When employees push back on AI, most leaders see a problem to be managed.
“They’re afraid of change.” “They need more training.” “They’re stuck in old ways.”
Sometimes that’s true. But often, the people resisting your AI rollout know something you don’t.
Their resistance is not just friction. It is information.
The gap between leaders and teams
In many businesses, leaders believe their teams are excited about AI. The teams themselves often feel very differently.
Some employees quietly avoid the new tools. Some use them only when someone is watching. Some go back to old methods as soon as the launch excitement fades.
Leaders who treat this as simple change resistance often launch pilot after pilot with little to show. Leaders who stop and ask why often discover blind spots no dashboard could reveal.
Four kinds of resistance, four different messages
1. Tool resistance: “This doesn’t work properly.”
Employees have tried the AI and found it wanting. It makes mistakes. It doesn’t fit their workflow. When your accounts team refuses to use an AI tool, they may be protecting you from errors you haven’t noticed.
2. Strategy resistance: “This is in the wrong place.”
AI is being used where it adds little value, while the real pain points are ignored. The people closest to the work often see this mismatch long before the numbers confirm it.
3. Trust resistance: “We don’t believe what you’re telling us.”
If leadership says AI is there to help people, but also talks about cutting headcount, employees respond logically to mixed signals. This isn’t fear of technology. It’s a reasonable reaction to contradictory messages.
4. Capability resistance: “We haven’t been prepared.”
If you expect people to adopt AI without proper training, time and support, you are setting them up to fail. And they know it.
Not all pushback is useful
Some resistance is simply obstruction. People who refuse reasonable change just because they don’t like it. Leaders can’t let personal discomfort become a veto over necessary progress.
The skill is telling the difference. Treat resistance as information first. Listen. Understand. Then decide whether you are dealing with a change problem, a technology weakness or a genuine concern that needs fixing.
What to do this week
1. Talk to your sceptics. Not to convince them. To understand them. Ask what they think is wrong and what would make it work.
2. Sort what you hear. Is it about the tool, the strategy, trust or capability? Each needs a different fix.
3. Compare where AI is with where the pain is. Is your AI investment going where your people actually struggle the most?
4. Be honest about jobs. If roles will change, say so clearly and explain how you will support people. Silence creates worst-case assumptions.
Example: A company rolls out an AI tool to check customer orders for errors. The warehouse team keeps ignoring its alerts. Management assumes they’re resisting change. When someone finally asks, the team explains that the tool raises far too many false alarms, flagging correct orders as wrong. Checking every alert takes longer than the old manual check. The tool is retrained using the team’s feedback, false alarms drop, and the warehouse team starts relying on it.
The bottom line
Before you try to overcome resistance, try to understand it.
You may find that your AI strategy needs more fixing than your people do.
Your most resistant employees might be your best source of truth.
What is your team’s resistance trying to tell you?