Why What Looks Like AI Resistance Is Actually a Trust Rupture — and What HR Can Do About It
.jpg){kind=spacer}
Last spring, something kept happening at commencement ceremonies. Eric Schmidt was booed at the University of Arizona when he spoke about artificial intelligence. Gloria Caulfield was booed at the University of Central Florida after calling AI the next industrial revolution. Scott Borchetta was booed at Middle Tennessee State University when he told graduates to "deal with it."
Think pieces called it fear. Resistance to change. The discomfort of disruption.
I wrote at the time that they were wrong. Those graduates were not booing AI. They were booing the betrayal. They had done exactly what institutions told them to do, taken on the debt, chosen the majors, built the expertise, and then, on the one day meant to celebrate the promise being kept, a powerful person stood at the podium and told them the promise was already void.
Why does this matter to HR professionals? Well, those same graduates landed in your organization. They’re in your buildings, your onboarding cohorts, your Slack channels. And the betrayal they felt at those commencement ceremonies didn’t end there. It became the operating condition of how AI is arriving at work, for them and for people who have been in the workforce for thirty years
Right now, HR is being handed a familiar story about the workforce. People have "AI resistance." They have "change fatigue." They need to be more adaptable, more curious, more resilient.
It’s the wrong diagnostic, and reaching for it is how we miss what is actually happening.
Dr. Jennifer Freyd's Institutional Betrayal Theory tells us that harm lands hardest when it comes from an institution that was supposed to protect you. The damage is compounded not just by what happened, but by who did it and whether they ever acknowledged it. At work, that institution is the employer. And a lot of what gets labeled resistance is not a mindset problem. It is a trust rupture with a specific name and outcome
"AI will augment you, not replace you." Almost every organization says a version of this. Then comes the reorg, the backfill freeze, the role that gets absorbed into a tool.
Researchers call the distance between stated values and lived experience the policy-practice gap, and they treat it as its own distinct form of betrayal. It is not the AI that breaks trust. It is the daylight between what leadership promised and what people watched happen. Employees notice that gap faster than any engagement survey can measure it, and they file it away.
"Upskill or be left behind" is the most common piece of workplace advice of the last two years. It sounds empowering. It is not.
It puts the entire burden of a decision no worker made onto the individual worker. The organization chose the tool, the pace, the rollout, and the level of support. The employee is then told that their stability depends on how quickly and quietly they absorb it all. That is "deal with it" wearing an L&D badge. It absolves every executive who automated a function and called it innovation, and it asks people to perform resilience on command while standing in rooms full of people who will never face the consequences of the decision being celebrated.
Watch what happens when an employee raises a real concern about an AI rollout. Often the response is not "tell me more." It is "you're resistant to change."
Freyd has a name for that move too. DARVO: deny, attack, and reverse victim and offender. It was first described in interpersonal harm, and it now runs at organizational scale. A person flags a legitimate harm, and the institution reframes the person as the problem. The concern never gets addressed because the messenger got recategorized. Do that a few times in a company and people stop raising hands. The silence gets read as buy-in. It is not buy-in. Freyd would call the quiet itself betrayal blindness, the unawareness we cultivate to protect a relationship we depend on. People stay quiet because they need the job. Silence is not consent. It is dependence.
Not everyone worried about AI is worried about their own paycheck. Some are being asked to build, deploy, or sell systems they believe will harm colleagues, customers, or the public. That is not burnout. That is moral injury, and the distinction matters more than almost anything else in this conversation.
Burnout is workload driven. You fix it with capacity, recovery, and boundaries. Moral injury is betrayal driven. It comes from being made complicit in something that violates your sense of how the world is supposed to work. A wellness webinar does not touch it. Resilience training does not touch it. Offering one to a person carrying the other is its own small insult, and people feel it.
In a recent Harvard Business Review piece, researchers documented employees increasingly turning to AI for social and emotional support during the workday, and warned it may erode trust, collaboration, and connection over time. The easy read is that people are getting weird about chatbots.
The harder and truer read is this. When someone reaches for a machine instead of a manager or a colleague, ask what withdrew first. People fill voids. If the human infrastructure of your workplace has thinned out to the point that a language model is the safest thing to talk to, the AI is not the only story; the void is too.
Freyd paired institutional betrayal with a second construct, institutional courage, and that is the actual assignment for leaders. Not another round of resilience training aimed at the people absorbing the risk. A different set of choices from the people creating it.
The graduates who booed were early. They said what a growing share of your workforce is feeling. You can call it resistance and try to coach it away, or you can change the conditions that produced it.
Resilience was never the missing ingredient. The missing ingredient is an institution worth trusting.
Change the conditions, not the coping.
Want to continue the conversation with Stephanie? Reach out to her at Hello@troophr.com.
.jpg){kind=spacer}
.jpg){kind=spacer}
.jpg){kind=spacer}
.png){kind=spacer}
.jpg){kind=spacer}
.jpg){kind=spacer}
