For decades, the US law firm ran on a pyramid. A wide base of junior associates handled high volumes of routine work — document review, research, first-draft contracts — billed to clients by the hour. A narrower band of mid-levels supervised them. A small group of partners sat at the top. The pyramid was not just a staffing model; it was a talent-development engine. The routine work at the base trained associates into the judgment and depth that, over five to seven years, produced the next generation of partners.
That pyramid is compressing, and the numbers are no longer subtle. According to the Thomson Reuters Institute Law Firm Financial Index, the average number of new first-year associates at Am Law 100 firms fell nearly 17 percent in a single recent fall hiring cycle. Among Am Law 200 firms, first-year class sizes dropped an average of 25 percent. NALP reported that median 2L summer program offers hit the lowest level on record, and that at the largest firms, summer class sizes fell to their smallest since 2012. In February 2026, Baker McKenzie cut 600 to 1,000 business-services roles citing AI integration — the largest AI-attributed cut in the legal industry to date.
This is not a temporary dip driven by an economic cycle. It is a structural shift, and AI is the primary force behind it. The work that justified hiring large junior classes is exactly the work AI now does fastest. As of March 2026, roughly 70 percent of attorneys reported using AI weekly. When first-pass document review and research can be done faster and cheaper than a junior associate can deliver them, the economic logic of the wide pyramid base erodes.
For firm leaders, this raises a question that is easy to defer and expensive to ignore: if you hire fewer juniors, and AI does the work that used to train the juniors you do hire, where do your future partners come from? The firms that answer this deliberately will still have a partnership in 2040. The firms that let the pyramid compress without redesigning how they develop talent are quietly dismantling their own future.
This guide explains why associate classes are shrinking, what the 2026 data actually shows, and — most importantly for firm leadership — what the compression means for your talent pipeline and what to do about it. SRA has designed and run performance, development, and engagement programs exclusively for US law firms since 1987, which puts us squarely in the middle of the question the pyramid’s compression forces: how do you develop and retain a smaller cohort exceptionally well?
Why associate classes are shrinking: four converging forces
The contraction is not driven by AI alone, though AI is the largest factor. Four forces are converging, and understanding all four matters for any firm trying to respond.
1. AI is removing the work that justified large junior classes. This is the dominant factor. First-pass document review, legal research, and routine drafting — the high-volume work that filled junior timesheets and justified large entry classes — is what AI handles best. As that work is automated, the economic case for a wide pyramid base weakens. Firms simply need fewer juniors to handle the same matter volume.
2. Clients are refusing to pay for AI-replaceable work. General counsel are increasingly unwilling to pay $900-an-hour associate rates for tasks a machine can do for a fraction of the cost. As clients scrutinize bills and route AI-replaceable work away from outside counsel, the revenue that supported large junior classes shrinks. The Thomson Reuters 2026 State of the US Legal Market documented this tension directly.
3. Firms are favoring laterals over new graduates. With juniors less able to be productive quickly through routine work, some firms are prioritizing experienced lateral associates who can contribute strategically from day one, bypassing the development curve. Lateral partner hiring surged even as associate hiring contracted — a sign firms are buying proven contribution rather than building it.
4. The recruiting timeline itself is breaking. Separately from AI, the recruiting calendar has accelerated to the point of dysfunction — firms locking in 2028 associate classes during the 1L panic cycle, before students have meaningful grades or experience. Some firms (Cooley, Susman Godfrey) are deliberately slowing down, hiring only part of their classes early and leaving seats open to fill later. This adds uncertainty to class-size planning on top of the AI-driven contraction.
Together these forces are reshaping not just how many associates firms hire, but who they hire, when, and what they expect those associates to do. The wide base of the pyramid is narrowing, and it is not coming back to its old width.
What the 2026 data shows
The contraction is well-documented across multiple independent sources. The table below pulls the key figures together.
Two caveats worth noting. First, not every firm is contracting — some, particularly in litigation-heavy and trial practices (Morgan & Morgan is one public example), are expanding their associate programs, betting that courtroom skills are less automatable. Second, a portion of the current contraction overlaps with a smaller graduating cohort, so the effect on overall employment rates is partially masked for now. But the structural direction — a narrowing pyramid base driven by AI — is consistent across every major source.
