and one round of edits. The extended version, at $2,000, adds policies written for up to three unusual roles, a second review call and a briefing note for managers. If the standard tier now takes about eight hours and the extended about fourteen (illustrative), both earn more per hour than the old $95 hourly rate, and a client who wants more of the consultant's time can see exactly what the extra $800 buys.
Fixed fees move risk onto you, and AI adds a new kind of it. If a draft comes out wrong and you spend three hours untangling it, that's your cost, not the client's. So set fixed fees on your realistic time including checking and the occasional bad run, not on your best day. Define the scope tightly too: "standard report, up to 200 pages of records, one examination" protects you when a case arrives with 900 pages.
The podiatrist in the worked example shows how easily a best day gets priced in. The first AI-assisted report took three hours, which would have pointed to a fixed fee of around $600 (illustrative). The next five averaged four and a half. One set of records arrived as poor scans of handwritten notes, and on another the chronology put two appointments in the wrong year, which took an hour to trace back to the originals. A fee set on that first report would have come in below the honest hourly bill of $675, the reverse of the gain the change was meant to capture. The $900 in the table rests on the five-report average, with a scope line for records over 200 pages.
Whichever you choose, the checking time is still real work and still yours to charge for. Reviewing AI output properly isn't a formality; setting up human review for AI work covers what good checking involves.
Working it out service by service
Don't reprice your whole business in one go. Take it one service at a time, starting with the ones AI has changed most.
- List your services and how each is priced today.
- Measure the time for each, before and after AI, on at least five real jobs. Include checking time. Estimates made in the first enthusiastic week are usually too optimistic.
- Ask what the client is buying. For some services it's mainly your time; for others it's your name on the result, your accountability, or speed.
- Pick a model per service from the four above, and write down why.
- Change prices at a natural point, such as a new engagement, a renewal or the start of a new year, with notice.
- Review after six months. Check whether clients noticed, whether you won or lost work, and whether the time saving held up.
A spreadsheet with one row per service and columns for old time, new time, old price, new price and effective hourly rate makes the decision visible. It also stops the most common mistake: discounting the service clients care about least while leaving an obviously automated one at its old price.
Filled in for a three-person bookkeeping practice, after measuring five jobs of each kind, it might look like this (all figures illustrative):
| Service | Priced as | Time before / after | Price before / after | Effective hourly rate | Decision |
|---|---|---|---|---|---|
| Monthly bookkeeping, up to 150 transactions | Fixed, monthly | 5 h / 3.5 h | $400 / $400 | $80 to $114 | Keep: clients buy reconciled books and our sign-off |
| Year-end accounts pack | Hourly at $85 | 9 h / 6 h | $765 / $680 fixed | $85 to $113 | Move to fixed fee, below last year's average bill |
| Payroll | $12 per payslip | Admin only | $12 / $12 | Not measured | No change |
| Monthly cash-flow commentary | Fixed | 3 h / 1 h | $250 / $180 | $83 to $180 | Cut: clients can see it's now largely drafted for us |
The commentary line is the one that would have caused trouble. It's the most visibly AI-assisted thing the practice sends, and leaving it at $250 while the effort fell by two-thirds is exactly what a sharp client notices. Cutting it still nearly doubles the practice's rate on that work.
What not to do
- Pad hours. Recording the time a job "would have taken" is billing for work you didn't do.
- Charge for your learning curve. Getting good at a tool is your investment.
- Add the subscription to invoices when it's part of how you run the business.
- Stay quiet when the contract is hours-based. If a client agreed a fee "based on an estimated 20 hours" and it now takes 8, raise it before they do.
- Cut prices across the board out of guilt. Most of your value was never the typing. Reprice item by item, where the effort genuinely changed.
The fourth point is where it most often goes wrong in practice. Picture a two-person copywriting studio that quoted a website rewrite at $1,800, "based on an estimated 20 hours at $90". With AI first drafts, the job took 11 hours, and the studio invoiced the full $1,800 without comment. Two months later the client's marketing lead asked for a timesheet before approving the next project, having recognised the drafting style. The studio ended up crediting $810, the nine hours not worked, and lost the goodwill it had built. Had it written at the halfway point, "this is running well under the estimate; we'll invoice actual hours, or we can agree a fixed fee for the next phase", it would have looked honest and probably kept the next project at a fixed price.
What to tell clients
If you're moving a service from hourly to fixed fee, say why, briefly and without apology. Something like:
Subject: A simpler fee for [service] from [date]
From [date], [standard reports / this service] will be charged at a
fixed fee of $[amount], instead of by the hour.
We now use AI tools, on business plans that don't train on your data,
to organise records and prepare first drafts. Every [report / document]
is still checked, completed and signed off by [name], and the
professional judgement in it is entirely ours.
The fixed fee is lower than the average we charged by the hour over
the last year, and you'll know the cost before we start.
Anything outside the standard scope will be quoted separately first.
If you'd like to discuss it, just reply.Adapt the middle paragraph to what's true for you. If your contracts need updating to cover AI use, mentioning AI use in client contracts and proposals has clause wording, and whether to tell customers you use AI covers the wider disclosure question.
