Developing Future Leaders When AI Does the Entry-Level Work

I keep coming back to one question as AI takes over more entry-level work:

How will people learn to become managers and leaders?

The tasks we're automating — drafting, summarizing, first-pass analysis, slide prep, research synthesis — were never just "junior work." They were how people learned. It's how early-career employees absorbed context, understood quality standards, saw how senior people think, and slowly built judgment.

So my concern isn't only that AI replaces tasks. It's that if we automate them without redesigning the learning path, we quietly remove the rungs people used to climb.

One study captures the risk. In research with Boston Consulting Group consultants, people completed professional tasks with and without generative AI. On most tasks, AI helped — they worked faster and produced stronger results. But on a task that required deeper judgment, where the answer wasn't obvious and AI was less reliable, the consultants using AI did worse than those working without it. Not because they were less capable — because they trusted the tool too much.

That's the leadership-development risk in miniature. If early-career employees never get enough practice judging the work, questioning assumptions, and recognizing when AI is out of its depth, they become faster — but not wiser.

I don't think the answer is to protect every old task. Work will change; it already is. But the development path can't disappear. Organizations need to be far more intentional about exposure, mentorship, feedback, and real participation in decisions. Junior people still need to sit in the hard conversations, see how senior people reason through ambiguity, and understand not just the answer but why it's the answer.

AI may produce the first draft. But humans still need to learn how to judge it, question it, improve it, and take responsibility for it.

My caution to leaders: don't let efficiency quietly hollow out your leadership pipeline. If AI frees up junior capacity and we just add more volume, we save money now and pay for it later — when there aren't enough people ready to lead.

Efficiency should create more room to develop people, not replace the process of developing them.

I've seen this in my own work too. The first drafts, summaries, and slide structures were not just outputs — they were how I learned what mattered, what was weak, and how to tell the difference between information and insight.

I'd love to hear from HR and talent leaders: how are you thinking about developing future managers when AI automates the traditional entry-level work?