McKinsey will make job applicants negotiate with an AI client

By Elena Vasquez

McKinsey & Company will ask some entry-level job candidates to hold a 10-minute conversation with an AI simulation of a client, according to The Wall Street Journal. The simulation, which the firm is rolling out as part of its interview process, puts applicants in a scenario consulting firms usually reserve for on-the-job training: a difficult stakeholder with demands, pushback, and limited patience.

The WSJ reports that the AI "aims to reflect 80% of the scenarios McKinsey consultants might encounter in their day-to-day work." That figure is the anchor of the whole experiment. McKinsey isn't building a stress test or a trick. The paper notes the simulation isn't designed to "mimic an unhinged client."

Humans, not the AI, will do the scoring. Assessors will be in the room to watch how candidates handle the simulated situations. The AI plays the client. People judge the performance.

What is McKinsey actually testing?

The tool is in testing for undergraduates applying to business-analyst roles, in what the WSJ describes as "some" offices. That phrasing matters. This is a pilot, not a company-wide mandate — at least not yet.

The logic behind it is straightforward. Entry-level consultants spend a large share of their first years managing client conversations: gathering requirements, defending recommendations, absorbing friction. A traditional case interview tests structured thinking. It does not test whether a 22-year-old can stay composed while a client pushes back in real time.

The AI simulation is McKinsey's attempt to close that gap before hiring day rather than after it. By modeling 80% of day-to-day consulting scenarios, per the WSJ, the firm is trying to compress the most common parts of the job into a 10-minute interview exercise.

Who scores the candidates?

The AI does not evaluate anyone. According to the WSJ, humans in the room assess the candidates and observe how well they deal with the simulated situations.

That detail separates McKinsey's approach from the more controversial end of AI hiring. Many companies now use automated systems to screen candidates, score video interviews, or rank applicants without a human in the loop. McKinsey's setup keeps the judgment with people. The AI is a prop — a sophisticated one, but a prop.

It also changes what the exercise measures. A human assessor watching in real time can read composure, improvisation, and empathy in ways no scoring model reliably can. The simulation supplies the pressure. The humans supply the verdict.

Why does this matter for the hiring market?

McKinsey is one of the most-watched employers in graduate recruiting. The firm receives enormous volumes of applications for its business-analyst positions, and its process is closely imitated across the consulting industry and beyond.

That gives the pilot outsized weight. If a firm of McKinsey's stature decides AI-simulated clients belong in interviews, competitors and large corporate recruiters tend to follow. Ten-minute AI role-play could become as standard a hiring fixture as the case interview itself.

The move also lands at a moment when AI is reshaping both sides of the hiring table. Candidates already use AI tools to draft applications and rehearse interviews. Employers are answering with AI-driven screening of their own. McKinsey's version is notable because it uses AI to test a specifically human skill — managing a live, difficult conversation — rather than to automate judgment about the candidate.

What are the limits of the simulation?

The WSJ's reporting frames a deliberate boundary: the tool targets 80% of real scenarios, not all of them. And it won't "mimic an unhinged client." In other words, McKinsey has calibrated the simulation toward the routine, not the extreme.

That choice keeps the exercise fair and predictable. It also narrows what it can measure. The hardest client moments — the ones that derail projects and burn out young consultants — often sit outside the tidy 80%. A simulation built to avoid the unhinged may miss exactly the scenarios that distinguish great consultants from adequate ones.

There's also the testing question. The pilot covers undergrads seeking business-analyst jobs in only "some" offices, per the WSJ. Whether McKinsey expands it to experienced hires, other regions, or its full entry-level pipeline will say a lot about how well the first cohorts of candidates actually performed.

What comes next?

For now, this is a bounded experiment: a 10-minute AI conversation, human assessors, some offices, undergraduate applicants. But the direction is clear. If the pilot holds up, expect the AI client to become a fixture of consulting interviews — and expect every major firm that competes with McKinsey for the same graduates to build one of its own.