Why Your Mid-Level Leaders Are Quietly Avoiding AI. And How to Change the Narrative.

"The leaders who look most resistant to AI are usually the ones who feel they have the most to lose. Change what they believe they are losing, and the resistance follows."

I have been in a version of the same conversation with senior leaders across Asia for the last six months. The shape of it is remarkably consistent.

The CEO is committed. The board is asking. The strategy is written. The tools are bought. And then they hit the wall.

The wall is not technology or budget. It is the next line of leadership. The senior managers, the department heads, the functional directors. The people who are supposed to lead the adoption are quietly avoiding the thing they are meant to champion.

Not openly, and not with resistance. With something harder to see and harder to address. They attend the training. They nod in the town halls. They sign off on the roadmap, and then they do not use it. Not genuinely. Not in a way that changes how their team operates.

When you sit with them one on one, the truth comes out. They are afraid. Not of the tool. Of what the tool means for them, for the expertise they spent fifteen years building, and for the role they have been told defines their value.

And here is the part most senior leaders miss. That fear is rational. Not because AI is coming for their job. Because no one has helped them answer the question that sits underneath all of it. If the machine can do parts of what made me valuable, what exactly am I for now?

You cannot logic someone out of that question, and you cannot training-session your way through it. You change the narrative they are telling themselves. That is the work.

Here are three narrative shifts that move mid-level leaders from quiet avoidance to genuine ownership.

Shift One: Replace "Will AI replace me?" with "What does AI free me to be better at?"

The question every mid-level leader is silently asking is not really about AI. It is about identity. They have built a career on being the person who knows, who analyses, who produces the first draft everyone else works from. AI does those things now. Not perfectly, but well enough that the question lands.

Most organisations try to answer the fear with reassurance. "AI won't replace you, it will augment you." That sentence has been said so many times it has lost all meaning. Your mid-level leaders hear it and dismiss it, because they have heard it enough to know it is a script.

The shift that actually works is reframing the question itself. Instead of answering "will I be replaced?", help them ask a better question. What does AI free me to do that I could not do before? What parts of my role were never the real value, and what parts were?

Most senior managers I coach, when they sit with this honestly, discover something uncomfortable. A significant portion of their week was spent on work that was never the reason they were promoted. The reports. The summaries. The first-draft thinking. The coordination. AI does not take their job. It takes the scaffolding that was never the point of their job.

What is left is the actual work. The judgment. The coaching of their team. The stakeholder relationships that cannot be automated. The difficult conversations no tool can have for them. The strategic interpretation that requires someone to own the consequences.

That is the narrative to build. Not "AI is safe." But "AI gives you back the job you were actually hired to do." The mid-level leader who hears that and believes it stops defending and starts designing.

Shift Two: Lead by Example Before You Ask Them to Lead by Example

This is the one I see most senior leaders get wrong, and it is the one that costs the most.

You ask your mid-level leaders to champion AI adoption. To model it for their teams. To lead by example. And many of them will, if they see you doing it first.

The problem is that most senior leaders are not doing it. They are sponsoring and funding it, but they are not personally using it in any way that is visible to the layer below them.

Your mid-level leaders notice. They may not say it, but they notice. And what they conclude is simple. If the person above me is not willing to change how they work, why should I? The ask lands as a directive, not an invitation. And directives from a position of "do as I say, not as I do" produce compliance, not conviction.

The shift is to make your own use visible. Not as a performance, as a practice. Share a prompt you ran this week and what it surfaced. Show the team the first draft AI produced and what you changed in it. Admit where it was wrong and what that taught you. Talk about the decision you made faster because the tool removed the first-hour work.

This does two things that matter. It normalises the practice. And it models the thing you actually want from them, which is not "use the tool." It is "be willing to be a beginner again in front of the people you lead."

Mid-level leaders are not avoiding AI because they cannot learn it. They are avoiding it because learning it means being visibly bad at something in front of their team. The senior leader who models that discomfort removes the permission problem before it starts.

Shift Three: Build a Culture of Experimentation Where "What Went Wrong" Becomes "What Did We Learn"

The first two shifts change the individual narrative. This one changes the environment that holds the narrative in place.

Most organisations say they want experimentation. Most do not actually tolerate it. The signal is in what happens when something goes wrong.

When a mid-level leader pilots an AI workflow and it produces a bad result, what happens next? Is the conversation about what went wrong, who is responsible, and why we should be more careful? Or is it about what we learned, what we would do differently, and what it tells us about where the tool helps and where it does not?

The first response teaches avoidance. The second response teaches ownership. And your mid-level leaders are reading which response they will get before they ever pilot anything.

This is where most AI adoption stalls. Not because people are unwilling to experiment. Because they have learned, through years of experience, that experiments that fail are punished rather than mined. So they wait. They let someone else go first. They adopt the tool in the smallest, safest way possible and call it progress.

The shift is to change the post-mortem question. From "what went wrong?" to "what did we learn?" That is not a semantic change, it is a cultural one. It changes what people bring to you, when they bring it, and how much risk they are willing to take.

A leader I coached recently did this with her leadership team. She asked each of them to run one AI experiment in their function within thirty days. The only requirement was that they had to share one thing they learned, and the learning had to include something that did not work. She was explicit. If you only share successes, you have not done the exercise.

The conversations that came out of that cycle were more valuable than any training program. Leaders shared where the tool helped and where it failed, and what it made them reconsider about their team's work. And because the frame was learning, not performance, they brought things they would have hidden in any other context.

That is what experimentation actually looks like. Not a slogan on a wall. A leader who makes it safe to report what did not work, and who treats that report as the most valuable data the organisation has.

The Leader Who Owns the Narrative

The mid-level leaders in your organisation are not the obstacle to AI adoption. They are the test of whether you have given them a narrative worth adopting.

If the story they are telling themselves is "AI is coming for my job, and my job is my identity," they will avoid it, quietly, for as long as they can.

If you change that story to "AI gives you back the work you were actually hired to do, I am doing it with you, and we learn out loud together," the avoidance turns to ownership. Not instantly, but reliably.

The leaders who get their organisations through this transition are not the ones with the most sophisticated AI strategy. They are the ones who understand that adoption is a narrative problem before it is a technology problem. And narratives are built by leaders, not purchased by procurement.

Change the story. The adoption follows.

---

If you are a senior leader watching your mid-level layer quietly avoid AI and want to think through how to shift the narrative in your organisation, I would welcome that conversation. Book a complimentary strategy call here.

---

Adeline Tiah is an ICF PCC Certified Executive Coach based in Asia, with 28 years of corporate leadership experience across global MNCs. She specialises in leadership transformation, high-performing teams, and culture change. She is the author of Reinvent 4.0, a guide to leading with purpose in an age of disruption.