As you've probably already seen, something structural is breaking in law firm economics—and it's happening quietly.

An estimated 63% of large law firms expect AI to fundamentally alter their leverage model by 2035.

The traditional pyramid is collapsing into a "cylinder": thin at the bottom (fewer juniors hired), bloated in the middle (too many senior associates and counsel), thin at the top. Associate attrition has roughly doubled—somewhere between 16-17% left firms in the past 12 months compared to about 9% the prior year.

Here's what's driving this. Firms see AI eliminate 60-80% of junior work (document review, due diligence, first-draft research) and think, "We can hire fewer associates and improve margins." And they're right—in isolated roles, strategic headcount reduction makes complete sense. The challenge is when firms apply this logic broadly without considering which roles actually train future partners.

Short-term: margins improve. Medium-term (2031-2033): you have no one qualified to promote to partner.

As Gretta Rusanow from Citi Law Firm Group put it: "If AI eliminates the work that used to train juniors into strategic advisors, how do we cultivate the next generation of partners?"

Most firms don't have an answer yet.

Look, as I see it, every firm using AI faces a similar fork in the road.

Path 1: Use AI to cut junior headcount broadly → capture short-term margin gains → face a talent crisis in 7 years.

Path 2: Use AI to cut strategically and accelerate associate development → invest in structured training → build the most valuable partnership in a decade.

Here's what each path looks like and why Path 2 is the only viable long-term strategy.

Most firms are choosing Path 1—and for understandable reasons.

The logic is rational: AI does 60-80% of junior work. You reduce junior hiring or don't replace attrition. You save on salaries, benefits, overtime, and training costs. Margins improve immediately. And in many situations, this is the right move—particularly for roles that were purely administrative or highly repetitive.

The issue isn't cutting headcount. The issue is cutting broadly without asking: which roles are we eliminating that used to create our future partners?

Here's what actually happens in 5 years when you cut without redesigning training. You create what's being called a "missing middle." Your 2031 senior associates can't validate AI output properly, can't train others, and can't develop judgment. You become dependent on expensive lateral hires to fill partnership gaps while competitors who cut strategically and invested in training are building stronger pipelines.

The firms winning this transition aren't avoiding headcount reduction—they're redesigning what the remaining juniors do.

The core principle: AI eliminates drudge work → use that freed-up time to accelerate associate development, not reduce headcount across the board.

To be clear: some headcount reduction is smart. If you had three juniors doing document review and AI can handle that workload, you don't need all three. But the question becomes: what do you do with the one or two you keep? That's where most firms are getting it wrong.

What changes: Work shifts from repetitive tasks to high-judgment work—client communication, strategic analysis, validating AI output, cross-disciplinary collaboration. Structured training programs replace osmosis. The metric shifts from hours billed to skills developed per quarter. Many firms bring in specialists to design these programs because building them in-house without expertise often leads to half-measures that don't stick.

The outcome: You build AI-fluent partners who use AI as a force multiplier. And because you've cut strategically rather than broadly, your margins improve while your talent pipeline strengthens.

You don't need to redesign your entire training program overnight. Start here:

Step 1: Audit your talent pipeline. Look at your current associate cohorts and honestly assess whether they're building strategic judgment or just validating AI output. This takes about 20 minutes but reveals where your real gaps are.

Step 2: Redesign one work allocation. Pick one practice area. Ensure every junior gets minimum 40% high-judgment work. Common workflows we redesign for law firms include contract analysis and client intake—shifting juniors from basic review to strategic work.

Step 3: Pilot one AI-enabled training program in Q1. Choose billable credit for AI learning, a simulation program, or manual-before-automation. This is where we run hands-on workshops for legal teams—introducing them to AI tools they can start using the same day and teaching them how to automate parts of their own workflows. Measure results in Q2, scale what works in Q3.

A) We're reducing junior hiring and relying on AI to fill the gap
B) We're keeping headcount flat but haven't redesigned training yet
C) We're investing in structured training programs to accelerate development
D) We're still figuring out what to do
E) We're not sure AI will significantly impact our training model

[Reply with your letter choice] - I'll share the results in the next edition.

Here's what I’m thinking about this past week.

Most partners seem to see AI eliminating junior work and think—great, we'll hire fewer associates and capture the margin. And look, they're absolutely right about the margin. Strategically reducing headcount where AI genuinely replaces entire roles? That's good business.

But here's where I see firms making a mistake: they're applying that logic broadly, making significant cuts to junior associate programs without redesigning what the remaining associates actually do.

In 7 years you'll need to promote someone to equity partner. If your associates spent 2026-2033 validating AI instead of building judgment, you won't have anyone ready.

The firms that solve this in 2026-2027 will have a 5-year talent advantage. By the time competitors realize they have a pipeline crisis, you'll have built the most capable partnership in your practice area.

So the question isn't whether AI changes your staffing model. It will. The question is: are you cutting strategically and building the partners you'll need in 2033, or are you cutting broadly and borrowing profitability from your future self?

My opinion, I think you've got about 12 - 24 months to decide.

— Liam Barnes

Wondering which workflows in your firm should be automated first—and which need human judgment to develop your future partners?

Grab some time to chat

(if you don’t see a suitable time, just shoot me an email [email protected])

How Did We Do?

Your feedback shapes what comes next.
Let us know if this edition hit the mark or missed.

Too vague? Too detailed? Too long? Too Short? Too pink?

👍 Loved it | 👎 Not for me | 💡 I Have a Suggestion

OR: [email protected]

Was this week’s newsletter forwarded to you?

Sign up, it’s free.

Last Week’s Reader Poll Results