Workplace AI doesn’t appear to be having a negative effect on wellbeing for now, research suggests

A new study published in the journal Nature: Scientific Reports offers a cautiously optimistic view of how artificial intelligence is affecting workers’ wellbeing. Contrary to common fears, the research finds no clear evidence that AI exposure is harming workers’ mental health or job satisfaction. In some cases, it may even be contributing to small improvements in physical health, particularly among workers without a college degree.

The study, titled Artificial Intelligence and the Wellbeing of Workers, uses two decades of longitudinal data from the German Socio-Economic Panel. Researchers Osea Giuntella (University of Pittsburgh and the National Bureau of Economic Research), Luca Stella (University of Milan and Berlin School of Economics), and Johannes King (German Ministry of Finance) examined how workers in occupations more exposed to AI have fared compared to those in less exposed roles.

Their analysis found no significant effects of AI exposure on life satisfaction, mental health or job satisfaction on average. However, some positive trends emerged, including slightly better self-reported physical health and a modest drop in the number of hours worked per week, without a corresponding decline in income or employment.

Professor Stella, who is also affiliated with CESifo and IZA, said that while public anxiety about AI is understandable, the worst-case outcomes are not inevitable. He suggested that some improvements could be the result of AI reducing the physical demands and risks of certain jobs.

However, the researchers caution that it is too early to draw firm conclusions. The study relies heavily on a task-based measure of AI exposure, considered more objective, but also explores self-reported exposure. That alternative approach revealed small negative effects on job and life satisfaction, raising concerns that perceptions of AI may be affecting worker morale even if the tangible impacts are limited so far.

The study also focuses only on Germany, a country with strong labour protections and slower AI uptake. Younger workers, who are likely to be at the forefront of future AI adoption, were not included in the sample.

“This research is an early snapshot, not the final word,” said Giuntella. “As AI spreads and deepens its influence across industries, its full impact on work and health remains to be seen. Outcomes will depend as much on policy and institutions as on the technologies themselves.”

The researchers recommend continued monitoring of AI’s influence on working conditions, particularly in less protected or more rapidly changing labour markets.