Law firms are expanding aggressively in 2026, and the hard data is beginning to complicate the popular narrative about AI dismantling the traditional legal profession.
According to Cushman and Wakefield’s Legal Sector Leasing Trends report for Q2 2026, law firm leasing totalled 7.3 million square feet, a record-setting quarter that exceeded the previous quarter by 23%.
That figure also represents a 27% increase over the same quarter last year, with the first two quarters of 2026 running 17% higher than the equivalent period in 2025.
New leases in the first two quarters composed approximately 57% of all legal leasing activity, suggesting firms are not simply renewing existing space but actively seeking more of it.
The Cushman and Wakefield report concluded that “law firms remained focused on growth, with expansion activity reaching its highest share since 2019 and downsizing activity continuing to decline.”
The financial picture reinforces that conclusion, with revenues reportedly growing 12.4% in the first half of this year and demand increasing 4.8% over the same period.
Associate salaries and bonuses have surged, partner compensation can top $40 million, and jobs across the legal market continue to increase rather than contract.
The conventional wisdom held that AI would hollow out the associate ranks, shrink office footprints, kill the billable hour, and fundamentally restructure how law firms operate as businesses.
If AI were genuinely replacing the need for lawyers, the logic of basic economics suggests firms would be reducing square footage, cutting headcount, and watching billable hours and revenues decline rather than rise.
Stephen Embry, a lawyer, speaker, blogger, and writer who publishes TechLaw Crossroads, argues there are three possible explanations for the disconnect between prediction and reality.
The first is simply that the forecasters were wrong, partly because the AI disruption thesis assumes demand for legal services stays flat or falls, which is not guaranteed.
Embry points to the Jevons paradox as one counter-argument, suggesting that greater efficiency in legal work could actually stimulate more demand for legal services overall.
New categories of litigation, such as product liability claims against technology companies over harm to younger users, illustrate how entirely novel legal questions can emerge and fuel fresh demand.
The second explanation is the enduring resilience of law firms themselves, institutions with a long track record of absorbing supposedly disruptive technology without allowing it to upend their core business model.
The billable hour has been declared dead repeatedly across decades of legal commentary, yet it continues to dominate how firms price and sell their services to clients.
Embry notes that e-discovery, cloud computing, data analytics, and remote working were each predicted to transform legal practice fundamentally, and in each case law firms adapted without structural collapse.
The third possibility is simply a matter of timing, that the disruption predicted is real but has not yet arrived, and that firms are still in the early stages of absorbing AI into their workflows.
There is also evidence that firms are buying AI tools without meaningfully deploying them, creating a gap between investment and genuine operational change that could delay any visible impact on headcount or revenue.
Risk aversion runs deep across the profession, both inside law firms and among in-house counsel at large companies, which tends to slow the adoption of any technology that threatens existing ways of working.
As Embry cautions, borrowing from Jimmy Buffett’s song “Mañana”: “there’s a danger in describing an ocean when we haven’t seen it.”