PublishedOctober 27, 2026
Last reviewedOctober 27, 2026
Editorial ownerFounder, Fortera Labs
Review status✓ Human-approved
AI assistedYes — disclosed
Sources5 cited
Back to Insights
StrategyPricingProfessional ServicesGenerative AILegal Ethics5 min

Billing by the Hour When the Work Takes a Fraction of the Time

When generative AI shortens the drafting, the ABA’s ethics guidance says an hourly biller may charge only for the minutes actually spent. Firms still pricing by the clock hand the efficiency gain to the client and keep the cost.

Fortera Governance PracticeReviewed by Founder, Fortera Labs · Last reviewed October 27, 2026

Billing by the hour assumes the work takes the hours it takes. Generative AI changes that assumption, and the American Bar Association’s Formal Opinion 512 says who keeps the difference.

The ABA has already said who keeps the saved time

On July 29, 2024, the ABA Standing Committee on Ethics and Professional Responsibility issued Formal Opinion 512, its guidance on lawyers’ use of generative AI (GAI) tools. The opinion’s section on fees addresses billing. In its words, “lawyers who bill clients an hourly rate for time spent on a matter must bill for their actual time.” The opinion also bars charging for “more hours than were actually expended on the matter.” If a lawyer “expends 15 minutes to input the relevant information into the GAI program” so the tool can draft a pleading, the lawyer may charge for those 15 minutes and for the time spent reviewing the draft. The hours the tool saved never enter the calculation.

The opinion also addresses costs a firm might try to pass through. A lawyer “may not charge a client to learn about how to use a GAI tool or service” that the lawyer uses regularly for clients. For a tool that works like equipping the practice, the lawyer “should consider its cost to be overhead and not charge the client for its cost,” absent a contrary disclosure to the client. A flat fee has its own catch: where the fee was set on the assumption of manual work, using a GAI tool may make it “unreasonable under Rule 1.5 for the lawyer to charge the same flat fee.”

Our reading: a firm that changes nothing ends up with a smaller number. Hourly clients pay only for the time spent, and a flat fee built on manual-work assumptions can fail the reasonableness test the opinion describes.

The time savings are real, but they arrive unevenly across tasks

A preregistered field experiment run with Boston Consulting Group was published by Harvard Business School as Working Paper 24-013, “Navigating the Jagged Technological Frontier,” and put 758 knowledge workers through 18 realistic knowledge tasks within the frontier of what the AI model could do. Workers using AI completed 12.2% more tasks and finished them 25.1% more quickly on average. The frontier is jagged, though. On a task selected to sit outside the model’s capability, the abstract reports that workers using AI were less likely to produce correct solutions than workers without AI.

The study covered consultants, not lawyers, so read it as a signal about knowledge work rather than a measure of legal tasks. Our reading of the two results is this: the first means the hours a firm can honestly bill fall as the tools reach each kind of task, and the second means the review time Opinion 512 permits is the part of the work that catches the tool’s errors. Our position: that review deserves its own price, and a pricing model that bills review at the same rate as data entry prices the wrong thing.

The 2025 Report on the State of the US Legal Market, released on January 7, 2025 by Thomson Reuters and Georgetown Law’s Center on Ethics and the Legal Profession, found “law firm billing rates accelerating at their fastest pace since the Great Financial Crisis.” Under “Shifting pricing paradigms,” it also predicted that the traditional billable hour model “will be challenged by alternative pricing structures.”

Practitioners have flagged the incentive problem and the room for value pricing

Writing for the Thomson Reuters Institute in April 2025, William Josten of Honigman headed one section “The fallacy of the billable hour.” He argued that, on a time-based reading, a lawyer who takes five times as long to complete a basic document, such as a will, by handwriting it would have a stronger claim to a larger fee. In ABA Law Practice Today, Michele Carney wrote in December 2024:

Lawyers should be able to set their fees using GAI tools without violating Model Rule 1.5 and remain profitable.

Lawyers should be able to set their fees using GAI tools without violating Model Rule 1.5 and remain profitable.

The opinion does not forbid value pricing. Josten reads the opinion’s caveat limiting the actual-time rule to lawyers who bill hourly as one that “clearly leaves room for the contemplation of alternative fee arrangements,” and he notes that even with an agreed fixed fee, the ABA still applies the requirement of reasonableness. What the opinion rules out is billing for time that was not spent. Price scoped fixed fees on outcome and risk, sell subscriptions for recurring work, and state a premium for review and judgment. Our reading: none of these depends on counting minutes, and writing down what you sell other than hours is the work.

You can start repricing this quarter by working through three steps

Start now, practice group by practice group. If you lead a practice group, work through this sequence before your next rate letters go out:

  1. Inventory the task mix per matter type. List your five most common matter types and tag each task as tool-handled or judgment, negotiation and advice. The first group is where hours disappear; the second is where the fee should move.
  2. Rebuild the fee basis in the engagement letter. Opinion 512 says that before charging a client for GAI use, “the lawyer must explain the basis for the charge, preferably in writing.” Apply the same discipline to the whole fee: write one paragraph per fee arrangement stating what the fee buys, and set a new number for legacy flat fees built on manual assumptions. Check how your state’s version of Rule 1.5 treats the change before you send it.
  3. Price review and judgment explicitly. Our reading: the Harvard Business School result on tasks outside the frontier is a reason to price supervision separately. Add a separately named review-and-judgment line to your next fixed-fee quote instead of billing it as leftover hours.

Under Formal Opinion 512, an hourly biller cannot keep the hours its tools save. Our position is plain: the firms that keep their margins will stop selling time before the tools make time the cheapest thing they have.

Source basis

  1. 01American Bar Association, Standing Committee on Ethics and Professional ResponsibilityFormal Opinion 512: Generative Artificial Intelligence Tools
  2. 02Harvard Business SchoolWorking Paper 24-013: Navigating the Jagged Technological Frontier — field experimental evidence on AI and knowledge-worker productivity
  3. 03Thomson Reuters and Georgetown Law Center on Ethics and the Legal Profession2025 Report on the State of the US Legal Market: release on billing rates, demand and pricing
  4. 04Thomson Reuters InstituteWilliam Josten: Is ABA Formal Opinion 512 off the mark, and what can law firms and GCs do about it?
  5. 05American Bar Association, Law Practice Division (Law Practice Today)Michele Carney: Formal Opinion 512 and the reasonableness of fees when using AI

Related reading

Working through a governance question like this one?

Talk to the practice