AI anxiety at work is usually treated as a mood. Something to calm down with a town hall, a training module or a reassuring email from the top. Two pieces of research released this week suggest that misses the point. The fear is not only about losing a job. It is about losing what the job means. And when that fear goes unanswered, it starts to show up in how people behave.

That is where leaders get into trouble. They see the behavior. They do not always see the fear underneath it.

So they treat a structural problem as a personal one.

Key takeaways

  • A new Elon University survey finds most Americans expect AI to eliminate more jobs than it creates, and many fear it will make work less meaningful.
  • New Washington State University research links fear of AI replacement to moral disengagement and more self-reported counterproductive behavior at work.
  • AI anxiety at work is a clarity problem before it is a motivation problem. Silence about AI’s role lets fear fill the gap.
  • Workers are asking for specific things: guidance on AI’s limits, clear and updated AI use policies, and honest information about which roles will change.

What’s happening: new research on AI anxiety at work

On October 13, Elon University’s Imagining the Digital Future Center released a national survey on how Americans see AI and the future of work. The survey, fielded by SSRS from August 20 to 24, reached 1,158 U.S. adults, with an oversample bringing the number of workers to 775. The headline finding is stark: 83% of adults expect AI to eliminate more jobs than it creates, and 62% expect most available jobs to be dull and unfulfilling. Sixty-five percent agree that if AI makes work less meaningful, people will lose purpose and dignity.

The survey also asked workers what paid work means to them. Eighty-one percent said it is extremely or very important to their role in their family. Seventy percent said it is extremely or very important to their sense of meaning and purpose. Sixty-four percent said the same about their personal identity.

Then it asked where workers would turn for help if AI significantly affects their jobs. Family, former colleagues and friends came first, at 33% saying they were extremely or very confident in that support. Employers came in at 22%. Among all adults, 43% said local employers are unprepared for the changes coming, versus 26% who said they are prepared.

The survey also asked what would actually help. Fifty-two percent of workers said training and guidance about AI’s limits and when human judgment should take priority would be extremely or very helpful. Forty-nine percent said the same about clear, regularly updated policies on AI use. Researcher Janna Anderson summarized the gap this way: “People have little faith that businesses, schools and communities will step up to effectively assist them in the AI transition.”

A day later, on October 14, Washington State University published research on what fear of AI replacement does to workplace behavior. The study, led by WSU Vancouver psychology professor Tahira Probst and published in Occupational Health Science, looked at workers across 18 industries. Workers worried that their occupation could become obsolete were more likely to morally disengage, which Probst describes as “a way of justifying bad behavior.” That disengagement was linked to more self-reported counterproductive behavior, such as surfing the web or showing up late for a shift. The effects were strongest in industries more exposed to AI, like finance and insurance, and weaker in industries such as forestry and construction.

Probst’s explanation is worth sitting with. A bartender whose bar closes can find work at another bar. That profession is not going away. Someone whose occupation is being automated does not have that fallback. The researchers’ recommendation to employers was simple: pay attention to how you talk about and implement the technology so employees stay engaged and supported.

What this reveals: the fear is about the definition of work

In The Next Turn, I describe two changes happening at once. The first is the definition of work: what people believe work is for, what it should provide and what it is worth. The second is the structure of work: how tasks actually get done, which AI is now reshaping.

Most AI conversations inside organizations live entirely on the second turn. Tools. Workflows. Adoption rates. Productivity.

The Elon data says workers are living on the first turn. When 70% of workers tie paid work to their sense of meaning and purpose, a change in the structure of work is never only structural. It reaches into identity. It reaches into family. It raises a question people do not usually say out loud: if a system can do what I do, what am I here for?

That is the collision. A structural change is landing on a human definition of work, and in many organizations nobody has explained how the two fit together.

In Chapter 1, “When What Worked Stops Working,” I write about how leaders tend to personalize what is structural. A capable employee starts coming in late. Someone who used to go the extra mile now does the minimum. The natural read is attitude, or effort, or commitment. The WSU research points to a different read. Some of that behavior may be what fear looks like when people do not know where they stand.

That does not excuse the behavior. The standards remain. Showing up, following through and treating colleagues fairly still matter. But if leaders only respond to the behavior, they will keep treating the symptom while the cause gets stronger.

Chapter 12, “Defining Work Again,” makes a point I keep coming back to:

Ambiguity is not neutral.

The Next Turn

When leaders say nothing specific about AI’s role in the work, people do not hear neutrality. They fill the silence with their own story. Often it is the worst-case story. One of the seven moves in Chapter 12 is to define AI’s role in the work before employees define it informally. The Elon survey shows what happens when that move is skipped. Thirty percent of workers who use AI on the job say a key reason is employer expectations, and another 30% say they feel they must keep up. That is pressure without definition.

Look at what workers say they want. Not reassurance. Not a promise that nothing will change. They want guidance on AI’s limits and on when human judgment comes first. They want clear policies that get updated. Those are requests for clarity, and they are requests leaders can actually meet.

Chapter 1 puts the shift plainly:

The burden is no longer simply managing people. It is creating clarity.

The Next Turn

This is not a generational issue. The Elon findings cover adults across the workforce, and the WSU effects tracked industry exposure to AI, not age. A Baby Boomer in claims processing and a Gen Z analyst in finance can carry the same quiet question. Leaders who treat AI anxiety as one group’s problem will miss most of the people who have it.

Your next turn

Here are five moves for this week.

  • Name the fear before it names itself. In your next team meeting, say out loud that people may be wondering what AI means for their roles. Credibility comes from naming the tension, not from pretending it is not there.
  • Define AI’s role in one workflow. Pick one process your team runs. Write down where AI helps, where it should not be used and which decisions stay with a person. Share it. Update it when it changes.
  • Separate standards from signals in how you read behavior. Before you label someone disengaged, ask what has changed in their work and how clearly you have explained it.
  • Tell people where human judgment still leads. More than half of workers in the Elon survey said guidance on this would help. Make it specific to your team’s work, not a slogan.
  • Be honest about what you do not know. You may not know how every role will change. Say that, and say when you will share more. A dated promise of an update builds more trust than a vague reassurance.

Frequently asked questions

What causes AI anxiety at work?

Research released this week points to two sources. Many workers expect AI to eliminate more jobs than it creates, and many also fear AI will make work less meaningful. Because most workers tie paid work to their sense of purpose and identity, the worry goes beyond job security to what their work is for.

Can fear of AI affect employee behavior and performance?

Yes, according to Washington State University research published in Occupational Health Science. Workers who feared their occupation could become obsolete were more likely to morally disengage, which was linked to more self-reported counterproductive behavior. The effects were strongest in industries with high AI exposure, such as finance and insurance.

How should managers respond to employees who are worried about AI?

Start with clarity rather than reassurance. Define AI’s role in specific workflows, explain where human judgment still leads and keep AI policies current. Workers in the Elon University survey said training on AI’s limits and clear, regularly updated policies would help them most.

Fear fills the space clarity leaves

Workers are not asking leaders to stop the change. Most of them have already accepted that it is coming. They are asking someone to explain how it will work here, on this team, in this job.

That is a leadership task, not a technology task. And it is one that cannot wait until the tools are fully rolled out, because by then people will already have written their own version of the story.

What would your team say AI means for their work, if you asked them today?

Sources