ChamberOne

Who benefits from AI productivity gains in legal fees?

By the ChamberOne team 5 minute read

A lawyer who drafts in two hours what once took six has a pleasant problem and an awkward one. The pleasant problem is capacity. The awkward one is the invoice. Billed by the hour, the saving goes to the client and the firm's revenue falls. Billed by the matter, the saving goes to the firm. Neither extreme feels entirely fair, and clients are beginning to ask.

Key points

  • Record time honestly; let efficiency show up in the pricing model, not in inflated entries.
  • Pass-through, fixed fees, value pricing and shared savings each answer 'who benefits' differently.
  • Prepare honest answers for clients who ask how you use AI and whether you charge for it.
  • Test fixed fees on repeatable tasks and review them with your own data.

The tension in hourly billing

Hourly billing ties revenue to time. If a tool halves the time, revenue on that task halves too, unless the firm bills for more tasks. Firms that rely on the hour therefore have an incentive, however unspoken, to use efficiency tools carefully, and clients know it.

Some firms fear that being open about AI will invite demands for a discount. Others fear that hiding it will damage trust when it comes out. Most clients, asked directly, would rather hear it from you.

First, honesty about time

Whatever pricing model you use, the starting point is honest recording. Time entries should reflect the time that the work actually took. Billing a client for hours that were not spent is a problem however the work was done. Check the guidance from your regulator or bar on time recording and charging, and on how technology costs may be treated.

If a task took forty minutes with a tool and a lawyer's review, record forty minutes. The efficiency gain should then show up in the pricing model, not in inflated entries.

Four ways firms respond

  • Pass it through: stay on hourly rates, bill the lower time and accept the revenue fall, trusting that happier clients send more work.
  • Fixed or capped fees: quote a price for the whole task or stage, based on the work required, and keep any gain if it is done faster, accepting the risk if it takes longer.
  • Value-based pricing: price by the importance of the outcome to the client, not by effort. This suits work with a clear benefit, such as a transaction with a deadline.
  • Share the saving: agree a price, then adjust for efficiency, for example by giving a discount or agreeing to split savings against an estimate.

What clients are entitled to ask

Sophisticated clients, especially corporate legal departments, are starting to put questions in their tenders and engagement terms. Do you use AI on our work? How do you check it? Will the saving show up in the fee? Are you charging us for the tool?

Have honest, consistent answers ready. A firm that can say 'we use these tools on these tasks, a lawyer checks every output, and here is how it affects our pricing' will do better than one that is vague. Do not claim that AI is used when it is not, or that it saves more than it does.

The cost of the tool itself

AI tools cost money, and firms spend time on training, policies and supervision. It is reasonable for these to be reflected in overhead and in pricing. But charging a client separately for software that the firm uses on all its matters can be controversial. Check your regulator's guidance on what may be charged as a disbursement or overhead, say plainly in the engagement letter what you will charge, and avoid surprises.

A worked illustration

Suppose a small commercial firm prepares standard supplier contracts for a client. Before using AI, each took around five hours of lawyer time. With a good template, a first draft by a tool and a careful review, it takes about two. The firm could keep charging on the old basis, which would be unfair if disclosed and unwise if discovered. It could bill two hours and lose revenue. Or it could agree a fixed fee for the contract at a level between the two, tell the client why, and use the freed time for work the client values more.

The client gets a lower, predictable price. The firm keeps a fair margin for the investment and the checking. Both sides know where they stand. That is not the only fair answer, but it shows that fairness can be designed.

Risks on the firm's side

  • Discounting too quickly before you know the true saving, which can leave the firm worse off.
  • Moving to fixed fees without good records of time, so that you cannot tell which fees are profitable.
  • Letting review time shrink because the tool is quick, and then paying for mistakes.
  • Training juniors less because routine work has gone, which may hurt the firm in future years.
  • Over-promising to clients about speed and cost.

Practical steps

  1. 1

    Keep recording time honestly, including on fixed-fee work, so that you learn what tasks really cost.

  2. 2

    Choose two or three repeatable tasks and test a fixed fee for each.

  3. 3

    Write a short statement of how you use AI and how it affects pricing, and share it with clients who ask.

  4. 4

    Review pricing every six months, using your own data.

  5. 5

    Talk to your clients. Ask what they value most, and whether they want savings, speed or certainty.

Do
  • Keep recording time honestly, including on fixed-fee work
  • Test a fixed fee on two or three repeatable tasks
  • Write a short statement of how you use AI and pricing
  • Review pricing every six months using your own data
Don't
  • Discount too quickly before you know the true saving
  • Move to fixed fees without good records of time
  • Let review time shrink because the tool is quick
  • Over-promise to clients about speed and cost
Pricing steps and risks

Fairness over the long run

A pricing model that clients see as fair will last. One that looks like quiet profit-taking will be challenged. The firms most likely to do well are those that can explain, in plain words, how the way they work benefits the client as well as themselves. That means sharing some of the gain, charging fairly for the skill that remains and being clear about what has changed. Clients value lawyers for judgment, and judgment is where the fee should rest.

It is worth revisiting the question as tools improve. The saving on a given task may grow, and what looked like a fair price this year may look generous next year. Build a review date into any fixed-fee arrangement, and be ready to talk to the client about it before they raise it.

This article is general information about the law, legal practice and technology, current when it was published. It is not legal advice. Laws and guidance change: check the sources linked above and take advice on your own circumstances from a lawyer qualified where you are.

Read next

Who benefits from AI productivity gains in legal fees?