Perceived skill devaluation as an indirect statistical pathway between perceived generative AI impact and algorithmic anxiety among college students: associations with adaptive coping strategies

Introduction Generative artificial intelligence (AI) is reshaping graduates' employment environments, job-competency requirements, and career competition. This study examined how perceived AI-related technological impact is associated with college students' perceived skill devaluation, algorithmic anxiety, career self-efficacy, growth mindset, and adaptive coping strategies. Methods A cross-sectional questionnaire study was conducted among college students. The instrument was developed in Chinese and refined in a preliminary sample through exploratory factor analysis and item diagnostics. The final measurement and structural models were evaluated using confirmatory factor analysis and structural equation modeling. Results Perceived AI technological impact was positively associated with perceived skill devaluation and algorithmic anxiety, and perceived skill devaluation was positively associated with algorithmic anxiety. The product-of-coefficients indirect association from perceived AI technological impact to algorithmic anxiety through perceived skill devaluation was statistically significant. However, the full indirect association extending through algorithmic anxiety to adaptive coping strategies was not statistically confirmed because its bootstrap confidence interval included zero. Career self-efficacy and growth mindset were not significantly associated with lower algorithmic anxiety and did not show significant moderating associations, but both were positively associated with adaptive coping strategies. Discussion These findings suggest that AI-related employment anxiety among college students is associated with their appraisal of whether technological change threatens the value of their career competencies, whereas adaptive coping is more strongly associated with positive psychological resources. Because the data are cross-sectional, the findings should be interpreted as theoretically consistent associations and statistical indirect pathways rather than demonstrated temporal or causal processes.