McKinsey & Company is bringing artificial intelligence into its hiring process. AI interviewers have appeared at technology companies for years, spawning a cottage industry of coaching on how to handle them, but their arrival at a consulting firm marks a departure from the norm.
The firm will introduce an "AI client simulation" this fall into interviews for business analyst roles at some of its North American offices, targeting university students, according to The Wall Street Journal on the 14th. Candidates spend about 10 minutes conversing on a laptop with an AI-generated "client avatar" while human interviewers observe. The interviewers do the evaluating; the AI plays no role in scoring. Hundreds of applicants are expected to take the test, and McKinsey plans to extend it to other roles next year if it proves successful.
Salim Zafar, the McKinsey partner who designed the AI client simulation, said the tool does not mimic "an unhinged client" or pose trick questions. Instead, he said, it reflects roughly 80% of the scenarios McKinsey consultants encounter in their daily work.
In one example, the mayor of a fictional city consults with a candidate on where to build a new stadium, with the AI avatar playing the mayor and peppering the consultant with questions. McKinsey believes AI can be fairer than humans in such role-playing exercises. When people take the role, difficulty can vary by interviewer, the firm says, whereas AI can hold it steady.
"What we watch closely is whether the candidate caught the important thing the client said, whether they acted on it or just ignored it and moved on, and whether they came in with their own playbook and tried to push their answer," Zafar said. "We are looking at how candidates respond to change."
Blair Ciesil, a McKinsey partner who co-leads global recruiting, said she expects the AI interview to reveal candidates' genuine qualities rather than rehearsed responses. "Sometimes candidates come in with a very well-practiced story. That doesn't mean it isn't a great story, but it makes it quite difficult for interviewers to get at who that person really is at their core," Ciesil said.
From Hiring to Firing, AI Seeps In
AI is working its way into personnel management and dismissals as well as recruitment. According to figures the Organisation for Economic Co-operation and Development released last year, 90% of U.S. managers surveyed said their companies had adopted at least one algorithmic management tool to direct, monitor or evaluate employees' work.
Controversy and pushback remain substantial. In California, a "robo boss" ban aimed at preventing unilateral AI-driven dismissals and discipline has passed and takes effect in July next year. It is the first U.S. law to require human review in firings and disciplinary actions that involve automated decision systems. Companies must notify employees in writing that AI is being used, and workers can demand an explanation of the personal data the AI used in reaching its judgment. State Sen. Jerry McNerney, who introduced the bill, said no worker should be fired or disciplined by a machine, and stressed that AI must remain a tool under human control.
In July, 26 Meta employees filed suit alleging the company used AI to select targets for job cuts and penalized workers who had taken legitimate medical or family care leave. Meta denied the allegations, saying humans made the final decisions.