Abstract
Objective: To quantify the associations between artificial intelligence use at work and psychosocial well-being among employed adults in the United States. Methods: Nationally representative data come from the Household Trends and Outlook Pulse Survey (United States, March 2026). Artificial intelligence use at work was measured as the number of distinct assisted work tasks. Outcomes included psychological distress, loneliness, and life satisfaction. Regression models adjusted for sociodemographic characteristics, region of residence, occupation, and telework frequency. Results: The data generalized to 152 million employed U.S. adults, with 51.7% using artificial intelligence at work. The most common artificial intelligence task was “searching for information or technical help” (35.8%), followed by “writing communications, documentation, or instructions” (31.6%). Greater artificial intelligence use at work was associated with greater life satisfaction (0.11 SD increase, 95% CI: 0.06, 0.16) and a higher prevalence of feeling lonely (aPR: 1.13, 95% CI: 1.02, 1.25). There was no significant association with moderate-to-severe psychological distress. Conclusions: The findings provide evidence that artificial intelligence use at work is associated with life satisfaction and loneliness, but not with overall psychological distress. Longitudinal studies are needed to examine the impact of increasing artificial intelligence use at work on employee psychosocial well-being.