Your board loves AI. Your workforce doesn't trust it yet.
Here is the number that should stop any AI rollout in its tracks. Only 9 percent of workers trust AI for complex, business-critical decisions. Among executives, the figure is 61 percent. That is not a rounding difference. It is two organisations pointing in opposite directions, and only one of them controls the budget.
Those figures come from WalkMe's State of Digital Adoption 2026, a global study of 3,750 executives and employees across 14 countries at enterprises of 1,000 staff or more. The report is full of uncomfortable reading for anyone rolling out AI at pace, but the trust gap is the one that matters most, because it explains all the others.
The workforce has already voted, quietly
While the board talks about AI transformation, the people who are supposed to use the tools have made their own decision. More than half of workers, 54 percent, bypassed AI tools and did the task by hand at least once in the past month. A third are not using AI at all.
In our experience, this is never laziness or resistance for its own sake. People route around a tool when, at the moment they need it, the approved way feels harder than the workaround. If the AI assistant is slow, buried, or wrong often enough to erode trust, they go back to what they know. The rollout looks complete on paper. Adoption never happened.
This is the same pattern we have seen with every enterprise platform for a decade. What is new is the speed and the spend. Firms are pouring money into AI and capturing only 55 percent of the potential value, with 40 percent of digital spend underperforming. The gap between what was bought and what is used has simply moved from CRM and ERP to AI, and it is widening.
The bit that should worry a regulated firm
For a bank, insurer or lender, the trust gap is not just a productivity story. It is a control story.
The same research found that 45 percent of workers used unsanctioned AI tools in the past 30 days, and 36 percent did so with confidential data. Read that again with a compliance hat on. More than a third of your people have pasted confidential information into tools your firm never approved, never assessed, and cannot see.
That is shadow AI, and it is the direct consequence of the trust gap. When the sanctioned path is hard and the unsanctioned one is easy, people take the easy one. You do not fix that with a policy memo or a mandatory training slot. You fix it by making the approved route the path of least resistance, and by being able to see where people fall off it.
Adoption is the control layer, not a nice-to-have
This is where digital adoption stops being a training topic and becomes governance. The firms getting real value from AI treat adoption as a measured discipline, not a launch event.
In practice that means three things. Guide people to the sanctioned tools inside the flow of work, so the compliant path is the easy one. Measure where they drop off, which prompts fail, and who has quietly stopped using the tool. And close the loop, fixing the friction that pushes people toward shadow AI in the first place.
Done well, this is not soft. On WalkMe-enabled processes we have seen a 60 percent reduction in support tickets, because the guidance sits where the work happens rather than in a manual nobody opens. The same mechanism that lifts adoption is the one that pulls shadow AI back into the light.
What "we deployed the AI" does not mean
"We deployed the AI" is not the same as "our people trust it and use it." Deployment is an event. Adoption is what shows up at the desk on day ninety, after the announcement has faded.
The trust gap will not close because you buy a better model. It closes when the sanctioned tools are easier to use than the workarounds, and when someone can actually see what is happening. That is an adoption problem, and it has an owner-shaped hole in most organisations right now.
Not sure where your own adoption and shadow-AI gaps sit? Answer 9 questions and get a personalised Digital Adoption Score across three dimensions: not just whether you have a problem, but why, and what to do about it.
Tell us what is not landing
Tell us what you are rolling out and where adoption, automation or AI is sticking. We will come back with a clear plan for the first steps, what success looks like, and what it costs. No fifty-slide pitch.
Book a call