AI Is Cutting Consultants. It Is Also Hiring Them.

The same technology is shrinking consulting headcount and growing consulting demand at once. Which side of that split a firm lands on is the whole question for the next few years.

Everyone agrees AI is coming for consulting. Almost no one has noticed it is also the industry's biggest new customer.

First, the part that deserves to be said plainly, before anything else. This has been a hard year in consulting. Accenture is cutting roughly eleven thousand roles, with leadership saying openly that people it cannot reskill will be exited. McKinsey is making its largest cut since 2008, several thousand roles. Cognizant is trimming thousands more under a restructuring it calls Project Leap, built on the premise that AI-augmented teams can deliver what large workforces once required. In most of these announcements, AI is named as the reason. These are real people, and the firms making these calls are not doing it carelessly. They are responding to genuine pressure.

But if you only read those headlines, you will draw the wrong conclusion about where this is going. Because at the exact moment the big firms are cutting consultants, the biggest technology companies in the world are spending billions to hire them.

The paradox nobody is putting together

In July, Microsoft committed 2.5 billion dollars to a new unit that embeds 6,000 engineers and industry experts inside client organizations. Amazon put a billion dollars behind the same model two days earlier. OpenAI stood up a deployment arm and a partner program aimed at certifying 300,000 consultants. These are the companies building the AI that is supposedly making consultants obsolete, and they are the ones now paying, at scale, to put consultants in the room.

Read those two facts together and the story stops being AI versus consulting. The services market is not shrinking. It is sorting. Demand for the right human, deployed to the right problem, is going up. Demand for undifferentiated hours is going down. The same technology is doing both, and it is doing them at the same time.

Most firms are cutting the wrong variable

When margin comes under pressure, headcount is the lever every leader can reach. It is visible, it is fast, and it shows up cleanly on the next quarter's numbers. So that is the lever they pull. The problem is that headcount is a blunt instrument for a precision problem. Cutting people reduces cost, but it does nothing to make the people who remain easier to deploy, and deployment is where the value has quietly moved.

Here is the mechanism, and it is worth being exact about it. When you bill by the hour, deploying the right expert faster actually works against you, because efficiency means fewer billable hours. Speed is punished. But the industry is moving off hours. More than thirty percent of McKinsey's fees are already tied to client outcomes rather than time, and the internal projections at firms like Deloitte have hours-based work shrinking toward a sliver of the market over the next decade. Once you are paid for the outcome instead of the time, the equation flips. Deploying the right person to the right problem faster is no longer lost revenue. It is pure margin.

That single shift, from hours to outcomes, is what makes deployment speed the decisive capability of the AI era. The firm that fields the right team first does not just win the work. Under outcome pricing, it keeps more of what the work is worth.

Fewer consultants is not the strategy

The firms that struggle over the next few years will be the ones who treated headcount as the only lever, who got smaller without getting faster, and who will find that a leaner firm that still cannot deploy its people quickly is just a smaller version of the same problem.

The firms that win will be the ones that got better at aiming. That can put the exact right expert on the exact right problem before a competitor can even staff it. That treat their people not as a cost to be trimmed but as capacity to be deployed, precisely and fast, into the work where AI has made expert judgment more valuable, not less.

AI is not making expertise worthless. It is making idle expertise unaffordable.

The proof is already on the table. The companies that build AI for a living are the ones paying most to put humans back in the room. They understand something the panic misses: as the routine work automates, the value concentrates in the judgment, the relationships, and the right person showing up at the right moment. That has not been devalued by AI. It has been made scarce, and scarce is the opposite of worthless.

The layoffs are real, and so is the pressure behind them. But fewer consultants was never the strategy. Better aimed ones is. The firms that understand the difference are the ones that will still be standing, and growing, when the sorting is done.

Deploy the people you have, faster. See how quickly Lumiere turns a client requirement into a staffed, client-ready team on the systems you already run. Book a 20-minute demo

A note on the numbers: figures here are drawn from reporting by CNBC, Reuters, Bloomberg, the Wall Street Journal, and industry trackers, and reflect reported figures, not audited results. Directionally consistent across sources, but treat them as reported.