A study led by psychologist Tahira Probst at Washington State University has found that employees who fear their entire profession may be rendered obsolete by artificial intelligence are significantly more likely to engage in counterproductive workplace behaviors, ranging from chronic lateness to the theft of company resources. Published in the journal Occupational Health Science in September 2026, the research suggests that the existential threat of AI does not motivate harder work, but instead triggers a psychological detachment from professional ethics.
The research was prompted by a shift in corporate hiring philosophies during the spring of 2025. On April 7, 2025, Shopify founder Tobi Lütke issued an internal memo requiring teams to prove why a task could not be performed by AI before requesting new human staff. Just weeks later, Duolingo founder Luis von Ahn announced the company would begin phasing out contractors for roles that AI could now handle. Probst and her team used these real-world signals of job insecurity to investigate how such news alters the psyche of the modern workforce.
How does the fear of occupational extinction differ from standard job insecurity?
Researchers distinguish between standard job insecurity, where a worker fears losing a specific position but can find work elsewhere, and occupational insecurity, where the profession itself is perceived to be disappearing. Tahira Probst illustrates this with the example of a bartender: if a specific bar closes, the bartender can find work at another establishment because the trade remains viable. However, for roles such as medical transcriptionists, the rise of automated dictation software means that if they lose their job, there may be no similar roles left in the market. This sense of terminal career prospects creates a unique psychological strain that differs from the temporary fear of a layoff.
To measure this effect, the team utilized the Occupation Insecurity Scale (OCIS) developed by Lara Roll in 2023. They surveyed employees across 18 different industries at three distinct intervals. The first phase measured their anxiety regarding the future of their profession; the second assessed their attitudes toward workplace rules; and the third recorded the frequency of counterproductive behaviors, such as wasting company materials, surfing the internet during work hours, or even taking a colleague’s food from the office refrigerator.
What is the link between AI anxiety and unethical workplace behavior?
The link between AI anxiety and unethical behavior is a psychological process known as moral disengagement, a concept first defined by Albert Bandura in 1986. Moral disengagement allows individuals to convince themselves that ethical standards do not apply to them in specific contexts, using justifications such as “everyone else is doing it” or “the company is being unfair to me.” Probst notes that these employees are well aware of the rules; they simply reinterpret their actions as acceptable responses to a perceived injustice or a looming threat to their livelihood.
The data confirmed a clear causal chain: workers who felt their profession was at risk showed higher levels of moral disengagement in the second phase of the study, which in turn predicted a higher frequency of harmful behaviors in the third phase. The authors suggest that when an employee views their career as endangered, they stop seeing the company’s success as their own. Instead, they may view cutting corners or minor acts of sabotage as a way to reclaim agency or compensate for the unfairness of their situation.
Which industries are most vulnerable to the psychological effects of AI?
Industries with high levels of routine cognitive tasks, such as finance and insurance, show the strongest correlation between AI fear and workplace misconduct. To determine which sectors were most at risk, the researchers utilized the AI Occupational Exposure index developed by Edward Felten. This index matches current AI capabilities with the specific skills required in various fields. While the path from anxiety to misconduct was pronounced in the financial and insurance sectors, it remained weak in industries like agriculture, forestry, and construction, where physical labor and unpredictable environments provide a natural buffer against current AI technology.
Global trends mirror these findings. Data from various national statistics offices shows a rapid rise in AI adoption, particularly in information, communication, and finance. For instance, reports indicate that the percentage of companies using AI has nearly tripled in recent years. As these technologies become more integrated into the white-collar workforce, the risk of widespread moral disengagement grows, suggesting that managers must address the psychological impact of automation rather than just the technical implementation.
Notes and Sources
- Roger Dooley, Forbes (2025). "Shopify CEO's AI-First Hiring Policy Is Job Security's Ticking Clock".
- Jibin Joseph, PCMag (2025). "Duolingo Adopts AI-First Strategy, Will Eliminate All Contract Workers".
- Tahira M. Probst et al. (2026). "In the Shadow of AI: How Industry Exposure to AI and Occupation Insecurity Fuel Moral Disengagement and Counterproductive Work Behaviors". Occupational Health Science.
- Shawn Vestal, WSU Insider (2026). "Workers who fear AI replacement show worse workplace behavior".
- Lara C. Roll et al. (2023). "Conceptualization and Validation of the Occupation Insecurity Scale (OCIS)". International Journal of Environmental Research and Public Health.
- Albert Bandura (1999). "Moral Disengagement in the Perpetration of Inhumanities". Personality and Social Psychology Review.
- Edward Felten et al. (2021). "Occupational, Industry, and Geographic Exposure to Artificial Intelligence: A Novel Dataset and Its Potential Uses". Strategic Management Journal.
- National Statistical Institutes (2025). "Artificial Intelligence Statistics Report".
