McKinsey plans to add a 10-minute conversation with a simulated client to interviews for some entry-level consulting candidates. The test will be used this fall for undergraduates seeking business-analyst roles in selected North American offices, with hundreds of applicants expected to encounter it.
The artificial client is designed to reproduce common workplace interactions rather than generate an extreme or hostile personality. McKinsey says the scenarios are intended to reflect about 80% of situations its consultants could encounter in day-to-day work.
Human interviewers will remain in the room and assess how candidates handle the exchange. The AI system will not assign a score or make the hiring decision. That distinction limits the pilot’s automation: the model supplies a repeatable conversational situation, while people interpret the applicant’s response.
The exercise could let candidates demonstrate communication and judgment in a more interactive setting than a static case prompt. It also introduces questions about whether the simulation behaves consistently across applicants and whether its scenarios reflect the work fairly. Because the tool is still being tested in only some offices, the rollout does not yet represent a company-wide replacement for McKinsey’s existing interview process.