Measurement and Analysis

AI Anxiety: How Worried Are Employees About the Future of Their Jobs?

Digital Institute
Last updated: November 17, 2026 · 12 min read
A brown-haired woman sitting at her office desk hides behind her laptop, with only her eyes visible, glancing sideways with a worried expression

When AI comes up in leadership meetings, the topics are usually predictable: which tools to buy, which processes to automate, what return to expect on the investment. The question on employees' minds is often far more personal: "What will happen to my job?"

In short
  • AI anxiety is an employee's worry that AI will threaten their job, role or value. It is fueled less by the technology itself than by uncertainty and a lack of information about the organization's intentions.
  • Anxiety is not a single feeling. Treating job security, loss of meaning in the role, fear of falling behind, surveillance and uncertainty as separate dimensions makes it clear where the problem lies.
  • Anxious employees do not necessarily use AI less; they use it more quietly. Adoption metrics collected without measuring anxiety therefore do not tell the full story.
  • The most effective ways to reduce anxiety are open communication about plans, real time for learning, involving employees in the design of the change and being clear about how productivity gains will be used.

This question is rarely asked out loud. It is not raised in meetings, it is brushed aside in one-to-ones and it is discussed in half sentences in the corridor. Yet its effects are felt throughout everyday work: in keeping a distance from new tools, in reluctance to share knowledge, in intent to stay and in trust in the organization.

In this article we look at AI anxiety from the employee's point of view rather than the organization's. We explain the dimensions that make up this anxiety, how it can be measured in an employee survey, how to read the results and what can be done to reduce it. If you want to measure how ready your organization is for AI, see our article on measuring AI maturity, and if you want to see the real impact of AI tools, see measuring AI effectiveness with employee surveys.

What is AI anxiety?

AI anxiety is an employee's worry that AI will threaten their job, their role, the value of their skills or their place in the organization.

An important feature of this anxiety is that it is fueled more by uncertainty than by the technology itself. Employees often worry less about what AI can do than about not knowing what the organization plans to do with it. When general forecasts in the news, reports of downsizing in the industry and plans that are not openly discussed inside the organization come together, the gap gets filled with the worst-case scenario.

AI anxiety should be distinguished from two similar concepts:

  • It is not the same as resistance. An employee who finds AI useful, and even actively uses it in their daily work, can still worry about the future of their job. Reading anxiety as resistance puts employees on the opposing side and leads to the wrong actions.
  • It is not the same as general job insecurity. Economic conditions, restructuring or a shrinking industry also create job insecurity. AI anxiety, however, is tied to a specific change and can therefore be directly influenced by the organization's communication and implementation decisions.

Why measure anxiety?

The most commonly tracked indicators in AI transformation are the number of licenses, the share of active users and frequency of use. These indicators show what employees do, but not how they feel. There are three main reasons to measure anxiety.

Anxiety hides adoption. Anxious employees do not always use AI less; sometimes they use it more quietly. They hesitate to show that their work has become easier, fearing that their workload will grow or that their role will look unnecessary. As a result, the organization cannot see the real use and impact of AI, and good practices do not spread across teams.

Anxiety affects engagement and intent to stay. Employees who cannot see their future in the organization first stop going the extra mile, then start looking for alternatives. This often happens faster among those who can most easily find another job, in other words the employees the organization least wants to lose.

Anxiety increases silence. Employees who think their job is at risk avoid making mistakes, asking questions and sharing ideas. This weakens exactly the environment of experimentation, learning and feedback that AI transformation needs most. In this sense, AI anxiety is closely related to psychological safety.

The five dimensions of AI anxiety

Measuring AI anxiety with a single question answers "Are employees worried?" but not "Why are they worried?" Looking at anxiety through the following five dimensions shows where action should be directed.

1. Job security anxiety

This is the most visible dimension: the worry that an employee's job or position will disappear because of AI. It tends to be high in roles built on repetitive, rule-based tasks and on processing text or data.

