Automating Junior Roles Cuts the Leadership Pipeline

I keep seeing the same prediction: AI will automate the work done by junior staff, so organisations won't need to hire them.

It is framed as inevitable. Efficient. The logical next step.

I struggle with the logic.

Where future leaders come from

If we stop bringing young people into organisations, where do future leaders come from?

Not the work itself — AI can absorb a significant portion of the execution. But the context, the judgement, the sense of how decisions actually get made in practice: that develops through exposure. You cannot promote an algorithm into a senior role. You cannot ask a model to take accountability for culture, to navigate the politics of a difficult stakeholder relationship, or to make a call in the kind of ambiguous situation that doesn't have a prompt.

Those capabilities are built over time, through the accumulation of experience. Junior roles are how that accumulation starts.

What junior roles were actually for

Yes, some entry-level tasks will disappear. They probably should. No one built a meaningful career on copying data between systems or reformatting reports for distribution.

But junior roles were never just about the task.

They were how people learned how decisions get made. How trade-offs actually play out. What good looks like in practice — and what good enough looks like when perfect would take too long. How to read a room. How to push back without damaging a relationship. How to tell when a project is in trouble before the metrics confirm it.

That learning happens through proximity. Through watching people more experienced than you handle situations under pressure. Through being wrong in a low-stakes context and understanding why.

If we automate all of that away, we are not just cutting cost. We are cutting the pipeline.

The better question

The better question is not how to remove junior roles. It is how to redesign them.

In a world where the mechanical work gets absorbed by tools, the remaining human work should shift accordingly. Less execution, more judgement. Less producing everything, more shaping and challenging what the tools produce. Less doing the work, more understanding why it matters and whether it's the right work to be doing.

That is actually a more interesting version of an entry-level role. More exposure to the decisions that matter. Less time in the operational weeds. If organisations designed deliberately for it, early-career people could develop judgement faster, not slower.

The risk is that most organisations won't design deliberately for it. They'll just remove the headcount, discover the gap a few years later, and wonder where all the senior people went.

The short-term efficiency trap

Optimising for short-term efficiency is a reasonable impulse. Organisations face real cost pressure. The argument for smaller teams is not without merit.

But efficiency is not the only thing that compounds. Capability compounds too. The organisations that invest in building a pipeline of people — that keep humans in the loop not just for compliance reasons but because they understand what developing people actually requires — will have a meaningful advantage in five years that doesn't show up in this year's headcount numbers.

Automate, yes. But also augment. Keep humans in the loop, especially the ones who are still learning what the loop is for.

The alternative is an organisation that gets leaner and leaner until the people left have no one to hand things on to.