Perspectives · Future of Work
Why Today's Graduates Can't Find the Bottom Rung
The corporate ladder isn't disappearing. Its shape is changing — and almost nobody has figured out what to do about the rung that's gone missing.
In April this year, at a panel discussion hosted by EDHEC Business School in London, an MBA student stood up and asked the room a question nobody could answer cleanly: How do you get to the middle of the career ladder when there are no junior jobs left to start on?
The panel — executives from Microsoft, QS, and a former McKinsey partner now running AI transformation work at Artefact — had spent the previous forty minutes describing, almost matter-of-factly, why that junior rung is vanishing. Stephen Bennett, the ex-McKinsey partner, put it the plainest:
"We will reduce the number of business analysts directly out of school, but we are not going to reduce the number of associates, we're not going to reduce the number of managers, because we will generate more work with AI. As a partner, I will have more teams working for me, but those teams will be smaller."
That sentence is worth sitting with, because it describes something more specific than "AI is taking jobs." It describes a shape. For decades, the standard organisational chart was a pyramid: a wide base of fresh graduates doing the grunt work, narrowing as people got promoted, until a handful sat at the top. AI doesn't flatten that pyramid — it pinches it in the middle of the base and pushes the weight upward, into a structure that increasingly looks like a diamond. A thin entry point. A fat middle of people who can direct, judge, and take responsibility for what the AI produces. A stable top.
If you want to see this in numbers rather than metaphors, look at what's happening to the firms that have always been the biggest first employers for business graduates. JPMorgan Chase, Citi, and Goldman Sachs together posted roughly $47 billion in first-quarter 2026 profits while cutting close to 15,000 jobs — a number large enough that the New York Times ran it as a standalone story on Wall Street's AI-driven job cuts. McKinsey is cutting an estimated 3,000 to 4,000 positions, about a tenth of its global headcount, concentrated precisely where Bennett said it would be: junior research and analyst work — part of a broader shift Bloomberg reported on in April, which also cited one widely circulated industry estimate putting the year-on-year drop in entry-level consulting inflows at 54%. Tellingly, Boston Consulting Group is one of the few firms actually growing headcount right now — but almost entirely by hiring AI and technology specialists, not the generalist MBA hire who used to walk in the door at twenty-four.
None of this means hiring has stopped. It means hiring has changed shape, and the shape it's changing into has no obvious door for someone with no experience to walk through.
India's Strange Contradiction
Nowhere is this more visible — or more confusing, if you only read the headlines — than in India. By the official numbers, Indian graduates are becoming more employable every year, not less. The India Skills Report's employability score climbed from 54.81% in 2025 to 56.35% in 2026, and more than nine in ten Indian employees now report using generative AI tools at work. On paper, this looks like a workforce getting sharper exactly on schedule.
On the ground, placement officers are describing the opposite experience. According to a 2026 investigation by The Quint, Devi Ahilya Vishwavidyalaya in Indore saw the number of companies visiting its campus rise from roughly 150 to over 250 — and the number of students actually placed fall from 1,744 to around 1,000, a 43% decline. More recruiters showed up; far fewer hires resulted. That is not a story about companies losing interest in campus recruitment. It's a story about companies showing up, looking harder, and being far choosier about who they take. Private engineering colleges across the country have reported declines of 50% to 70% in placement volumes over the same window. Even branded institutions aren't fully insulated — IIM Tiruchirappalli's count of unplaced students nearly doubled, from 34 in one recent batch to 59 in the next.
Meanwhile, India's IT services sector — historically the single largest absorber of fresh graduates, business and engineering alike — is seeing an estimated 20% to 25% decline in entry-level roles — an EY estimate cited in the same Quint reporting — even as the broader tech workforce grows at roughly 2.3% for the year, per NASSCOM. One comparison of task-level automation, cited in the same reporting, found that 37% of entry-level work tasks in India are already being performed by AI tools, slightly ahead of the 33% global figure — plausibly because the routine, repeatable, well-documented tasks that generative AI is best at are disproportionately the tasks Indian IT services has built its fresher pipeline around for twenty-five years.
Put the two halves of the story together and the picture sharpens: the skill bar Indian graduates are clearing keeps rising, but the number of doors willing to let them in is shrinking faster. Rising employability and falling placement rates aren't a contradiction. They're the same diamond-shaped structure showing up in a different country's data.
The Part That Should Worry Educators More Than the Layoffs
The job-cut numbers are the visible part of this story. The less visible part may matter more in the long run, and it has nothing to do with headcount.
Researchers at Wharton recently studied what they call "artificial cognition" — the habit of treating a generative AI tool as a default external judgment-maker rather than an occasional assistant. Their finding should give every business school pause: when the AI's answer happened to be correct, people's own reasoning accuracy on a task rose by 25%. But when the AI's answer was wrong, accuracy fell by 14%, because people followed the flawed output anyway rather than catching the error themselves. A separate survey across six business schools, presented at the same EDHEC event, found that roughly one in four undergraduates say they use AI specifically because they're afraid of falling behind their peers — not because it's the best tool for the task in front of them — and one in five say their reliance on AI is already affecting their confidence in their ability to do the work without it.
That's a feedback loop, not a one-time effect: lower confidence drives more reliance, more reliance crowds out the independent practice that used to build genuine skill, and the erosion compounds. A recent global workforce survey put a number on where that loop currently lands — only 19% of entry-level job seekers describe themselves as "very confident" about their careers. That is not a generation lacking ambition. It is a generation whose traditional way of building competence — junior-level reps, repeated under real stakes, with someone senior around to catch the mistakes — is being removed at the exact moment the tools available to them quietly undercut their own judgment.
What Doesn't Change, and Why That Matters
It would be easy to read all of this as an argument that business and management education has lost its relevance. The evidence says almost exactly the opposite. Employers across every source in this reporting — UK hiring panels, Indian placement officers, global workforce surveys — describe the same skills gap, and it isn't technical. It's critical thinking. It's communication under pressure. It's judgment, and the willingness to take responsibility for a decision an algorithm merely suggested. QS's chief executive, speaking at the same London event, put it simply: employers say the real gap between what graduates have and what employers need shows up in critical thinking, communication, and resilience — not in AI fluency, which is rapidly becoming table stakes rather than a differentiator.
The World Economic Forum's widely cited projection of 78 million net new jobs globally by 2030 is true and almost beside the point. Net job creation at the top of a diamond doesn't help someone standing at the bottom, looking for the entry point that used to exist. The honest version of the optimistic case is: there will be enough work, eventually, for people who can already exercise judgment — and a serious shortage of structured ways to become one of those people, because the apprenticeship model that built judgment for the last fifty years assumed a wide base of junior roles that is no longer there.
Nobody at that EDHEC panel had a confident answer for the student who asked how to reach the middle of a diamond with no junior rung to start from. Six weeks and an ocean away, an Indian placement officer was watching the same structural shift play out in a noisier, more chaotic form — more recruiters, fewer offers, scores rising, placements falling. They are describing the same problem from two different vantage points. Until business schools, employers, and policymakers build a deliberate, repeatable answer to that one question, every other statistic in this story is just commentary on the same unresolved gap.
Sources — EDHEC panel, the diamond thesis, and confidence research
Sources — Wall Street and consulting-industry cuts
Sources — India placements and employability