Beyond Current AI Use: Perceived AI Automation Potential and Workers’ Expectations of Future Job Opportunities

Abstract

The rapid integration of artificial intelligence (AI) into workplaces has increased uncertainty about its implications for future employment. While existing research has examined workers’ current exposure to AI, less attention has been given to how workers perceive the extent to which their current job tasks could potentially be performed by AI. This study examines the relationships between current workplace AI exposure, perceived AI automation potential, and workers’ expectations regarding future job opportunities. Drawing on Cognitive Appraisal Theory, the study analyzes data from 3,975 employed respondents in the Pew Research Center’s American Trends Panel Wave 157 (2024). Weighted multinomial logistic regression models were used to examine whether perceived AI automation potential provides incremental explanatory value for workers’ expectations beyond current workplace AI exposure. The findings indicate that perceived AI automation potential is strongly associated with expectations of fewer future job opportunities, even after controlling for current AI exposure, age, gender, education, income, and industry. Adjusted predictions further show that expectations of fewer opportunities increase from 28.8% among workers perceiving no automation potential to 58.3% among those perceiving that all their work could be performed by AI. The findings highlight the importance of distinguishing between what AI currently does and what workers believe AI could do in the future.