What the compression means for your talent pipeline
Here is the part firm leaders most need to think through, and the part the hiring data alone does not surface. A narrower pyramid base has consequences that compound over years, and most of them are invisible until it is too late to fix them cheaply.
Fewer juniors means fewer future partners — eventually. Partners come from associates. If a firm hires 25 percent fewer first-years and loses some of those to attrition, the cohort from which it will draw partners in a decade is materially smaller. The firms that do not plan for this will face a partner gap in the 2030s that is being created by hiring decisions made now.
The juniors you do hire develop differently — or don’t. Even for the smaller class a firm hires, AI has removed the routine work that historically built judgment and depth. If a firm does not deliberately redesign how those associates develop — through supervised AI use, earlier substantive responsibility, and structured feedback — it will end up with mid-levels who are thinner on judgment than previous cohorts. A smaller class developed poorly is worse than a larger class developed well.
Retention matters far more when the cohort is small. When a firm hired a wide base, losing a few associates to attrition was absorbable — there were others. When the base is narrow, every departure is a larger percentage of the future partner pool. Losing one strong third-year from a class of twelve is a far bigger problem than losing one from a class of forty. Retention shifts from a cost-management concern to a strategic survival concern.
The development you do provide becomes a competitive advantage. In a market where most firms are contracting and under-investing in developing their smaller cohorts, the firms that develop their juniors exceptionally well will attract and keep the best of a shrinking pool. Development quality becomes a recruiting and retention differentiator in a way it never was when everyone hired wide and trained through volume.
The strategic reframe: in the wide-pyramid era, talent was a numbers game — hire a lot, work them hard, keep the ones who survive. In the compressed-pyramid era, talent is a development game — hire fewer, develop each one deliberately, and retain them because losing any one is expensive. The firms that make this shift will own the next generation of legal talent.
Developing and retaining a smaller cohort exceptionally well
This is precisely where SRA works. As the pyramid compresses, the firms that win are the ones that get the people side right — measuring associate development accurately, surfacing and addressing the issues that drive attrition, and building the structured feedback and engagement systems that turn a smaller cohort into a strong partnership pipeline.
We do not do legal recruiting and we do not sell AI tools. We design and run the performance review, evaluation, and engagement programs that determine whether the associates a firm does hire develop into the partners it needs. In a contracting market, that work is no longer a nice-to-have. It is how a firm protects its future.
→ Talk to SRA about developing and retaining your associate pipeline → Explore SRA’s review and engagement programs
What US law firms should do about it
The compression is a structural market force no individual firm can reverse. But how a firm responds to it is entirely within its control, and the responses that work share a common logic: stop relying on volume to develop talent, and start developing it deliberately.
The Ropes & Gray approach — making AI exploration billable so associates build fluency as part of their development — is one concrete example of priority 1. The Law360 guidance that summer programs should teach students “how to think about using AI” rather than just the tools themselves points the same direction. The throughline across all six priorities: deliberate development replaces development-by-volume. The firms that internalize this will be fine. The firms that wait for the old pyramid to come back will not.
The bigger picture: where will tomorrow’s lawyers come from?
There is a question larger than any single firm’s pipeline, and firm leaders should at least be aware of it. The compression of first-year hiring across the industry raises a profession-wide concern: if entry-level legal jobs contract structurally, the system that has historically turned law graduates into experienced lawyers — and experienced lawyers into the judges, regulators, and senior counsel the legal system depends on — comes under strain.
Individual firms cannot solve a profession-wide problem. But individual firms can decide whether they will be part of the solution or simply optimize for the next quarter. Firms that maintain a deliberate commitment to developing junior talent — even at a smaller scale — are investing in both their own future partnership and the broader health of the profession they depend on. Firms that treat juniors as a cost to be minimized are, collectively, eroding the foundation the whole system stands on.
For a firm leader, the practical takeaway is narrower and more immediate: the contraction is real, AI is driving it, and the firms that respond by developing a smaller cohort exceptionally well will have a partnership in twenty years. That is a choice being made now, in this year’s hiring and development decisions, whether firms are conscious of making it or not.
Frequently asked questions
Are all US law firms shrinking their associate classes? No. The contraction is most pronounced at large firms — Am Law 100 and 200 — and in practice areas where routine work is most automatable. Some firms, particularly in litigation and trial practice, are maintaining or expanding their programs, betting that courtroom and advocacy skills are less automatable. But the overall structural direction across the industry is a narrowing entry-level pyramid.