2. Loss of meaning and value in the role

An employee who does not fear losing their job may still worry that their work will lose its meaning. The devaluation of expertise built over many years, the reduction of work to "checking AI output" or effort becoming invisible are all part of this dimension. Among experienced specialists, this dimension is often more pronounced than job security anxiety.

3. Fear of falling behind

This is the worry that there will not be enough time, support or ability to learn new tools. Comments such as "younger colleagues pick this up much faster" or "there is training, but no time for it in the middle of the work" are signs of this dimension.

4. Surveillance and evaluation anxiety

This is the worry that AI will be used to monitor, measure or evaluate employees' performance. Not knowing how decisions are made heightens this anxiety and directly weakens trust in the organization.

5. Uncertainty and lack of information

This means not knowing the organization's plans for AI, how these plans will affect employees and the principles by which decisions will be made. This is often the dimension that feeds the other four. The good news is that it is also the one the organization can influence most directly.

Questions to ask in the survey

The following statements can be used with a five-point agreement scale (Strongly disagree – Strongly agree). Some statements are worded positively and some negatively; negatively worded items should be reverse-scored when reporting.

Job security

  • I think developments in AI will put my job at risk within the next few years.
  • I trust that our organization will not treat employees as expendable because of AI.

Meaning and value of the role

  • I worry that my expertise will lose its value as AI becomes more widespread.
  • AI allows me to spend more time on the more meaningful parts of my work.

Skills

  • I receive the support I need to use AI tools in my work.
  • I can genuinely find time within my workload to learn new tools.
  • I worry that I will not be able to keep up with the rapid change in AI.

Surveillance and evaluation

  • I know how AI is used in our organization to monitor or evaluate employees.
  • I trust that AI will be used fairly in decisions about my performance.

Uncertainty and communication

  • Our organization's plans regarding AI are openly shared with us.
  • I can talk with my manager about how AI will affect my team's work.
  • Employees' views are taken into account in decisions about AI.

In addition, a short open-ended question provides very valuable information: "What concerns you most about the impact of AI on your work?" To analyze these responses at scale, see our article on AI analysis of open-ended comments.

Tips for question design

Avoid leading language. Questions such as "Are you afraid AI will take your job?" both amplify anxiety and invite exaggerated answers. Use calm, balanced statements that let employees assess their own situation.

Ask about use and anxiety separately. "I use AI tools" and "AI puts my job at risk" measure different things. Asking both makes the group that uses AI but is anxious about it visible.

Keep anonymity strong. Anxiety is one of the topics where employees are most concerned about being identified. Set a minimum group size so that small groups are not reported, and tell employees about it clearly.

Pay attention to timing. Measuring right after a restructuring announcement makes it hard to separate AI-specific anxiety from general job insecurity. If possible, schedule the survey independently of major announcements, or read the results in that context.

How to read the results

Look at the breakdowns

The company-wide average is often misleading. AI anxiety varies greatly by role, unit, tenure, age group and collar type. Groups whose work consists of the tasks most affected by AI and groups with the most limited access to tools and training should be examined separately. Among white-collar employees, job security and the value of the role may stand out, while for blue-collar employees uncertainty and lack of information may be more decisive.

Read anxiety and use together

Looking at anxiety and use scores together reveals four different groups:

  • Using and not anxious: Natural ambassadors of the change. Give them room to share their experience.
  • Using but anxious: The most critical group. They see the benefits of AI but worry that it will work against them. Quiet use is common in this group.
  • Not using and anxious: A lack of both support and trust. Communication comes before training.
  • Not using and not anxious: They may not yet have connected the topic to their own work. Concrete use cases work well for this group.

Look at the gaps between dimensions

If job security anxiety is high and communication scores are low, a large part of the anxiety may stem from uncertainty and could drop quickly with open communication. If skills anxiety is high but support scores look good, training may exist but there may be no time left to learn. If surveillance anxiety is high, the issue is less about technology and more about organizational trust.

