The Real Reason AI Rollouts Stall
It is not the tool. It is trust, culture, and leadership.
There is a question I hear constantly from leaders right now: “Why is my team resisting the AI tool?” It sounds like a technology problem. In most cases, it is not.
I spoke recently with Rui Bashir, who leads global HR for a Fortune 500 multinational operating across more than 20 countries. What he shared reframed the whole conversation for me. His opening statement cut straight to it: AI rollouts in global organisations succeed when leaders focus on trust and authority, not just the technology. That is the part of the story most people are not telling.
The Wait That Looks Like Resistance
Rui made a point that I think every leader needs to hear. When a team appears to be resisting a new AI tool, they are almost certainly not refusing to learn it. They are doing something entirely rational: they are doing their own quiet calculation.
What might I lose? Whose role gets bigger? Do I still matter?
That is the trust question. Whether the team believes someone will handle those concerns fairly. The authority question is whether people whose roles are being changed were ever actually asked about it. When the answers to both questions are unclear, the team waits. And in most organisations, that wait gets read as resistance. The gap between what leaders are communicating and what the team is trying to figure out is where years of stalled momentum live.
The solution Rui describes is, in one sense, straightforward: explain the why and the how clearly, and momentum follows. But as Rui was quick to point out, that framing is easier to state than it is to execute, especially at scale.
Same Strategy, Completely Different Outcomes
This is where the conversation genuinely surprised me. Rui’s perspective is shaped by two decades of sitting in rooms where global strategy meets local reality. He has watched identical rollouts land in completely different ways depending on geography, culture, and language, and he gave one of the most striking examples I have heard to illustrate it.
Take a single word: courage. In Anglo-American corporate culture, courage is the lone voice at the table, dissent as a virtue, someone who says the thing nobody else will say. In Portuguese, the equivalent word means something closer to collective resolve: you go to the front because everyone behind you goes with you. The lone voice is not courageous; it is suspect. In certain Spanish-speaking contexts, calculated restraint is the mark of a skilled leader. Knowing when not to speak is part of the skill. The courageous thing is sometimes what you do not say in the meeting.
Three languages. Three definitions. One word. Now design an AI rollout that asks teams to “speak up if they have concerns”, and three teams hear three entirely different things. One team complies. Another interprets the silence as deference to the metrics. A third performs adoption while quietly killing the initiative.
Same product, same training, same KPIs. Completely different outcomes. And the standard headquarters read is always a change management problem.
As Rui put it: the local team is not struggling with change. The local team is doing exactly what the culture taught them to do when they sense the calculation is not safe.
None of that nuance is in the standard AI rollout playbook.
The Diagnostic Gap
Beyond culture, Rui made another point that applies universally, whether you lead a team of five or a function spanning continents: the importance of diagnosis before intervention.
He offered a vivid illustration. Imagine two leaders who say exactly the same thing in the same way: “I cannot trust my own judgement.” Same words. Same energy. Completely different problems.
The first leader was promoted for technical brilliance and has never had to make a hard call under real pressure. They need practice: real decisions, real stakes, building the muscle from scratch. The second leader used to have that muscle. They built it over a decade of hard calls. Then years of being second-guessed and overridden wore it down. They do not need more practice. They need someone to trust them again, so they can trust themselves.
If you get the diagnosis wrong, you spend six months telling the first leader they are brilliant when what they actually need is to make the call. You push the second leader to make more calls when what they need is someone to say: that judgement was always sound, it just got beaten down. Either way, you make things worse.
Orchestration: What the Human Actually Does
As AI takes on more of the operational doing, Rui argues that the most important human skill is not using the tools, it is orchestrating them. He described orchestration as the allocation problem: where does direction need to be given rather than absorbed? Where does trust need to be granted rather than withheld? Where does courage need to be spent rather than saved?
AI will not make those distinctions. The human will. Most leaders are already making these calls every day, instinctively. The opportunity, as Rui sees it, is to make those judgements more visible and more deliberate, so that scarce human resources go to exactly the right place.
When I asked him what trust looks like in an AI-augmented team, his answer was clear: how you use the information AI generates either builds or breaks trust. And when you operate across cultures, that translation of information must preserve trust, or the whole initiative begins to fracture.
The Apprentice Gap
Rui also picked up on something I explore in depth in Enhanced Leadership: what I call the apprentice gap. AI is already taking away the operational, entry-level work that silently shaped how people learned to do the harder work above it. The note-taking, the data entry, the first drafts. These tasks felt like graft. But they were also the training ground.
Rui sees this playing out in real time. Jobs at every level are being reshaped. Companies are actively seeking AI-native skills. And if schools and universities do not start building AI literacy into their curricula at pace, the cost of that adjustment will be felt across society. The concern is not that AI makes people redundant. The concern is that it removes the ladder that taught people how to think, decide, and lead.
For Leaders Right Now
If you are leading an AI rollout, Rui’s advice is to resist the temptation to treat it like any other change programme: train people, communicate, measure adoption. That works on the surface. What is going on underneath is trust and authority.
Before the announcement, before the training plan, ask yourself two questions. Do my people believe this will be handled fairly? And have the people whose roles are being changed actually been asked?
If you lead globally, add a third: what does this message mean in the room I am not sitting in?
For coaches, the pattern to look for is the leader who frames a team’s hesitation as resistance without ever asking what that hesitation is made of. The diagnostic question is not “how do we get them to adopt the tool?” It is “what calculation are they doing, and what are they not yet sure of?” The answer to that question changes everything about the intervention.
