Gen Z Can Hear What Leaders Won’t Say About AI

My favorite aspect of being a professor was the gift of working with young people.

Not because they were naïve—they weren’t. In fact, the best students I taught could be brutally poignant about what was broken. They could spot contradiction quickly, challenge inherited logic, and ask the question many in the room had learned to swallow.

At its core, humanities education trains young people to think critically about the world around them. What made the college classroom magical, for me, was that students were not yet fully jaded by years of institutional compromise, incremental progress, and the kind of professional realism that teaches people to lower their expectations and call it “maturity.” They critiqued the world because they still believed broken things could be fixed.

That kind of critique does not always translate neatly into the workplace. Gen Z workers are often framed as entitled, difficult to manage, anxious about AI, resistant to innovation, or unprepared for the future. Recent coverage has described young workers as openly rebelling against AI at work, and other reports have pointed to job anxiety among students and early-career workers as AI disrupts entry-level roles.

With those looming anxieties in mind, when recent graduates laugh, boo, or go quiet as commencement speakers tell them to embrace AI, we must hear more than generational defiance. We must hear a public version of what often shows up more quietly inside organizations.

Rumbi Petrozzello and I talk about this pattern often. We see it in organizations navigating high-stakes change: Leaders announce a future, employees hear the trade-offs, and the gap between the two becomes the place where trust starts to thin.

Gen Z has real reasons to question the story they are being sold around AI.

They are graduating into debt, instability, and a labor market already being reorganized around automation. They have spent years being told to build skills, find their edge, and stay adaptable. Now, at the ceremonial threshold between education and work, they are hearing leaders celebrate the technology that may reshape the first decade of their careers before they have even had a chance to begin them.

That reaction deserves more than an eye roll. For CHROs and senior leaders, Gen Z’s response is an early signal.

The Risk Is Bigger than Job Displacement

The easiest way to explain Gen Z’s reaction is job anxiety. Anyone graduating into this labor market would be foolish not to wonder what AI means for their first job, earning power, and career prospects.

However, displacement is only the most visible fear. The deeper risk is what happens when AI starts to absorb the very work that teaches people how to become skilled.

Entry-level work has always included tasks that look inefficient from a distance: drafting the first version, sitting in meetings where you understand only half of the conversation, shadowing someone who is more experienced, making low-risk mistakes, revising after feedback, learning which questions matter and which details are noise, and—most important—failing. And then learning how to be better.

That work can look repetitive, slow, and ripe for automation. It is also how people gain the critical skill of judgment.

This is where the “just upskill” argument breaks down. We keep telling young workers AI will not replace them if they learn to use it. But that assumes the workplace still gives them room to develop expertise. If AI handles what once was the training grounds, the next generation may be asked to become experts without apprenticeship.

That should worry every CHRO. A company can move faster and actually become less productive.

A Future of De-Skilled and Weakened Expertise

The early research related to AI and skill development should make leaders pause.

Anthropic’s research on AI coding assistance found that junior developers using AI support while learning an unfamiliar coding library showed weaker understanding afterward, particularly in areas such as debugging and code comprehension. Yes, AI can help people work, but speed and learning are not the same thing.

A worker can complete a task faster and—at the same time—have a low level of understanding about the underlying process.

Organizations are investing heavily in AI with the promise of productivity. But productivity gains that depend on thinner judgment and diminished apprenticeship will create a different cost later. The resulting output, which may initially appear to be an efficiency win, may look fine until it needs to be explained, defended, corrected, adapted, or taught to someone else. This is how de-skilling can hide in plain sight.

Complicating matters further, here’s the risk that scares Rumbi and me most: Today’s entry-level work becomes tomorrow’s leadership pipeline crisis.

Messy first drafts become strategic discernment. Low-risk mistakes become the ability to catch high-risk ones before they escalate into big problems. Early repetition becomes managerial discernment and good judgment.

Automate too much of that away, and the loss may not show up immediately, but … it likely will show up down the road. It may surface years later, when fewer people are ready to lead and more managers are reviewing work they do not fully trust.

From Human in the Loop to Human in the Lead

Rumbi has been pushing those of us at Seramount toward a stronger frame: human in the lead. The phrase has been gaining traction in AI leadership conversations, including Accenture and AICPA & CIMA.

Distinct from the now common “human in the loop”—in which people may be positioned as a final check on a system already moving at high speed—“human in the lead” emphasizes people retaining authority over judgment, pace, accountability, and the design of work itself.

Our current research at Seramount urges business leaders to decide where automation strengthens the work and where it weakens the people doing it. It means protecting the developmental repetitions that build judgment. It means asking which tasks should be faster, which should remain relational, and which should stay slow enough to allow time for skill development.

It also means telling the truth about trade-offs.

If AI is being used to reduce headcount, say so. If entry-level roles are changing, explain how people will still learn, adapt, and evolve. If managers are expected to absorb quality control, coaching, emotional labor, and AI adoption all at once, name the load they are carrying. If the organization wants to be trusted, it cannot communicate as though employees are too unsophisticated to understand what is happening.

Inside organizations, employees may not boo at a company town hall. However, they may nod, complete the training, open the tool, and quietly wonder whether the future being described has room for them.

Before You Scale AI, Sharpen the Axe

In one of our conversations, Rumbi reminded me of an old story about two people chopping trees.

One keeps swinging without stopping. The other pauses throughout the day. By evening, the one who paused has cut more. The difference was not effort. It was the willingness to stop long enough to sharpen the axe.

Many organizations are in an AI rush that treats pausing as failure: They do not have time to map the work. They do not have time to understand where trust is thinning. They do not have time to ask what young workers are learning, what managers are carrying, or what expertise may be disappearing behind a temporary productivity gain.

But that pause is not a mistake. It is the work that determines whether AI becomes a force multiplier or another expensive transformation that leaves people overwhelmed, underdeveloped, and quietly less able to execute.

Gen Z may be hearing the dull blade before the rest of us are willing to admit it.