You Can’t Mandate Trust in AI
![[OPT 1] IMG PMP – Editorial – AI Adoption Pace (1)-90731](https://peoplemanagingpeople.com/cdn-cgi/image/format=auto,quality=82,onerror=redirect,width=792,height=446,fit=crop/wp-content/uploads/sites/3/2026/09/OPT-1-IMG-PMP-–-Editorial-–-AI-Adoption-Pace-1-90731.jpg)
Trust Gap: AI adoption often lags deployment because employees need time to understand changes and relinquish familiar control.
Founder Trap: Founders can build quickly with AI, but limited participation may leave organizations unprepared to adopt resulting changes.
Psychological Safety: Leaders accelerate learning when they normalize uncertainty, admit limitations, and let employees question tools without embarrassment.
Useful Resistance: Repeated questions and hesitation can expose gaps in understanding, trust, or the sacrifices a new process requires.
Balanced Involvement: Effective change gives employees meaningful input without allowing constant oversight to interrupt experimentation and progress.
AI can have an impact inside an organization faster than the organization can make sense of it.
A new tool appears. Within months, leaders expect employees to adopt AI and change how they work. Wagner Denuzzo thinks that expectation misunderstands how people adapt.
“Humans do not trust what they don’t fully understand,” he said. “Adoption is therefore gradual by nature.”
Continue Reading for Free
Create a free account to finish this article, plus get ongoing access to timely insights and practical resources.
Have an account? Log In
That assessment carries a little more weight coming from Denuzzo than from the average retired executive. Before holding talent and leadership roles at IBM and Prudential, he trained and worked as a licensed psychotherapist. He now lives in Portugal, watching the AI transition with excitement, but also with an eye on what organizations routinely underestimate about human behavior.
One of his sharper observations concerns founders.
Founders are smart people who have no patience for bringing others to where they have arrived.
AI makes that tendency more consequential. Leaders increasingly need fewer people to turn an idea into a functioning prototype, workflow or product. But Denuzzo argues that needing fewer people to build the technology doesn’t mean engaging fewer people in what gets built.
This creates a new leadership trap. AI makes it easier for a small group to get remarkably far without broad participation, creating the illusion that the rest of the organization is ready to follow.
Deployment Moves Faster Than Adjustment
Denuzzo describes human adaptation as moving through stages of maturity. Technical deployment can happen quickly. Emotional integration usually cannot.
That isn’t an argument to slow AI programs until everyone feels comfortable. It is an argument about sequence.
Organizations can install a tool overnight. They cannot instantly replace the habits, expertise and sense of control built around the process it changes.
Denuzzo notes that AI is already forcing organizations to let go of processes people were accustomed to controlling, observing and completing themselves. That makes adoption more complicated than learning a new tool. People are also being asked to give up familiar ways of working, seeing how work gets done and judging whether it’s done well.
Some of what looks like reluctance to adopt something new may be discomfort with relinquishing something old. Rollout plans tend to be much better at teaching the former than accounting for the latter.
People Need Room to Admit What They Don’t Know
This is where Denuzzo gets practical.
Psychological safety, in his view, is not created with a framework or an executive speech. Leaders establish norms that remove shaming and judgment. They acknowledge what they do not know. They can laugh at previous failures. They give people room to express an idea without dismissing someone else’s experience.
“That’s how literacy works,” he said. “People learn better without fear.”
For AI adoption, that matters enormously.
If employees believe admitting confusion will make them look obsolete, they will hide confusion. If questioning a tool makes them look resistant, they will stop questioning it. Leaders then get apparent compliance without much learning underneath it.
Denuzzo offered a small example from his own work. When the founders asked him to help develop an AI leadership experience, he tried to build on their thinking rather than dampen their enthusiasm with criticism. That example gets at what engagement can look like without draining urgency from the work. People do not need every decision slowed down for consensus, but they do need enough room to question, contribute and understand what is changing.
Denuzzo also warns against the opposite mistake: involving people so heavily through updates and progress checks that leaders begin interrupting the work they are trying to support.
There is a difference between monitoring people and giving them room to work through the change. The first can become interference. The second gives people space to ask questions, admit what they do not understand, test ideas and contribute to how the technology is actually used.
That is why resistance is worth paying attention to rather than simply overcoming. Hesitation, repeated questions or attachment to familiar processes can reveal gaps in understanding, trust or what the change asks people to give up.
None of that means leaders should wait for universal comfort before moving. It means the friction itself contains information.
Trust developing slower than technical capability is not necessarily a flaw in the workforce to engineer around. It can be a signal about what the organization is prepared to absorb.
