McKinsey now wants entry-level recruits to talk their way through a fake client crisis - with an AI playing the client.
The firm is testing a 10-minute interview segment where undergraduate candidates angling for business-analyst roles spar with an AI simulation of a client, according to the Wall Street Journal. The simulation is built to mirror roughly 80% of the situations a McKinsey consultant actually runs into on the job, not some worst-case meltdown scenario. The AI itself doesn't grade anyone. Human interviewers sit in the room and judge how candidates handle the simulated back-and-forth. The tool is currently in testing at a handful of offices, not a firm-wide rollout.
This matters because it signals how consulting firms plan to hire for an AI-saturated job market: they're not just testing whether candidates can use AI tools, they're testing whether candidates can out-think one in a live conversation. If junior analysts are expected to manage AI-assisted workflows on the job, screening for that skill at the interview stage is a logical next step, even if it sounds like a gimmick on paper.
Plenty of companies have added coding tests and case studies to interviews over the decades. This is the same instinct, just pointed at a different simulated opponent - and it says more about what entry-level consulting work is turning into than about the AI doing the talking.