Is AI the only reason associate classes are shrinking? It is the dominant reason but not the only one. AI removing routine junior work is the largest factor. Client pushback on paying for AI-replaceable work, a shift toward lateral hiring, and a dysfunctional recruiting timeline all contribute. AI is the structural driver; the others are reinforcing forces.
If we hire fewer juniors, do we still need formal performance and development programs? More than ever. A smaller cohort raises the stakes on every individual associate — you cannot afford to develop them poorly or lose them unexpectedly. The wide-pyramid model could absorb weak development and attrition through sheer numbers. The compressed pyramid cannot. Formal, deliberate development and retention programs become more important as the cohort shrinks, not less.
Where will our future partners come from if we hire fewer first-years? From the smaller cohort you do hire — if you develop and retain them well — supplemented by lateral hiring. The risk is a partner gap in the 2030s for firms that contract hiring now without intensifying development. The firms planning succession against the smaller cohort today will have a partnership; the ones that simply cut and hope will face a shortage of partner-ready lawyers in a decade.
How does the shrinking pyramid affect associate retention? It makes retention dramatically more important. In a wide pyramid, losing a few associates was absorbable. In a narrow one, each departure is a larger percentage of your future partner pool. A strong third-year leaving a class of twelve is a strategic loss in a way it never was leaving a class of forty. This is why engagement and retention programs move from cost-management to strategic priority.
How should we measure associates differently in this environment? Shift away from billable hours and toward judgment, AI-output validation, substantive depth, and AI fluency. The smaller the cohort, the more important it is to assess each associate accurately on what actually predicts partner potential. We cover the new measurement framework in
How Should US Law Firms Measure Associate Performance Now That AI Has Broken the Billable Hour?.
Sources
- Jordan Furlong (2026). Where will tomorrow’s lawyers come from? (citing Thomson Reuters Law Firm Financial Index, NALP, Decipher, Firm Prospects). jordanfurlong.substack.com
- LSAC / NALP (January 2026). NALP: Accelerating Recruiting Trends Amidst Market Slowdown. lsac.org
- JD Journal (March 2026). Law Firms Cut Summer Associate Hiring to Record Low as Recruiting Timeline Moves Earlier. jdjournal.com
- The Agency Recruiting (May 2026). 2026 Legal Hiring Trends: AI Impact on Law Firm Staffing (Baker McKenzie cuts; 70% weekly AI use). theagencyrecruiting.com
- Above the Law (February 2026). Top Biglaw Firm Pumps The Brakes On Early Recruiting (Cooley). abovethelaw.com
- Law360 Pulse (May 2026). The 2026 Summer Associates Survey. law360.com
- Thomson Reuters Institute & Georgetown Law (January 2026). 2026 Report on the State of the US Legal Market. abovethelaw.com
Related reading on srahq.com
- → How Should US Law Firms Measure Associate Performance Now That AI Has Broken the Billable Hour?
- → Legora vs. Harvey AI: What US Law Firms Should Actually Know in 2026
- → Partner Performance Review: How US Law Firms Evaluate Equity Partners in 2026
- → Which Employee Engagement Software Should US Law Firms Actually Use in 2026?
- → What Is the Difference Between a Performance Evaluation and a Performance Review at a US Law Firm?
- → How Should US Law Firms Separate the Coaching Conversation from the Performance Review Record?
The pyramid that built the modern US law firm is compressing, and AI is the reason. Fewer juniors, less routine work to train them on, and higher stakes on every associate a firm does hire. The firms that respond by developing a smaller cohort exceptionally well will still have a partnership in twenty years. The firms that simply cut and hope will not.
SRA designs and runs confidential performance reviews, evaluations, upward review programs, 360-degree feedback, and firm engagement surveys exclusively for US law firms. As the talent pipeline narrows, the work of developing and retaining the associates you do hire becomes the work that protects your firm’s future. Built for US law firms since 1987.
Performance Reviews | Upward Reviews | 360-Degree Feedback | Firm Engagement Survey | Schedule a Consultation
Exclusively serving United States law firms since 1987.
.jpg)
.jpg)
.png){kind=small}