The Last Word: Not every graduate role will disappear. The issue is not the end of the junior career, but the need for firms to redesign it.
For years, the first rung of professional life was built from work nobody especially loved. Junior bankers checked models. Consultants cleaned slides. Lawyers reviewed documents. Graduates wrote first drafts, chased details and learnt, slowly, that competence is not the same as confidence. It was imperfect, often inefficient and sometimes needlessly punishing. But it did one useful thing: it turned raw talent into judgement.
This is not quite the AI jobs apocalypse. Bank of America is welcoming nearly 4,000 summer interns and full-time campus recruits this year, drawing talent from more than 500 colleges and universities. Reuters framed the decision against growing concern that generative AI can now handle the complex, data-heavy tasks once handed to junior banking staff. Goldman Sachs, meanwhile, is taking a more cautious line: David Solomon has said the bank’s entry-level hiring may “contract a little” as AI changes the talent mix, while rejecting the more theatrical idea of mass extinction for young professionals.
The numbers matter. But the better question is not whether every graduate role will disappear. It is whether the work that used to teach graduates how to be useful is being quietly stripped away.
Professional apprenticeship was never especially elegant. Junior bankers built models at absurd hours. Consultants tidied research, cleaned charts and learnt how senior people framed messy problems. Lawyers reviewed documents line by line. Journalists wrote briefs no one praised. Much of it was repetitive. Some of it was inefficient. A little of it was character-building in the way that delayed trains are character-building: mostly by accident.
Yet the dull work did something important. It exposed novices to the raw material of judgement. You learnt what a bad assumption looked like because you had made one. You learnt how numbers could be technically correct and commercially meaningless. You learnt that a client’s hesitation in a meeting often mattered more than the third decimal place in the spreadsheet.
New routes to competence
AI now promises to remove the drudgery. Good. Nobody should mourn unnecessary toil. But friction was also part of formation. If a machine produces the first draft, first model, first scan and first summary, the junior employee may be spared the grind without being given a new route to competence.
Nor is AI acting alone. The New York Fed has argued that remote work has left younger workers sidelined, estimating that it can explain 64 per cent of the recent increase in unemployment among young college graduates. Its point is not anti-flexibility; it is pro-formation. Early-career workers learn through proximity: the overheard call, the quick correction, the manager who says, “Not like that — try this.” Remove the office and automate the task, and the first rung becomes very slippery indeed.
There is also a widening gap between education and employment. Deloitte Asia-Pacific CEO Rob Hillard has warned that graduates are arriving at work with the idea that using AI is cheating, while employers increasingly expect fluency with it. That is a curious settlement: universities police yesterday’s definition of originality while companies hire for tomorrow’s collaboration.
The young are not necessarily frightened. A PwC and World Economic Forum briefing found entry-level workers more curious and excited than worried about AI, although many doubt how long their current skills will remain relevant. Strada’s employer research adds an important counterweight: senior talent leaders are 2.7 times more likely to expect AI to increase entry-level hiring in 2026 rather than decrease it. The issue, then, is not the end of the junior career. It is the redesign of it.
Future-relevant firms will stop treating apprenticeships as a by-product of cheap labour. They will make it deliberate. Juniors should be taught how to interrogate AI outputs, not merely generate them. They should get earlier exposure to clients, structured feedback, judgement drills and chances to make small mistakes before they become expensive ones. Managers, too, will need to relearn the craft of teaching.
AI can produce answers. Apprenticeship produces professionals. The companies that win will not be those that automate junior work fastest, but those that reinvent the first rung before the next generation has nothing left to practice on.
Photo: Dreamstime.