Did AI Really Cause Baker McKenzie's Layoffs? What Actually Happened
Baker McKenzie cut hundreds of business-services roles and cited AI. The real story involves margin pressure, business model economics, and a convenient headline.
On 6 February 2026, Baker McKenzie confirmed it was reducing its global business-services team, and the firm’s own spokesperson pointed to “rethinking the ways in which we work, including through our use of AI.” Within days, the story had become shorthand across the legal press for “AI is coming for law firm jobs.” The reporting on the cuts themselves is largely accurate. The causal story built on top of it is not, and it is worth separating the two.
What Was Actually Confirmed
Reports differ slightly on the precise number. Legal Cheek put the figure at under 10 percent of the firm’s global business-services headcount, roughly 600 roles. Above the Law reported “a little more than 700 people” on the same basis. Baker McKenzie has not confirmed an exact total.
The roles affected span business professionals, not lawyers: IT, knowledge management, administration, diversity and inclusion, learning and leadership, secretarial support, marketing, and design functions, across multiple offices.
The firm described the cuts as proposals, subject to consultation in applicable jurisdictions, with some roles phased out and others changed in scope rather than eliminated outright.
The firm’s stated rationale was to “position the firm for continued growth” and remain agile, citing its use of AI, efficiency gains, and a redirection of investment toward roles that serve clients directly.
This was not an isolated move. Legal Cheek reported in November 2025 that Clifford Chance planned to cut roughly 50 London support roles and change about 35 more, citing AI alongside falling demand for certain services. Freshfields and Irwin Mitchell have made similar cuts to paralegal roles.
None of that is in dispute. What is doing a lot of unexamined work in the coverage is the leap from “the firm mentioned AI” to “AI caused this.”
Why the AI-as-Single-Cause Story Does Not Hold Up
First, no lawyer roles were cut. Every account of the layoffs is consistent on this point. The reduction touched business-services and support functions, which is exactly where document-heavy, process-driven AI tools currently add the most value, and exactly the layer of a firm’s cost base that comes under pressure whenever revenue growth slows. That is a real effect of AI tooling. It is not evidence that AI is displacing legal judgement.
Second, the firm’s underlying economics were already under strain independently of AI. Across the Global 100 rankings, revenue per lawyer fell roughly 3 percent even as gross revenue rose about 2 percent, a divergence the American Lawyer described as “the real story” behind the headline numbers. Baker McKenzie operates one of the largest, most geographically dispersed practices in the world, a volume-and-scale model that behaves very differently from smaller, premium-positioned firms. By way of comparison, Wachtell’s profit per equity partner sits at roughly $4.01 million against Sullivan & Cromwell’s $2.94 million, figures that illustrate how differently firms at the top end of the market are built, and why a scale-based firm under margin pressure has cost-cutting reasons that have nothing to do with AI.
Third, an industry recruiter said out loud what the framing is doing. Scott Love, president of The Attorney Search Group, told Above the Law plainly: “If they use AI to justify it, they won’t look as cruel when they make that many staff cuts.” Whatever the actual mix of causes behind the decision, citing AI is also a public-relations choice, and other firms are reportedly watching Baker McKenzie’s reception closely as cover for their own cuts.
The Salesforce Parallel Is Instructive
Baker McKenzie is not the first large organisation to have an “AI replaced our workforce” story outrun the facts. In early February 2026, reports claimed Salesforce had replaced roughly 4,000 customer-support roles with its Agentforce AI agents. Salesforce disputed this directly, describing the change instead as a “strategic rebalancing” built from redeployment, unfilled vacancies, and natural attrition, not a wholesale swap of people for software. Critics had also pointed to reliability problems, including one client whose AI agent failed to trigger customer satisfaction surveys, a gap Salesforce fixed with what it called “deterministic triggers” rather than treating it as evidence the agent could not be trusted.
The lesson is not that either company is lying. It is that “AI did this” is currently the most convenient, least scrutinised explanation available to any organisation making a workforce decision, and legal journalism has been no more sceptical of it than tech journalism was.
What This Actually Means for Law Firms and Lawyers
The realistic risk here is not that a large language model takes over a lawyer’s work. It is that:
- Support-function headcount will keep shrinking at firms with high-volume, document-heavy business services, as AI tools genuinely do reduce the labour needed for that layer of work.
- Margin-pressured, scale-based firms will keep using AI as the explanation of choice for workforce decisions that are substantially about cost discipline, because it is a more palatable headline than “demand growth has not kept pace with headcount.”
- The competitive risk to lawyers themselves is indirect. Client work moves toward leaner teams and AI-fluent firms, or in-house departments that have built their own verification and governance capability, not because an algorithm replaced a lawyer’s judgement, but because better-run teams are winning the work.
That last point is the one worth building a firm-wide response around, and it is exactly what our Productivity & Practice Management module addresses: how to deploy AI tools in a way that genuinely improves throughput and client service without the governance gaps that turn “we use AI” into a liability story six months later.
Baker McKenzie’s cuts are real. The business-model pressure behind them is real. The claim that AI alone caused them is the part of the story that was never actually checked.