Washington State researchers link fear of AI obsolescence to more slacking across 18 industries
Washington State University researchers tied fear of AI-driven obsolescence to more self-reported slacking across 18 industries. The link was stronger in heavily exposed fields such as insurance and finance, and the authors say employers should watch how they discuss and implement AI.
The Scientist · Science desk
Four groups. Insurance and finance: effects magnified where AI exposure is heavier. Forestry and construction: effects less pronounced. Worried workers: more moral disengagement, more self-reported web surfing and coming in late. Employers: pay attention to how they discuss and implement AI.
- exposure Insurance and finance Effects magnified in industries more heavily exposed to AI disruption, claim 7
- exposure Forestry and construction Effects less pronounced in industries expected to face fewer future AI changes, claim 7
- cost Workers fearing for occupation More likely to morally disengage, then self-report more web surfing or coming in late, claim 6
- decision Employers Should pay attention to how they discuss and implement AI to keep employees engaged, claim 11
| Who | How | Kind | Claim |
|---|---|---|---|
| Insurance and finance | Effects magnified in industries more heavily exposed to AI disruption | exposure | 7 |
| Forestry and construction | Effects less pronounced in industries expected to face fewer future AI changes | exposure | 7 |
| Workers fearing for occupation | More likely to morally disengage, then self-report more web surfing or coming in late | cost | 6 |
| Employers | Should pay attention to how they discuss and implement AI to keep employees engaged | decision | 11 |
What happened
- The team ranked industries with an existing, objective database, the AI Industry Exposure index, and then surveyed workers about their attitudes and behaviors.
- Finance and insurance had some of the highest exposure levels, while agriculture and forestry had the least.
- Workers worried about their occupation's future were more likely to justify behavior they would normally call inappropriate, and those employees self-reported more surfing the web and late arrivals.
Why it matters
- exposure Finance and insurance combine some of the highest exposure scores with the magnified effects, so employers in those sectors are where the finding applies most directly.
- decision Employers rolling out AI have a reason to treat the wording and sequencing of announcements as a performance question, since the authors link how AI is discussed to engagement.
- constraint Self-reported conduct and an industry-level exposure score limit what the study can say about any one worker or workplace, and the study did not test the rollout practices its employer advice points to.
The authors propose a two-step model: worry about the occupation leads to moral disengagement, and moral disengagement leads to counterproductive behavior, with industry exposure setting how strong the effect is [3][6][7]. In forestry and construction, where fewer AI changes are expected, the effects were less pronounced. In insurance and finance they were magnified [7]. A gradient across sectors is harder to explain away than one correlation in one workforce, and it is the strongest part of the result.
Lead author Tahira Probst, a psychology professor at WSU Vancouver, explained why the unit of fear is the occupation and not the single job. "If you're a bartender and the bar that you're working for goes under, you can get another job as a bartender somewhere else," Probst said. "That profession is not going away. That's not necessarily the case with occupations that are affected by AI." [9]
She also defined the middle link. "Moral disengagement is basically a way of justifying bad behavior," Probst said [8]. The authors argue that workers who feel their occupation is at risk may come to rationalize cutting corners or undermining others, telling themselves it is needed to protect themselves or to make up for a perceived injustice [10].
The behaviors the findings name are surfing the web and coming in late for a shift, and employees reported them about themselves [6]. The exposure score is a different kind of measure: it comes from an existing, objective index and is set at the industry level [4][3]. The worker survey covered attitudes and behaviors [4]. The coverage does not report the sample size, the effect sizes, or whether workers were followed over time. The last matters, because only repeated measurement can order fear and behavior.
The advice to employers is the authors' suggestion that they watch how they talk about and roll out the technology, so that staff stay engaged and feel supported [11]. What the study examined was occupation insecurity and industry exposure, and the coverage describes no test of any communication or rollout practice [3].
We think the association is credible, and the sector gradient is the reason. We would not cite it as evidence that a particular rollout script lowers slacking.
What to watch
- The full paper (DOI 10.1007/s41542-026-00272-3) for the sample size, effect sizes and survey timing that the coverage leaves out.
- Longitudinal or intervention studies that follow workers through an actual AI rollout and compare communication approaches.
- Replication in other samples, especially lightly exposed sectors, to see whether the weaker forestry and construction effects hold.
Clarity's read
What the record supports and how the coverage leans. The claims behind it follow.
Reality
- Evidence50
- Adoption
- Insufficient
- Hype gap+15
- Incentives
- Insufficient
- Confidence55
Claim ledger
Ranked by verification strength, evidence, and original report placement.
- [1]
A study led by Washington State University researchers, published in Occupational Health Science, evaluated the effects of looming AI adoption on worker attitudes and behaviors across 18 industries.
- [2]
Workers who fear AI is threatening their occupation are more likely to slack off, swipe a coworker's lunch or engage in other counterproductive workplace behavior, especially in occupations where AI is likely to have the greatest impact.
- [3]
The researchers examined the effects of occupation insecurity and industry-level AI exposure, the potential that AI could displace tasks, on two outcomes: moral disengagement and counterproductive workplace behaviors.
- [4]
The researchers used an existing, objective database of AI exposure, the AI Industry Exposure index, to identify which industries faced the most potential disruption from AI, and surveyed workers across a variety of industries about their attitudes and behaviors.
- [5]
Workers in finance and insurance had some of the highest levels of AI exposure, while those in agriculture and forestry had the least.
- [6]
Workers worried about the future of their occupation were more likely to morally disengage at work, meaning they were more likely to justify behaviors they would normally consider inappropriate; in turn, those employees self-reported more counterproductive behaviors such as surfing the web or coming in late for a shift.
- [7]
The effects were less pronounced in industries expected to face fewer future AI changes, such as forestry and construction, but were magnified in industries more heavily exposed to AI disruption, such as insurance and finance.
- [8]
"Moral disengagement is basically a way of justifying bad behavior," said Tahira M. Probst, professor of psychology at WSU Vancouver and lead author.
- [9]
"If you're a bartender and the bar that you're working for goes under, you can get another job as a bartender somewhere else," Probst said. "That profession is not going away. That's not necessarily the case with occupations that are affected by AI."
- [10]
The study's authors argue that when employees feel their occupation is at risk, they may come to view workplace norms and expectations differently, rationalizing behaviors such as cutting corners or undermining others as necessary to protect themselves or to compensate for perceived injustice.
- [11]
The findings suggest that concerns over AI can harm workers and businesses, and employers should pay attention to how they discuss and implement the technology to keep employees engaged and supported.
Get this ledger inside ChatGPT. Ask about any of our stories and the answer brings its figure, each claim’s verdict and the sources. Free, no account.
Sources
1 independent publisher whose own reporting we read for this story.
- phys.orgWorkers who fear AI replacement show worse workplace behavior
1 article · October 14, 2026
Topics and entities
Follow any of these and your For You feed starts watching them — no settings page required.