The harder message is the reply to a client who raises it first. Suppose a bookkeeping client from the table above writes: "Now you're using AI, we'd expect our monthly fee to come down by half." A defensive first draft tends to read: "AI doesn't really change much, and our fees reflect our expertise." It dodges the question, and the client can tell. A better reply answers with facts:
Fair question. AI now does the first pass of coding your
transactions, which saves us about an hour and a half a month on
your books. The reconciliation, the checks against your bank
statements and the sign-off take the same time as before, and we're
still responsible if anything is wrong.
Your fee has been $400 a month since [date]. Rather than raise it at
renewal, we'll hold it there. If your volume stays under 100
transactions a month, we can also move you to our lighter package at
$340. Happy to talk it through on a call.It gives the client a real concession and a real option, without pretending the work vanished.
When lowering your prices is the right call
Sometimes the market decides for you. Lower the price of a specific service when:
- Competitors have already repriced the same work and clients are comparing.
- The work has become routine enough that clients could get most of it from a tool themselves.
- A cheaper price would win noticeably more of the work, and you have the capacity AI freed up.
- A valued client asks, and the saving to them matters more than the margin to you.
Look at what AI did to your costs before deciding. If a service's effort halved, you can cut its price by a quarter and still earn more per hour. In a small translation agency, for instance, a batch of 40 product descriptions for an online shop might once have taken four hours and been priced at $320. With AI first drafts and a fluent translator's edit, it takes two. Dropping the price to $240, in line with what the client is being quoted elsewhere, still earns $120 an hour against $80 before. That's a decision you can make calmly, with numbers, rather than one forced on you by a client who noticed first.
A pricing rule for each way you charge
- Hourly: bill actual time, always. If AI makes that uneconomic, move the affected work to fixed fees, openly.
- Fixed fee: keep the price if the result and your accountability are unchanged; revisit fees that were quoted as estimates of hours.
- Per appointment: nothing to change. Spend the saved admin time on patients or customers.
- Everyone: never pass on normal software costs or learning time, and tell clients how AI is used in their work.
Pricing and AI: follow-up questions
Do I have to tell clients I use AI?
There's no single rule for every business, but your contract, your professional body's guidance and any client instructions may require it, especially where client data goes into an AI tool. Even where it isn't required, telling clients how you use AI and what stays with you avoids an awkward conversation later if they find out another way.
Can I charge more because I use AI?
You can charge more if the client gets more: faster turnaround, extra checks, or a service you couldn't offer before. What you shouldn't do is add a charge for the AI tool itself when it's part of your normal software, or bill hours for work the tool did. Price the result and the judgement, not the software.
What if a client asks for a discount because AI did the work?
Take it as a pricing conversation, not an accusation. Explain what the fee covers: your review, your judgement, your accountability for the result, and the time you still spend. If you bill hourly, show that the bill already reflects the actual time. If the work genuinely became routine, a lower fixed price for that item may be the right answer.
Further reads
- How to Calculate AI ROI for Your Business (Worked Example) — Work out what the time saving is worth to you first.
- How Much Time Can AI Really Save a Small Business Each Week? — Realistic time savings before you reprice anything.
- How Much Time Can AI Save a Solo Consultant Each Week? — The same pricing question for one-person firms.
- How to Write an AI Disclosure Statement for Your Website — Public wording on how you use AI.
- Is It Safe to Put Customer Data Into ChatGPT? — Before client files go into any AI tool.
- How to Redesign Job Roles Once AI Handles Routine Tasks — What to do with the hours AI frees up.
- Will AI Replace My Employees? An Honest Answer for Small Firms — An honest answer for owners: what the evidence shows, a task-by-task method to score each role, a bookkeeping firm example and what to tell staff.
- Pros and Cons of Using AI in Your Business, With Real Costs — Every pro and con of business AI with a price attached, a first-year ledger for a small law firm, and the costs that never reach an invoice.
- Why AI Saves Time but Not Money, and How to Capture the Gain — Why hours saved by AI vanish from the accounts, the four routes that turn them into money, and a worked example from a small brewery.
- AI for Freelance Designers: Faster Concepts Without Rights Issues — Where AI speeds up concept work, how to pick generators by their terms, a process log that proves authorship, and the checks before you hand over files.
- Turning Hours Saved by AI Into Advisory Work Clients Pay For — How a practice stops AI time savings leaking away and turns them into a fixed-fee advisory service clients actually buy.
- Why AI Adoption Stalls in Accounting Firms, and How to Restart It — The five reasons AI stalls in accounting practices, a twenty-minute diagnosis, and a 30-day restart timed around the deadline calendar.
- Should Freelancers Tell Clients They Use AI? — Decide when to disclose AI use, ask for specific approval and explain your process without making promises you cannot support.
- AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.
Sources: Formal Opinion 512 on generative AI tools (July 2024), as summarised in published legal commentary; ChatGPT Business and Claude Team list prices from vendor pricing pages.