In January 2026, Bob Sternfels - McKinsey's global managing partner - appeared at the Consumer Electronics Show in Las Vegas and said something that should have made every business leader put down their phone. "How big is McKinsey? How many people do you employ? My latest answer would be 60,000 - but it's 40,000 humans and 20,000 agents."

He then updated the number. As of that week, the agent count had reached 25,000. And his stated goal: parity. Equal numbers of AI agents and human employees by the end of 2026.

The world's most prestigious consulting firm - the one that has shaped how global corporations think about strategy, structure, and the future of work for over a century - just told you exactly where business is heading. Most people filed it under "interesting" and moved on. They shouldn't have.

What 25,000 agents actually means

These are not chatbots answering FAQs. McKinsey's agents handle entire job functions independently. They generated 2.5 million charts in six months. They saved 1.5 million hours last year on search and synthesis tasks alone.

McKinsey's QuantumBlack division - 1,700 people driving all of its AI work - now accounts for 40% of the firm's total output. Think about that ratio. The world's most prestigious consulting firm has restructured nearly half its business around AI in less than two years. Not as a side project. Not as a pilot programme. As its core operating model.

And Sternfels is not done. His next goal: every single human employee paired with at least one agent within 18 months.

"In another 18 months I think every employee will be enabled by one or more agents. We'll have a workforce that is human and agentic, and we're going to have to navigate that." - Bob Sternfels, McKinsey CEO

The business model change nobody is talking about

The agent story is dramatic. But buried inside it is something more consequential for every business leader: McKinsey is changing how it charges.

For decades, McKinsey sold advice by the hour. The fee-for-service model that built the entire consulting industry. That model is ending. Sternfels is moving McKinsey toward outcomes-based pricing - identifying a business goal with the client upfront, then tying McKinsey's fees to whether that goal is actually achieved. Not hours delivered. Not recommendations produced. Outcomes.

This shift is only possible because autonomous agents change the economics of delivery fundamentally. When software handles the research, the synthesis, the analysis, and the reporting - the cost of producing an outcome drops dramatically. And when the cost drops, you can afford to bet on the outcome rather than bill for the process.

Every business that relies on a service model should be watching this closely. The consulting industry just showed you what happens when autonomous software meets professional services. Your industry is next.

What McKinsey is hiring for instead

Here is where it gets interesting for anyone thinking about talent. McKinsey is now incorporating AI into its graduate recruitment process itself. Candidates face an AI-led interview stage - tasked with using McKinsey's internal AI tool as a "thinking partner" to solve business scenarios. The evaluation focuses not on technical AI expertise, but on judgment and reasoning.

Separately, McKinsey is shifting its graduate hiring toward liberal arts majors. Not engineers. Not data scientists. People who can think clearly, communicate precisely, and exercise judgment in ambiguous situations.

The message is unambiguous. The skills that AI cannot replicate - creativity, judgment, genuine human insight - are becoming more valuable, not less. The skills that consist primarily of processing information and producing structured outputs are being absorbed by the agents.

"For young professionals coming into the workforce: focus on honing skills that AI can't do. Human judgment and true creativity." - Bob Sternfels, McKinsey CEO

The number that puts it in perspective

McKinsey went from a few thousand agents to 25,000 in eighteen months. Gartner projects that 40% of enterprise applications will embed AI agents by end of 2026, up from less than 5% in 2025. That is an eight-fold increase in a single year - and McKinsey is already living at the far end of that curve.

Amazon cut 16,000 corporate roles this year citing a shift to AI-driven automation. BCG, PwC, Deloitte, and EY are all racing to match McKinsey's model. The pattern is not isolated. It is systemic.

What this means if you lead an organisation

Most leaders are still asking the wrong question. They are asking how to add AI to what they already do. McKinsey is asking something different. It asked: if we had access to an unlimited number of tireless, highly capable digital workers, what would our organisation look like? And then it built that organisation.

The answer involved fewer human layers handling information and more human judgment applied to decisions that genuinely require it. It involved a business model redesigned around outcomes rather than process. And it involved a hiring strategy that explicitly prioritises the capabilities software cannot replicate.

You do not need to be McKinsey to ask the same question. You do need to ask it. Because your competitors are.

By the numbers: 25,000 AI agents at McKinsey as of January 2026, alongside 40,000 human employees. Goal: parity by end of year. Key idea: McKinsey is not adding AI to its existing model. It is building a new model around AI - and changing how it prices, hires, and delivers as a result. That is the part most people missed.