Your Team Isn't Resisting AI, Just Unowned Change
Somewhere around week six of a stalled AI rollout, a diagnosis gets written: the team is resisting. They're set in their ways, they're threatened by the technology, they need more training, maybe a change-management workshop. The diagnosis feels generous, even. Nobody's blaming the tool.
I've watched this diagnosis get written for thirteen years, through SharePoint, through Teams, and now through AI, and it's almost always wrong in the same way. The team isn't resisting the technology. The team is responding, rationally, to change that nobody owns. Those look identical from the executive floor and they are completely different problems, with completely different fixes.
What "resistance" actually is, case by case
Take the three behaviors that get filed under resistance and hold each one up to the light.
"They're not using it." The tool was introduced alongside the existing way of working, not in place of any part of it. Every task now has two paths: the familiar one and the unfamiliar one, and taking the unfamiliar one is optional. Then the quarter heats up. Under deadline, a person choosing the path they can walk without thinking is not resisting; they're prioritizing, which is what you hired them to do. Optional new habits lose to mandatory old habits every single time. If nothing old was retired, nothing new was adopted, and no amount of enthusiasm training changes that arithmetic.
"They're skeptical." Ask what happened to the last initiative. The pilot that trailed off, the platform that was announced and never mentioned again, the tool that was mandatory for a month. Skepticism isn't a personality trait of operations people; it's pattern recognition, earned honestly, from watching changes arrive without owners and leave without verdicts. A team that has seen three unowned initiatives die is not resisting the fourth one. It's forecasting, and its forecast has a better track record than the rollout plan does.
"They're quietly working around it." Someone keeps building the report by hand, keeps walking over to ask instead of asking the tool, keeps the old spreadsheet alive in parallel. Look closely and you'll usually find that nobody with authority ever closed the old way. The workaround isn't defiance. It's the org chart telling you the truth: the person who could retire the old path either doesn't exist or wasn't given the job.
Three behaviors, one root. In every case the team is doing something reasonable, and in every case the missing ingredient is the same: a specific person whose job it is to make this specific work happen differently.
The owner test
Here's the fastest diagnostic I know. Ask, of the stalled rollout: who, by name, is responsible for this changing how work gets done on this team?
If the answer is "IT deployed it," you have availability, not adoption; IT can make a tool exist, and cannot change how the operations team runs its Friday reporting, because IT doesn't own that workflow. If the answer is "everyone, really," you have a subscription with a communications plan. And if the answer is a name, ask the follow-up: does that person have the authority to retire the old way of doing things, and the obligation to report whether the new way is better? This is the same question as the third of the three questions worth asking before you even buy the tool. Ownership without both halves is sponsorship wearing a badge.
Teams calibrate to this instantly, because they always know whether a change is owned. They watched who announced it, whether anything old actually stopped, and whether anyone came back to check. The behaviors above are how a team prices that information. Change the information and the behavior follows without a single workshop.
What this means for the next attempt
The uncomfortable implication of the reframe: the fix isn't aimed at the team. It's aimed at the rollout.
- Name the owner before the tool arrives.
- Decide what stops, and say it with a date.
- Capture what the old way costs this week, so there's a number to come back to. The real math on why that number matters runs to twenty thousand dollars for a fifteen-person pilot once you count what the postmortem leaves out.
- Put the decision on the calendar: a day when someone compares before to after and says expand, fix, or kill, out loud.
Do those four things and the "resistance" mostly evaporates, not because anyone was persuaded but because the rational response to owned change is different from the rational response to unowned change. The same people who quietly kept the old spreadsheet alive will retire it themselves when someone with a name and a date asks them to, because now the safe bet has flipped.
The team was never the problem. The team was the instrument telling you what the problem was. Most organizations replace the instrument.
If your rollout has been diagnosed with resistance, the second opinion is usually cheaper than the workshop. Get in touch.
Rosemarie Withee has spent thirteen years helping operations teams get real work out of their software, first Microsoft 365, now AI. She’s written six books for Wiley and builds AI products at Portal Integrators.