For years, people have worried that artificial intelligence (AI) would take jobs, increase stress, and make work more difficult. But a major new study suggests the reality might be very different. Researchers looking at two decades of worker data from Germany found no evidence that AI exposure has reduced job satisfaction or damaged mental health. In fact, it may be quietly improving physical health—especially for workers without college degrees—by taking some of the most physically demanding tasks off their shoulders. The study, “Artificial Intelligence and the Wellbeing of Workers,” published June 23 in Nature: Scientific Reports, analyzed long-term data from the German Socio-Economic Panel. The research team—Osea Giuntella from the University of Pittsburgh and the National Bureau of Economic Research (NBER), Luca Stella from the University of Milan and Berlin School of Economics, and Johannes König from the German Ministry of Finance—compared the well-being of workers in jobs more exposed to AI with those in less-affected occupations.
What the Study Found:
- Job satisfaction and mental health remain stable. No significant average effects of AI exposure were seen on job or life satisfaction.
- Physical health may be improving. Workers reported slight improvements in self-rated physical health, particularly among those with lower levels of education.
- Less physical strain. AI appears to be reducing job intensity by taking on physically demanding tasks.
- Working hours declined slightly. Weekly working hours dipped modestly, though income and employment rates stayed steady.
- A note of caution. Self-reported AI exposure did show small negative effects on well-being, suggesting that perceptions of AI may differ from task-based data.
Too Early for Final Conclusions:
The researchers are careful not to overstate the findings. Because the data only covers the early stages of AI adoption in Germany, and excludes younger workers, the long-term impacts remain uncertain. “Public anxiety about AI is real, but the worst-case scenarios are not inevitable,” said Professor Stella, who is also affiliated with CESifo and IZA. “If anything, physical health seems to have slightly improved, likely due to reduced physical demands in AI-exposed occupations.” Still, Stella warned that AI’s influence could shift dramatically as the technology spreads further. “We may simply be too early in the adoption curve to see its full impact,” he said. Giuntella agreed, stressing that outcomes will depend not only on technology but also on policies and institutions. “As AI adoption accelerates, monitoring its broader effects on work and health is essential,” he explained. “Technology alone doesn’t determine outcomes—how societies manage it will decide whether AI improves or erodes work conditions.”
Why Germany?
The study focused on Germany, a country with strong labor protections and a relatively gradual pace of AI adoption. The researchers note that results could look different in countries with looser labor laws or among younger workers entering increasingly AI-heavy workplaces. Journal Reference: Osea Giuntella, Johannes König, Luca Stella. Artificial intelligence and the wellbeing of workers. Scientific Reports, 2025; 15 (1). DOI: 10.1038/s41598-025-98241- Source: University of Pittsburgh