Key driver analysis can show how much AI anxiety affects engagement and intent to stay in your organization. In groups with high anxiety, signals of quiet cracking and change readiness scores should also be examined together.

What should be done to reduce anxiety?

Share plans openly. You do not need to have every answer. But sharing where AI will be used, which decisions have not yet been made and the principles by which decisions will be made prevents the gap from being filled with the worst-case scenario. Saying "we don't know yet, but we will talk again on this date" builds far more trust than staying silent.

Be clear about how productivity gains will be used. Employees' most basic question is where the time saved by AI will go. Will it be devoted to more meaningful work, learning or new responsibilities, or will it become a justification for headcount cuts? As long as the answer remains unclear, quiet use and anxiety will continue.

Give real time for learning. Offering a training catalog is not enough. Clearly defined time within the workload for employees to try out new tools is one of the fastest ways to reduce skills anxiety.

Prepare managers for the conversation. Employees share their concerns most often with their direct manager, but most managers do not know how to answer these questions. Give managers a short, clear guide on the organization's plans, the promises they can and cannot make and how to talk about the topic with their teams.

Involve employees in designing the change. Deciding together with the people who know the work best how AI will be used in which tasks produces better solutions and increases the sense of control. The strongest antidote to anxiety is being an active participant in the change.

Draw clear lines on surveillance. Share in writing what AI will and will not do with employee data. Explain the role of AI in performance decisions.

Measure again. AI anxiety changes quickly. Alongside the annual survey, track how anxiety develops with short pulse surveys at critical stages of the transformation. To share the results and the actions taken with employees, see our article on sharing employee survey results with employees.

Conclusion

AI anxiety is the hidden cost of transformation. It does not appear in license reports, usage dashboards or productivity calculations, but it shows up in quiet use, in teams that fall silent and in unexpected resignations. Because much of this anxiety is fueled by uncertainty rather than by the technology itself, it can be directly influenced by the organization's decisions on communication, learning and involvement.

Measuring anxiety not with a single question but through the dimensions that make it up shows why employees are worried and ensures that action is directed to the right place. The organizations that adopt AI successfully are often not those with the best tools, but those that answer their employees' question "What will happen to my job?" most honestly.

If you would like consulting support to measure AI anxiety with dimensions tailored to your organization, analyze the results by role and unit and turn the findings into a concrete communication and action plan, get in touch with us.

Start preparing your survey
Our guide includes sample questions and a step-by-step preparation checklist.
Download the Guide

Frequently asked questions

Is AI anxiety the same as resistance?

No. Resistance is a deliberate stance against a change. Anxiety is often found among employees who support the change as well. An employee can believe AI is useful while worrying about the future of their job at the same time. Measuring the two separately prevents the wrong actions.

Will asking about AI anxiety make employees more worried?

Not if the questions are worded well. Instead of leading questions such as "Are you afraid of losing your job?", use balanced statements that let employees assess their own situation. Explaining in advance why the survey is being run and how the results will be used also builds trust.

How many questions are needed to measure AI anxiety?

Three or four questions added to the annual engagement survey give an overall picture. If the organization is in the middle of a major AI transformation, a separate section of eight to twelve questions covering all five dimensions, or a short pulse survey, gives more reliable results.

Which employee groups tend to be more anxious?

Groups whose work consists of repetitive, rule-based tasks and groups with limited access to AI tools and training are usually more anxious. However, this differs from one organization to another. Breaking the results down by role, unit, tenure and collar type shows the picture in your own organization.

Should anxiety be eliminated completely?

No, that is not a realistic goal. A certain level of concern can show that employees take the change seriously and are open to learning. The real goal is to keep anxiety from reaching a level that freezes employees, silences them or pushes them away from the organization.

Author
Digital Institute
A research and consulting firm that runs employee satisfaction and engagement surveys, bringing research, technology and consulting together.