AI Is a Change Leadership Problem First

Most organizations are solving for the technology. The real work is human, and until leaders engage with that, adoption won't happen.

Keith Laplante

Thirty-five years leading inside organizations (TD Waterhouse, BMO, HomeEquity Bank) and advising alongside them (Accenture, ExperiencePoint, The Refinery Leadership Partners). Facilitated leadership programs at Harvard, Duke, and Stanford. Read more about Keith

"AI is not a technology problem for most organizations. It's a people problem wearing a technology costume."

I want to start with an observation that I think most leaders instinctively know but haven't quite let themselves say out loud.

AI is not a technology problem for most organizations. It's a people problem wearing a technology costume. And until leadership recognizes that and starts leading accordingly, most AI initiatives will continue to produce the disappointing results they're producing today.

I've been in a lot of conversations about AI adoption over the past two years. Boardrooms, leadership teams, workshops, one-on-ones with senior executives trying to figure out what to do. And the pattern I keep seeing is the same. Leaders talk about the tools. They talk about the vendors. They talk about the training programs. They talk about the transformation roadmap and the change management workstream. What they mostly don't talk about, at least not directly, is the actual thing that's driving how their people are responding.

Their people are afraid.

Not all of them. Not visibly. Not in ways that would show up in an engagement survey. But if you sit with anyone in any organization long enough and talk honestly about AI, you'll eventually get to the question that lives just underneath the surface of every training rollout and every pilot program.

Am I going to lose my job to this?

Until leaders engage with that question directly, nothing else they do about AI is really going to work.

The hammer and the carpenter

There's an analogy I've come to use in this conversation that I think reframes it usefully.

A hammer is a powerful tool. In the hands of a skilled carpenter, it can help build a house. In the hands of someone who doesn't know what they're doing, it's dangerous and unproductive. The hammer isn't the story. The carpenter is.

AI is the hammer. And most organizations are spending enormous amounts of time evaluating hammers, comparing hammers, procuring hammers, training people to hold hammers, and building governance around hammers. What they're spending very little time on is the carpenter. Who they are. What they're worried about. What they need to feel safe experimenting with a new tool that some part of them suspects might make them redundant.

The technology is important. But the technology is the easy part. Any organization with a reasonable budget can acquire the tools. What separates the organizations getting real value from AI from the ones getting mediocre results is what they're doing with the people who have to use it.

The fear leaders aren't addressing

I want to be direct about this fear, because I think leaders are systematically underestimating how much it's shaping what their people do.

When a person is told to adopt a new tool that might make their role redundant, they have two choices. They can lean in, learn the tool, get good at using it, and hope that positions them well for whatever comes next. Or they can go slow, resist quietly, and hope that if the tool underdelivers, their existing role stays intact.

The rational calculation depends entirely on what they believe about the intent behind the adoption. If they believe their leadership is genuinely investing in them, that the tool is being introduced to make their work more valuable rather than to eliminate it, they'll lean in. If they believe, even suspect, that the real agenda is cost reduction and headcount, they'll go slow.

Most organizations have not answered this question for their people. They've talked about AI as an opportunity. They've talked about productivity gains and competitive advantage. They've talked about the transformation of work. What they mostly haven't said, clearly and specifically, is what all of this means for the person sitting in front of them.

And in the absence of a clear answer, people assume the worst. That's not paranoia. It's pattern recognition. Most technology transformations in the past decade have, in fact, resulted in workforce reductions somewhere in the organization. Your people know that. They're watching what leadership does far more carefully than what leadership says, and they're making their own judgments about what's coming.

What leaders who are doing this well look like

The organizations getting real traction with AI adoption look, in my experience, quite different from the ones that aren't.

They talk about it honestly. They don't pretend AI won't change jobs. They engage directly with the question of what it means for the people doing those jobs today, and they say more than platitudes. They talk about specific commitments to retraining, redeployment, or in some cases, honest conversations about roles that are genuinely at risk and what the organization will do to support the people in them.

They think bigger than efficiency. Efficiency is the easy story with AI, and it's also the least interesting one. The leaders who are getting the most out of AI are asking a different question. Not, how do we do the same work with fewer people, but what could we do that we've never been able to do before, if we had this capability alongside our people. That's a fundamentally different framing, and it produces fundamentally different results.

They involve the people who do the work in figuring out how AI fits into it. The best AI use cases in any organization are almost never the ones designed by the executive team or the consultants. They're the ones designed by the people who actually do the work, because those people know where the friction is, what the real bottlenecks are, and what would make their jobs meaningfully better. Organizations that engage those people directly in shaping AI adoption get better use cases, faster adoption, and far less resistance.

They protect the space for experimentation. Learning a new tool takes time. Getting good at it requires making mistakes. Organizations that are still measuring people primarily on short-term output metrics are, in effect, punishing the exact behaviour they need if they want AI adoption to succeed. The leaders who understand this create explicit space for their people to learn, experiment, and get temporarily slower before they get faster.

The change leadership foundation

None of what I've just described is really about AI. It's about change leadership.

The organizations doing AI well are the ones that were already good at change leadership, or that have quickly realized they need to become good at it. They understand that people who feel threatened will not fully engage. They understand that unstated fears drive behaviour more powerfully than stated opportunities. They understand that adoption of anything meaningful requires involvement, not just communication. And they understand that the human side of a technology transformation is where most of the value is either captured or lost.

If you're a leader trying to figure out why your AI initiative isn't producing the results you'd hoped for, I'd suggest the answer probably isn't in the tool selection or the training curriculum. It's in the change leadership you're doing, or not doing, around it.

Are you addressing the fear that's shaping your people's response? Are you being honest about what AI actually means for the workforce? Are you involving the people who do the work in shaping how the tools get used? Are you creating space for experimentation, or are you inadvertently punishing it?

These are the questions that determine whether AI adoption succeeds or stalls. And they have almost nothing to do with the technology.

The uncomfortable truth

The uncomfortable truth about AI adoption is that the technology, for once, is the easy part.

The tools exist. The training is available. The vendors are eager. What's genuinely hard is the human side. And most organizations, for reasons that are entirely understandable, would rather spend money and attention on the parts they can procure than on the parts that require honest conversation, real leadership, and the willingness to engage with what their people are actually feeling.

The organizations that are getting AI right have made a different choice. They've recognized that the technology is a tool, that the tool works only in the hands of the people who use it, and that everything about how those people show up depends on the leadership around them.

AI is not a technology problem for most organizations.

It's a change leadership problem.

And leaders who understand that are the ones whose organizations will figure out what to do with the hammer.