VIEW AT BOTTOM OF SET OF CONCRETE STAIRS LOOKING UP TO THE TOP INTO A BLUE SKY (Photo by H. Armstrong Roberts/ClassicStock/Getty Images)

H. Armstrong Roberts/ClassicStock

Early in my career, I deployed a technology that I believed had tremendous potential across the organization. I still believe it did. However, it accomplished nothing. Our post-mortem called it "lack of adoption." The reality was that I failed to convince the users it had any value to them.

I think about that rollout constantly now, because I see so many companies reliving it at scale.

Gallup just released a study confirming it. Only about 1 in 10 employees inside companies that have adopted AI strongly believe it changes how their work gets done. Ten percent. Meanwhile, regular AI use jumped to 45% of workers in ManpowerGroup's latest data. And confidence in using the technology fell 18% over the same stretch.

So people are touching the tools more frequently (and token costs keep climbing), but fewer people trust them. That looks less like adoption and more like a mandate pushing behavior while leadership fails to make the case.

Here is the tough part for those of us who hold the technology seat. We are wired to read rising usage as progress. License counts go up and we call it a win. But we need to recognize that a license is not a changed behavior. You can force people to log in. You cannot require them to think differently. That gap is the hard part. And we are measuring the wrong thing.

Foundry’s research this year shows the seat is shifting under us and 46% of CIOs now see themselves as business leaders. Aligning technology strategy with the business has become the defining CIO priority, ahead of the keep-the-lights-on work that used to top the list. For decades, the job was uptime and breach prevention. Now the top of the list is alignment, which is a polite word for influence. The CIO's value is no longer in what we own. It is in what we can move.

That is the real story. The modern CIO does not win on infrastructure. We win on belief. And belief is a different problem than deployment.

I learned this scaling technology at lululemon as we grew from $2B to past $10B. In a stretch like that, you ship a lot of systems fast. The ones that stuck were never the ones I mandated. They were the ones a store leader started swearing by, or a planner quietly built her whole week around, and then mentioned to three peers over coffee. Belief traveled sideways, not down. By the time it reached me, it was already real.

The systems that failed were the ones I tried to push from the top. I had the org chart. I had the budget. I had executive backing. None of it mattered. People clicked the button I required and changed nothing about how they actually worked. That is compliance, and compliance is worthless in a transformation. It looks like motion and produces none.

So if you cannot mandate belief, how do you earn it?

You make it safe to go first. Every new tool needs an early believer, and being the first to bet on something unproven is socially risky. Nobody wants to look foolish championing a thing that gets killed in six months. The CIO's job is to lower that risk. Name your early adopters. Protect them. Make trying the new thing the smart move, not the exposed one.

You make it safe to say it is not working yet. This is the one most leaders skip. If your AI rollout is quiet, you probably think that is acceptance. It is usually the opposite. It is people keeping their heads down because admitting confusion feels like admitting they are behind. You cannot scale what people are afraid to question. When someone tells you the tool is clunky or the output is wrong, that is not resistance. That is the most valuable feedback you will get, and it only surfaces where saying so is welcome.

And you show instead of tell. Conviction does not spread through a town hall. It spreads through one person watching another work faster, calmer, better, and thinking, I want that. Your job is to engineer those moments, not to give the speech.

None of this is technical. That is the point. The hardest part of the CIO job in 2026 has almost nothing to do with technology and almost everything to do with how people decide to trust a change. We were trained for the first problem. We are being graded on the second.

The gap closes one believer at a time, and believers are made, not assigned. The CIOs who understand that are about to pull away from the ones who do not. Same tools. Same budgets. Same vendors. The only variable is whether their people believe, and belief was never something authority could buy - but it’s the one thing that matters most.