Everyone agrees AI makes lawyers faster. But faster at the billable hour is a pay cut. And nobody has a clean answer for what replaces the lost revenue.
When Anthropic’s General Counsel Jeff Bleich stood up at the ABA’s White Collar Crime Institute this month and declared the billable hour’s days numbered, the room nodded along. GCs from Liberty Mutual, IBM, and Google joined the chorus. The consensus was clean: AI is eliminating the need for large teams doing tedious work, and clients will no longer tolerate paying for time that a machine compressed.
What nobody said, not Bleich, not the panel, not the trade press coverage, is the part that actually keeps managing partners up at night.
If AI cuts the time to do a task in half, what exactly happens to revenue?
Run the Numbers
Let’s say your associate spends 25 hours on a complex brief at $300/hour. That’s $7,500 to the client.
Now your firm deploys AI. The same brief takes 10 hours.
You have three choices:
| Hours billed | 10 |
| Rate | $300/hr |
| Revenue | $3,000 |
| Revenue lost per brief | −$4,500 |
| Target revenue | $7,500 |
| Hours worked | 10 |
| Rate required | $750/hr |
| Rate increase required | +150% |
| Hours billed | 25 |
| Hours worked | 10 |
| Rate | $300/hr |
| Revenue | $7,500 |
| This is called | Fraud |
| “There is no Option D hiding in this math. That is the problem. The billable hour and AI efficiency are structurally incompatible, And every firm that hasn’t solved this is currently bleeding revenue or quietly overbilling.” |
Why a 150% Rate Hike Doesn’t Work Either
Option B looks clean on paper. If AI makes you more efficient, charge more for your expertise. That’s the “strategic value” argument Bleich and the GC panel were gesturing at.
Here’s why it breaks down in practice:
First, your clients have procurement departments now. They benchmark rates across firms. A sudden jump from $300 to $750, even if justified, triggers a sourcing review. You’re not just justifying the work; you’re justifying the rate in a climate where GCs are explicitly bragging to audiences that they’re leaning on AI to cut outside counsel spend.
Second, the rate increase has to be uniform across all work, or the math gets impossible to manage. But not every matter benefits equally from AI. A nuanced appellate brief and a routine NDA are not the same problem. Blunt rate increases punish clients on routine work while still undercharging on complex work.
Third, and this is the part nobody says out loud, associates at your firm still need to eat. Their billable hour quotas haven’t changed. If AI compresses a 25-hour task to 10 hours, that associate now needs to find 15 hours of billable work somewhere else to hit their annual target. Where does it come from?
The Actual Problem: Firms Are Structured Around Volume, Not Value
The billable hour didn’t just emerge as a billing preference. It became the operating system of the entire law firm. Headcount decisions, associate promotions, partner compensation, office lease size, everything is downstream of billable hours times rate.
When GCs say “the value is your strategy, your results” as Liberty Mutual’s GC said at the same ABA panel, they’re describing what they want to pay for. They’re not describing how law firms are currently built to deliver it.
Rebuilding around value delivery isn’t a billing conversation. It’s a restructuring conversation. It means:
- Rewriting engagement letters and scope definitions from scratch
- Renegotiating partner compensation that currently rewards hours originated and hours billed
- Rebuilding associate development pipelines that currently run on repetitive work
- Repricing client relationships based on outcomes and risk, not time
That’s not a memo you send. That’s a multi-year transformation. And it has to happen while you’re still running the current model.
What Firms That Are Actually Moving Are Doing
A small number of firms aren’t waiting for the industry to converge on a new model. They’re testing alternatives now, matter by matter:
Fixed-fee engagements for defined scopes. This works well where the work is bounded — M&A due diligence, standard contract review, regulatory filings. The firm keeps the efficiency gain. The client gets cost certainty. The key is defining scope tightly enough that AI savings aren’t eaten by scope creep.
Subscription retainers for mid-market clients. One model gaining traction: a flat monthly fee that covers a defined set of legal needs, with the lawyer acting more like an embedded advisor than an outside vendor. AI makes this economically viable by dramatically cutting the hours required to serve each client. The lawyer who used to need 20 clients to hit revenue targets can now serve 20 clients in the time it used to take to serve 10.
Outcome-based fees on litigation. Contingency and partial-contingency arrangements aren’t new, but AI changes the risk calculus. If AI makes case assessment and document review dramatically faster and more accurate, firms can take on more matters on contingency without proportionally increasing risk, because the cost of being wrong on case selection is lower.
Hybrid billing with an AI transparency line item. A handful of progressive firms are experimenting with explicit AI cost disclosure in invoices, showing the client what the matter would have cost at traditional rates, what AI savings were passed through, and what the firm retained. This is the most honest model and probably the most client-relationship-preserving. But it requires a level of internal cost tracking most firms don’t currently have.
The Only Question That Matters Right Now
Bleich’s speech was aimed at GCs, not managing partners. The message was: we know you’re spending too much, here’s the intellectual cover to demand less.
The question for law firm leaders is different: before your biggest clients come to you with that demand, and they will, do you know what your actual economics look like in a post-billable-hour model?
Not philosophically. Not in a McKinsey deck. Concretely: if your top 10 matters moved to fixed fees today, using your real AI-adjusted hours, at a margin you could sustain, what does revenue look like?
Most firms don’t know. That’s the open loop. And the firms that solve it first will be, as Bleich put it, “leapfrogging” the ones that are still waiting for the industry to hand them an answer.