McKinsey is piloting a shake-up in how it recruits its next generation, asking graduate candidates to use an AI assistant to complete tests designed to reflect consultants’ new ways of working.
The elite consulting firm has begun asking candidates to use its AI tool Lilli as part of its notoriously taxing tests for business school graduates, according to people familiar with the matter.
Candidates that were part of the pilot were asked to use the chatbot during one of their interviews to replicate how the firm expects its consultants to work as the technology transforms office jobs.
The candidates used Lilli to help analyse a case study and refine their conclusion, one person said. The interview tested how applicants prompted Lilli and whether they had the “curiosity and judgment” to “take stuff that Lilli spits [out] and work with it, challenge it, put it into context of the client’s specific requirements”, they added.
The move underlines how AI-driven disruption is reaching even one of the world’s most competitive recruitment processes at a firm that has long been a training ground for chief executives including Alphabet’s Sundar Pichai, Jane Fraser at Citigroup and Lloyds Banking Group’s Charlie Nunn.
McKinsey would roll out the test to all junior recruits in the coming months if the pilot proved successful, the people said, adding that it was just one evaluation rather than a pass-or-fail exercise.
Mayank Gupta, chief executive of CaseBasix, which trains McKinsey hopefuls for the rigorous interview process, said other prestigious consulting firms such as Boston Consulting Group and Bain were also likely to incorporate AI into the interview process.
Many consulting firms are buying or building AI expertise as they spend less time on traditional strategy advice and more on helping companies adopt the technology.
Last week Accenture agreed to buy Faculty in a $1 billion deal to improve its AI abilities and to install the start-up’s chief executive on its own global management committee.
Major consulting firms could be forced to shift away from their traditional pyramid set-up, which uses a relatively small number of senior consultants to oversee armies of junior analysts.
McKinsey encouraged its weakest-performing consultants to quit in 2024 and shrank its workforce by more than a tenth between 2023 and the middle of last year, amid a consulting industry slowdown, after hitting a peak of 45,000 staff.
It is also planning more job cuts, in part to reflect efficiencies from AI, according to a person familiar with the matter. It reportedly set a target of axing 10 per cent of non-client facing roles over the next two years – potentially more than 1,000 jobs.
At the same time, McKinsey has increased its internal use of “AI agents”. Chief executive Bob Sternfels told a podcast this month that the firm had a “workforce” of 20,000 agents on top of its 40,000 staff. That figure would grow over the next 18 months to “get to one agent per human”, he added.
“We’re migrating pretty quickly away from . . . pure advisory work . . . to much more of an outcomes-based model,” he said on Harvard Business Review’s IdeaCast. “The stuff that I did when I joined as an associate 32 years ago, we wouldn’t consider even doing right now. Why? Because clients do that stuff themselves . . . the imperative will then be to move to the even more complicated questions.”
McKinsey has also changed multiple biases in its recruitment processes to adapt to the skills it needs in the AI age and will prioritise candidates who learn from failure, Sternfels said.
Candidates with liberal arts degrees “might have been deprioritised in the past” but they have more “truly novel” ways of thinking that will complement some AI models’ inability to make “discontinuous leaps” in logic, Sternfels added.
McKinsey has long been a first mover to respond to changes in consulting work in its hiring practices. In 2018 it led the transition to gamified problem-solving tests over the traditional exam-style format.
McKinsey did not respond to a request for comment. –Copyright The Financial Times Limited 2026