Short answer: Employer-employee trust is already near a low point, and AI is an easy thing to blame for it. Data Society’s Catie Maillard argues that’s backwards: the real driver is change fatigue, years of reorgs, return-to-office mandates, and new tools stacked on top of each other, with AI added as a tool that never stops changing and almost never comes with a plan. Employees are told to use it without being told why, without usage guidelines, and without knowing what gets them in trouble. That combination, not the technology itself, is what erodes trust.
The trust numbers were already bad before AI showed up
Maillard’s starting point is blunt: blaming AI for a trust problem makes the problem worse, because it distracts from causes organizations can actually fix.
“Blaming AI for anything right now is already backfiring on trust. Most of the articles I see in the HR world are about how trust is the lowest it could possibly be between employers and employees.”
Catie Maillard, Global VP of People, Data Society
The numbers back her up. A 2026 LiveCareer survey of 900 U.S. workers found 80% rarely or never trust leadership to prioritize employee interests over profit when the two conflict, and 71% expect employers to downplay or avoid discussing how AI affects their jobs. Gallup data cited in the same reporting found 29% of U.S. employees say leaders don’t communicate clearly, honestly, or consistently. AI didn’t create this gap. It’s landing in the middle of one that was already there.
The real driver is change fatigue, not the technology
Maillard’s argument is that AI gets blamed because it’s the most visible item on a list that was already too long.
“I think the main reason is change fatigue, to be honest. We’ve been throwing change at our employees at a pace that’s more rapid than our brains can handle since the pandemic started. There’s lots of research on this that shows the number of change initiatives, whether it’s a reorg, a business strategy change, return to office, work from home, new technology, and keep in mind that an AI change is never-ending. These tools are constantly changing, so we need to constantly figure out how our workflows change and how we share that with other folks.”
Catie Maillard, Global VP of People, Data Society
Gartner’s research on organizational change gives the fatigue a number: the average employee absorbed 10 planned enterprise changes in 2022, up from just 2 in 2016, a fivefold increase in six years. Over the same period, employee willingness to support organizational change collapsed from 74% to 43%. AI isn’t one more line item on that list. It’s a line item that, as Maillard points out, never finishes changing, which means the fatigue it adds doesn’t taper off the way a single reorg eventually does.
“Figure it out” is a governance failure, not an employee problem
The sharpest part of Maillard’s argument is about what happens after a company rolls out an AI tool without a plan.
“If you’re telling me I need to be using this but aren’t telling me why, and you’re not giving me guidelines around when I use this and when I don’t, I don’t know if I’m going to get fired because I used too many credits or not enough. There’s just so much confusion out there that it’s really hard for employees to navigate. And most of the time, they just don’t have time to. So they’ll use your tools in between meetings, they’ll figure things out as they go along. It’s not effective, it’s not efficient, and I understand why we’re degrading trust this way.”
Catie Maillard, Global VP of People, Data Society
That confusion is measurable, not anecdotal. An IT Brew survey of 241 IT professionals found 35% aren’t confident employees understand their company’s AI usage and data security policies, and only 12% said they were “very confident.” At the same time, a Culture Amp survey found 77% of HR professionals say they’re bought into AI, but only 24% feel comfortable letting it act autonomously within bounded workflows. Leadership is enthusiastic about the technology and unsure about the rules for using it. Employees are left holding both halves of that gap at once.
What a thoughtful rollout actually requires
Maillard’s fix isn’t slower AI adoption. It’s the governance work that’s supposed to come with it: clear guidelines on when a tool is required and when it isn’t, a stated reason for the change instead of a mandate with no explanation, and a rollout built for people who are already stretched thin by everything that changed before this. None of that has to slow adoption down. It has to happen at the same time as adoption, not after employees have already been degrading trust by guessing.
Frequently asked questions
Is AI actually the main cause of declining employee trust?
Maillard argues no. Trust between employers and employees was already low before AI, driven largely by change fatigue. AI adds to that fatigue because, unlike a reorg or a policy change, it never stops changing.
How much has employee tolerance for change actually declined?
Gartner research found employees absorbed 10 planned enterprise changes on average in 2022, up from 2 in 2016, while willingness to support organizational change fell from 74% to 43% over the same period.
Do employees actually understand their company’s AI usage rules?
Often not. An IT Brew survey of IT professionals found 35% lack confidence that employees understand their company’s AI usage and data security policies.
Why does rolling out AI without guidelines erode trust specifically?
Because employees are told to use a tool without being told why, without knowing when its use is expected versus optional, and without clarity on what usage could get them in trouble. That ambiguity, not the tool, is what damages trust.
What should organizations do differently when rolling out AI tools?
Pair the rollout with real change management: a stated reason for the change, clear usage guidelines, and governance, delivered at the same time as the tool rather than left for employees to piece together on their own.
Next steps
If your organization rolled out an AI tool without a stated reason or clear usage guidelines, that’s the gap to close before the next rollout, not after. Data Society’s AI Advisory work helps leaders build the governance and change management that actually protects trust, and AI upskilling programs give employees the shared understanding a mandate alone can’t.
