For two decades, the pitch a young analyst heard at McKinsey was simple: you will work brutal hours, you will build the models and the decks, and in exchange you will get the pedigree that opens every door afterward. The deck-building was the tuition. In 2026 the firm decided it no longer needs quite so many people paying it, and the reason it gives is the one it has spent the last two years selling to everyone else.
The consultant who told you to automate
McKinsey is cutting roughly 10% of its global workforce — an estimated 3,000 to 4,000 roles — with the reductions concentrated in back-office functions, junior research, and practice areas where generative AI has collapsed how long an engagement takes to deliver. The HR Digest reports the cuts as part of a multi-year reshaping under global managing partner Bob Sternfels, who had already signaled that non-client roles would shrink as the firm leaned into its own AI tooling.
The irony writes itself, so we may as well name it plainly. McKinsey’s most lucrative recent product line has been telling other companies how to reorganize themselves around artificial intelligence — which teams to shrink, which processes to automate, how to justify it to the board. The slide that recommends cutting a client’s analyst layer looks exactly like the slide someone at McKinsey just built about McKinsey.
What actually got automated
Strip away the framing and the mechanism is unremarkable, which is the point. A traditional consulting engagement ran on a pyramid: one partner, a couple of engagement managers, and a wide base of analysts who pulled the data, ran the interviews, cleaned the spreadsheets, and turned it all into a hundred-page deck. That base was where the hours lived, and the hours were the bill.
Generative AI eats the base first. Tasks that used to occupy an analyst team for two weeks — synthesizing research, drafting the first version of a deck, building the model scaffolding — now come back in an afternoon, per Quartz, which frames the McKinsey cuts as a signal that white-collar reductions are spreading past tech into the professional-services core. When the work that justified a headcount shrinks, the headcount follows. This is not AI doing something magical. It is AI doing the boring 60% that a firm was quietly charging premium rates for.
Not just McKinsey, and not just prestige
The tell is that this is not one firm’s stumble. Fast Company calls the McKinsey cuts a warning signal for consulting as a category, noting that Bain, BCG, and Deloitte have all trimmed headcount or slowed hiring into the same market. When the entire top tier of an industry adjusts in the same direction at the same time, it is not a management error — it is the cost structure changing underneath all of them.
For anyone building a career, the McKinsey number is worth more as a data point than as gossip. The consulting analyst job was one of the most reliable on-ramps in the white-collar economy: high pay, transferable skills, a name that traveled. If AI can thin that role at the firm that practically invented it, the comfortable assumption that a prestigious employer is a safe employer needs retiring. The protective thing was never the logo on the business card. It was being harder to automate than the person next to you — and the analyst pyramid was, it turns out, exactly the shape AI climbs fastest.
The firms will call this a productivity story, and in the narrow sense they are right. Fewer people can now deliver the same engagement. The part the productivity slide leaves off is that “the same engagement, with fewer people” is just a layoff wearing a forward-looking caption. McKinsey knows that. It has been drawing that slide for its clients for two years.
