McKinsey is piloting AI client avatars in entry-level interviews

What's going on? McKinsey & CompanyDealroom has a profile for this one. Try Dealroom → is testing a roughly 10-minute conversation with an AI simulation of a client as part of interviews for some undergraduate business-analyst candidates in North America this fall, according to reporting by the Wall Street Journal relayed by The Verge. Hundreds of candidates may sit the exercise; humans stay in the room and score performance — the avatar does not.

How it works: The simulation is meant to cover about 80% of everyday client situations consultants face, and is not designed to mimic an "unhinged" client. Partners or assessors watch how candidates listen, adapt and present recommendations — including scenarios such as pitching to a simulated mayor. The exercise sits alongside existing steps such as personal-experience interviews and McKinsey's Solve assessment; the firm has also used its internal AI assistant Lilli in some final-round cases since January 2026.

Context: Entry-level business-analyst roles at McKinsey are among the most competitive openings in consulting. Other employers are separately experimenting with AI avatars in hiring, but few have put a simulated client into a brand-name professional-services interview loop with human assessors still deciding the outcome.

Why it matters: It is a clear signal that generative AI is moving from back-office tools into the gatekeeping rituals of elite white-collar hiring — testing how candidates handle a client, not only how they use an assistant.

What's not settled: The pilot is limited to some North American offices. McKinsey has not publicly set success metrics for expanding it to other roles in 2027, and press coverage does not detail how performance with the avatar will be compared with candidates who never see it.

Sources: The Verge (Jay Peters, 14 Oct 2026), citing the Wall Street Journal; New York Post (Emily Lang, 14 Oct 2026)