A new study, led by Washington State University researchers, found that employees who fear their industry is threatened by artificial intelligence are more likely to show worse workplace behavior, like stealing a coworker's lunch or showing up late.
Someone fearing their occupation will become obsolete from automation is different from job insecurity, said Dr. Tahira Probst, a professor of psychology at WSU Vancouver. She’s been researching employee well-being and economic precarity for the past 30 years.
“AI and automation and other kinds of new technologies are not only impacting job insecurity, so worries that you might lose your specific job within an organization, but that it might actually transform your entire occupation as a whole,” Dr. Probst said. “So it doesn’t matter if you’re in organization A or organization B. These changes that are being brought about by organizations adopting AI can impact that entire occupation.”
For the study, researchers collected data from more than 600 workers in different industries across the United States over six years, including finance and construction.
Dr. Probst said that the study wasn’t made to determine how much AI is going to replace certain occupations; Instead, it looks at how people are reacting over worries of occupation and security concerns with AI.
“If you’re a bartender, let’s say, and you lose your job at a particular bar, there are lots of other bars that exist,” Dr. Probst said. “But if you’re a call center worker, or a medical transcriptionist, these are occupations that are increasingly at risk of being [completed] by AI. So if you lose your job as a medical coder, you can’t just turn around and get a job with another organization, because it’s an entire occupation that is being impacted by it, not just that specific job within a particular company.”
Researchers found that people who worry their occupation will become obsolete or transformed by automation are more likely to justify behaviors at work that they would typically consider inappropriate.
Dr. Probst said the onus is now on organizations and how they roll out AI.
“I think that, for me, the main takeaway is that how organizations roll out these AI tools matters enormously,” Dr. Probst said. “And that if you leave it up to individual employees to make sense of what's happening, these fears, whether warranted or not, of occupation insecurity…will result in these negative outcomes that harm not only individual employees, but also organizations.”
In research, there’s always a margin of error, Dr. Probst said.
“I teach statistics, and I always tell my students that a single study doesn't prove anything,” Dr. Probst said.
But, she’s hoping that this study and others like it get replicated around the world.
“Replication and extension to different geographic areas with different levels of labor market protections, different levels of AI regulation, that would all be really important to look at, because we just don’t know whether the results would be the same,” Dr. Probst said.
The study was published in Occupational Health Science, an online science journal.
