Cutting entry-level roles is the easiest AI saving. It quietly starves your leadership pipeline.

AI can now handle a large share of the work you used to give your most junior people, so the logic writes itself: if the software does the graduate-level tasks, stop hiring graduates. It looks like a clean saving. The tasks are automatable, the headcount is junior, and the budget line is right there to cut. But entry-level roles were never just cheap task execution. They are how your organisation manufactures judgment, institutional knowledge and its future leaders. Remove the bottom rung and the saving shows up this quarter, while the hole shows up in five years, when the senior people that rung would have produced simply are not there.

The scale is already real. A 2026 World Economic Forum report, AI and the Future of Entry-Level Work, finds that 37% of young workers globally are employed in occupations with medium-to-high exposure to AI-driven task change. Its central warning is one most cost-cutting misses: entry-level roles have long been the foundation for building the institutional knowledge, judgment and tacit skills that underpin long-term organisational capability. Automate the entry rung purely to save money and you are not just shedding junior labour, you are breaking the pipeline that produces your future specialists, managers and leaders. The risk is not only to young workers. It is to the organisations that stop investing in them.

Why is cutting entry-level roles more dangerous than it looks?

Because you are eating your seed corn. Entry-level work is where people absorb the things no manual contains and no model will hand you: how the business actually works, how to exercise judgment, how to sense when something is off before it becomes a problem. That tacit learning happens by doing the junior work, being close to it, and being corrected. Delete the rung and you do not merely lose cheap execution, you lose the only place that grows your senior people. The damage is invisible today, because nothing breaks, and severe later, because the bench is empty exactly when you need to promote from it.

So should you keep hiring juniors to do work AI can now do?

No, and that is the false binary. The choice is not "hire juniors for automatable tasks" or "stop hiring juniors." It is to redesign the entry-level role around what actually builds capability. The WEF report frames it exactly this way: rather than simply automating early-career tasks, some organisations are redesigning junior roles so they continue to build judgment, oversight and future leadership rather than just execute tasks. The concrete examples are instructive. IBM is tripling its US entry-level hiring in 2026 while rebuilding junior roles around judgment, oversight and customer engagement. One banking organisation redesigned its junior analyst roles so that new analysts challenge AI outputs for incorrect assumptions, missing data and logical flaws. In both cases the junior's job becomes checking, questioning and steering the machine, which is precisely the apprenticeship that forms a capable senior.

How do I make the call deliberately?

By treating early-career hiring as strategic workforce planning, not a line to trim.

1. Decide it where you plan your leadership bench. Entry-level hiring shapes who runs the company in a decade, so make the call at that altitude, not in a quarterly cost review.
2. Redesign the role, do not delete it. Point AI at the routine tasks and rebuild the junior job around judgment, oversight and challenging AI output.
3. Make juniors the humans who check the machine. The apprenticeship that used to come from doing the task now comes from supervising and questioning it. Design that in on purpose.
4. Protect the pipeline explicitly. Track whether you are still building future managers and specialists, not only whether this year's costs fell.
5. Rebuild onboarding for the new shape. Onboarding that assumed years of grunt work needs redesigning for a role that starts much closer to oversight and judgment.

What does this change for me as a leader?

It reframes the cut from a cost decision into a capability decision. The seductive logic, AI does the junior work so stop hiring juniors, optimises the one number you can see and quietly destroys the asset you cannot. The leaders who get this right will use AI to redesign the bottom rung rather than remove it, so early-career people spend less time on rote execution and more time learning to exercise and check judgment. That way you take the efficiency and keep the pipeline.

This is the part of the end of business as usual that hits slowest and hurts longest: the organisation that hollows out its entry level looks leaner for a few years and then discovers it has no one ready to lead. Judgment is the thing AI does not give you, and it is your genuine advantage, in line with the truth that your advantage is what is distinctly yours, not the tools everyone can buy. The apprenticeship that builds that judgment is worth protecting, precisely when AI makes it easiest to cut.

About the author

Thomas Green

British technology futurist, AI keynote speaker and advisor. Thirty years across enterprise technology and AI strategy, helping leaders navigate the future of work. The futurist who died.