---
title: Your team is not resisting AI. It is resisting being redesigned without a say.
description: Employee resistance to AI comes in more than one kind. How to tell which one you face, and the learning actions that change it, on your own unit.
image: https://managementkits.com/hubfs/Management-Kits_One-Word-Four-Kinds.png
---

[Skip to content](https://managementkits.com/blog/employee-resistance-to-ai#main-content)

![Management Kits](https://managementkits.com/hubfs/Element%201_svg.svg)

- [The MK Practice](https://managementkits.com/practice)
- For organizations
  
  Show submenu for For organizations 
  
    - [Companies and teams](https://managementkits.com/enterprise-solutions)
    - [International schools](https://managementkits.com/education)
- [Blog](https://managementkits.com/blog)
- [About](https://managementkits.com/about)

Open main navigation

Close main navigation

- [The MK Practice](https://managementkits.com/practice)
- For organizations
  
  Show submenu for For organizations 
  
    - [Companies and teams](https://managementkits.com/enterprise-solutions)
    - [International schools](https://managementkits.com/education)
- [Blog](https://managementkits.com/blog)
- [About](https://managementkits.com/about)
- [Contact us](https://managementkits.com/blog/employee-resistance-to-ai#contact)

[Contact us](https://managementkits.com/blog/employee-resistance-to-ai#contact)

 2 Nov 2026

# Your team is not resisting AI. It is resisting being redesigned without a say.

Share this blog: [LinkedIn](http://www.linkedin.com/shareArticle?mini=true&url=https://managementkits.com/blog/employee-resistance-to-ai) [Twitter](https://twitter.com/intent/tweet?url=https://managementkits.com/blog/employee-resistance-to-ai) [Email](mailto:?body=https://managementkits.com/blog/employee-resistance-to-ai)

**What looks like employee resistance to AI is, more often than not, resistance to being redesigned without a say. It also comes in more than one kind. This note shows how to tell which kind you are facing, in two learning actions on your own unit, and what to do next. For many teams what comes next is one 45-minute session in which you use the tool on a real case in front of your team, and the team decides how one step of the work should run from now on.**

The pattern is familiar. An AI tool arrives in a team's work: a drafting assistant, a triage step, a summary that goes to the client. It was chosen somewhere else, announced in a meeting and followed by a training. A month later, half the team uses it and the other half has found good reasons not to. The word that comes up next is resistance, and the answer that comes after it is usually more training.

## Why is my team resisting AI (and is it, actually)?

"Resistance" is one word for several different things, and they need different answers. The Stanford team we cited in our last note looked at [51 enterprise AI deployments that worked](https://digitaleconomy.stanford.edu/publication/enterprise-ai-playbook) and found the most frequent resistance not among the people using the tools (23% of cases) but in legal, HR, risk and compliance (35%), where the worry is liability and who answers when the output is wrong. The people using the tools mostly worried whether they could trust output that changes from one run to the next. Fear of being replaced was the concern everybody talked about, and it showed up in only two of the 51 cases. In one company, the most resistant layer was middle management.

The staff functions need a role in the rules, which is what the decision-rights map is for. Your team needs something else. It is reacting to how the tool arrived: a change to their work that was designed somewhere else, announced in a meeting and followed by training. From the outside that looks like resistance; from the inside it looks like common sense. In our work with leadership teams, the word "resistance" tends to come up before anyone has asked the team what actually changed in their work.

None of this is new. In a [1948 experiment at a factory in Virginia](https://doi.org/10.1177/001872674800100408), the group that was simply told about a new way of working lost about a fifth of its output and did not get it back; the groups that helped work out the change were back within weeks and ended above where they had started. Same factory, same change, a different say. One factory is not a law of nature, and pooled studies since find the effect real but modest; that is a reason to make the say real, not to skip it.

The numbers from the current wave fit that reading. In Gallup's 2026 research in organizations that have introduced AI, [78% of employees who say their manager actively supports AI use](https://www.gallup.com/workplace/713063/ai-workplace-productivity.aspx) use it frequently, against 44% of everybody else, and only about a third say they have that kind of support. A survey cannot tell you which way the arrow points, so treat it as a strong pattern rather than proof. The experimental evidence is older and narrower, and it points the same way: in a series of experiments published in [Management Science](https://doi.org/10.1287/mnsc.2016.2643), people were considerably more willing to use an imperfect algorithm when they were allowed to adjust its output, even slightly. *A say does not need to be large to matter; it needs to be real.*

At Management Kits we take a clear position here. Most of what gets labelled resistance to AI is a reasonable response to being redesigned without a say, and the lever is not more training but **participation in the redesign of one concrete piece of work**, led by the person the team watches most closely: you. We made a related point in [our post on Gioia and Chittipeddi](https://managementkits.com/blog/change-doesnt-start-with-a-plan.-it-starts-with-meaning): change starts as a negotiation of meaning, not as a rollout. Skip the negotiation, and you get the told group.

*Working on this in your own organization? [See how the MK Practice works →](https://managementkits.com/practice?utm_source=blog&utm_medium=cta&utm_campaign=p1-practice&utm_content=pn03-mid)*

## How do you address resistance to AI?

Not with a training, and not with a better announcement, but with learning actions on your own unit. They will not be the same for every team, because the resistance is not the same. What every path has in common is how it starts: first you understand the dynamics at play, then you assess your situation. What follows depends on what you find.

