AI is testing Big Law’s leverage model, not ending it yet

The traditional Big Law model converts hierarchy into economic scale. Equity partners attract clients, exercise judgment and carry ultimate responsibility, but teams of associates perform substantial research, drafting, review and transaction work. When that work is billed by time, adding capable lawyers below each partner can expand revenue faster than the number of owners entitled to the residual profit.

AI changes the production function beneath that structure. Research, document comparison, due diligence and first-draft preparation are among the tasks that generative systems can accelerate. The immediate result need not be fewer lawyers. A firm facing strong demand can use the saved time to process more matters, shorten turnaround times or redirect associates toward higher-value work. In that scenario, technology reinforces leverage: partners still oversee teams, but each team produces more.

The pressure begins when efficiency reduces chargeable hours without producing equivalent additional demand. Under hourly billing, a task completed faster normally generates a smaller fee. ABA Formal Opinion 512 makes the ethical constraint explicit: a lawyer who agreed to bill by time cannot charge for hours that were not worked merely because AI made the task faster. (americanbar.org) A firm can respond by raising rates, winning more work or negotiating a price based on the matter’s value. Each response, however, depends on client acceptance.

This is why AI threatens the shape of leverage more directly than the existence of large firms. Citi’s 2026 advisory anticipates a movement away from the classic pyramid toward a more cylindrical structure, with fewer junior associates and a larger proportion of experienced lawyers performing strategic work. That is Citi’s forecast, not an observed industry-wide outcome. Its mechanism is nevertheless plausible: automating repetitive tasks weakens the commercial reason for maintaining a broad base of junior fee earners while increasing the value of lawyers who can verify outputs, understand clients and make defensible judgments. (citiglobalwealth.com)

The billable hour slows that transition because it remains familiar to firms, clients and procurement departments. Thomson Reuters reported that 90% of tracked legal dollars still used standard hourly arrangements, despite investment in tools intended to save time. Its report described clients and firms as caught in a pricing impasse: buyers want efficiency, but often compare proposed alternatives by converting them back into implied hourly rates. (thomsonreuters.com)

AI therefore creates two competing versions of leverage. The older version scales people and hours beneath partners. The emerging version scales expertise through software, standardized workflows and smaller teams. Established firms may operate both systems for years, applying conventional staffing to unpredictable, high-stakes matters while using technology-heavy delivery for repeatable work. The winners will not necessarily be the firms with the fewest associates. They will be those able to measure AI’s contribution, protect confidential information, verify the output and price the resulting service without making efficiency financially self-defeating.