1. **Understand the dynamics at play (45 minutes).** Ask the three people who use the tool least what has changed in their work since it arrived. Write down their words, not your reading of them. What you learn: which kind of resistance you are actually facing.
2. **Assess your situation (30 minutes).** Sort what you heard into four boxes:
     1. trust in the output
     2. rules and accountability
     3. fear for the job
     4. no say in the change.

Decide which boxes are yours to work on and which belong to someone else, such as the rules that [our last note](https://managementkits.com/blog/who-may-overrule-the-ai) puts on one page. What you learn: where your lever is, and where it is not.

Here is how the path went on for a team where most of what the manager heard landed in the last box, no say in the change:

- **Use the tool in front of your team (45 minutes).** On a live case, not a demo, including the part where it gets something wrong. Say upfront what is open for discussion and what is not; participation on a decision already taken elsewhere is theatre, and people can tell. What you learn: what the team sees when you use it, and what they ask.
- **Let the team redesign one step (45 minutes).** One point where work passes from one person to the next with the tool in between. The team decides who does what now, what gets checked and by whom, and logs three decisions, each with a name. What you learn: how much your team already knows that the rollout never asked for.
- **Look again after two weeks (30 minutes).** Who uses the tool now, and what became of the three decisions? Write down what you would do differently the next time a change lands. What you learn: whether it was the step that was wrong, or the people. It is rarely the people.

A team whose concerns land in the first box needs different actions: time with the tool on its own cases, and an agreed way to flag output that is wrong. A team that fears for its jobs needs a straight answer about what the saved time is for, which is the subject of the next note. In our work with leadership teams, the listening is the action managers are most tempted to skip, and the one that changes the room. Teams are rarely short of opinions about a tool. They are short of evidence that the person asking them to use it has listened.

If you want the same move with more structure around it: in the MK Practice, the follow-through sits in the working topic *Define interventions to involve stakeholders*, with frameworks for change interventions and for activating networks; its learning actions go onto your work agenda and come back for review. For the climate the session depends on, see [our post on Edmondson's original psychological safety study](https://managementkits.com/blog/psychological-safety-isnt-niceness.-the-original-study-never-said-it-was).

![The yellow card in the MK Practice workspace: Turn resistance to AI into a redesign your team owns, five learning actions](https://managementkits.com/hubfs/Management-Kits_Workspace_Resistance-Milestone.png)

A decision rule we find useful: if the tool has been in the workflow for a month and fewer than half the team use it, do not buy more training. Start the first two actions this week and look at usage again two weeks after the session. If nothing has moved after a quarter, the step is wrong, not the people.

The people on your team who hold back are not lazy and not against progress; they do what people do with a change they had no part in. So before you plan the next training, ask three of them what has changed in their work, and write down their words. It takes 45 minutes, and it is the one thing no course can do for you. The next note is about the time the tool gives back, and the decision almost nobody has taken about what that time is for.

You don't read a practice. You run it.

The MK Practice is a structured process you run on your own organization: a diagnostic that reads where your unit needs work first, working topics like the ones in this series, a work agenda that keeps the moves coming, and a practitioner who looks at your problem within 48 hours. Founding Member: USD 990 a year, 30-day money-back guarantee.

[See how the MK Practice works →](https://managementkits.com/practice?utm_source=blog&utm_medium=cta&utm_campaign=p1-practice&utm_content=pn03-participation)

*Want to talk it through first? [Book 20 minutes with a Management Kits team member](https://calendly.com/managementkits-demo/talk-first?utm_source=blog&utm_medium=cta&utm_campaign=p1-practice&utm_content=pn03-call). Running it with a whole leadership team? [See the MK Practice for organizations](https://managementkits.com/enterprise-solutions?utm_source=blog&utm_medium=cta&utm_campaign=p1-practice&utm_content=pn03-org).*

[Leadership](https://managementkits.com/blog/tag/leadership), [Organizational Change](https://managementkits.com/blog/tag/organizational-change), [AI transformation](https://managementkits.com/blog/tag/ai-transformation)

Copyright © 2023 by Management Kits AG

```json
{
  "@context" : "https://schema.org",
  "@type" : "BlogPosting",
  "author" : {
    "@type" : "Person",
    "name" : "Olaf Bach",
    "url" : "https://managementkits.com/blog/author/olaf-bach"
  },
  "dateModified" : "2026-11-02T21:59:25.546Z",
  "datePublished" : "2026-11-02T21:59:25.000Z",
  "headline" : "Your team is not resisting AI. It is resisting being redesigned without a say.",
  "image" : [ "https://managementkits.com/hubfs/Management-Kits_One-Word-Four-Kinds.png" ],
  "mainEntityOfPage" : {
    "@id" : "https://managementkits.com/blog/employee-resistance-to-ai",
    "@type" : "WebPage"
  },
  "publisher" : {
    "@type" : "Organization",
    "logo" : {
      "@type" : "ImageObject",
      "url" : "https://managementkits.com/hubfs/Element%201_svg.svg"
    },
    "name" : "Management Kits AG"
  }
}
```