When self-control is no longer a protective factor: the moderating effect of institutionalized identity on attitudes toward artificial intelligence
Ying Long, Yu Zhou, Die Luo, Ming Chang
As artificial intelligence (AI) is increasingly applied in high-risk industries, understanding the attitudes of different occupational groups toward AI and their underlying psychological correlates has become an important research issue. Existing studies typically conceptualize self-control as a psychological trait with universal adaptive value, suggesting that it helps buffer “individuals” negative attitudes toward new technologies; however, this assumption is primarily based on samples from the general population. Drawing on a situated perspective of psychological functioning, the present cross-sectional study examined whether institutionalized identity moderates the relationship between self-control and negative attitudes toward AI, using a convenience sample of 289 participants (179 university students, 76 pilots, and 34 flight instructors). Negative AI attitudes were assessed with the Negative Attitudes toward Artificial Intelligence Scale (NARS), which captures concerns and reservations toward AI rather than acceptance, trust, reliance, or behavioral adoption. Results showed that among university students, self-control was stably and negatively associated with negative AI attitudes; among pilots, this relationship was no longer significant; and among flight instructors, a non-significant positive trend was observed, which—given the small subgroup size ( n = 34)—should be regarded as a preliminary, exploratory finding rather than a confirmed effect. These results are consistent with the interpretation that the functional priority of self-control may shift from emotional buffering toward risk vigilance in high-responsibility occupational contexts; however, because the proposed underlying psychological processes (e.g., capability anxiety vs. responsibility-based caution) were not directly measured, this account is offered as a plausible interpretation rather than a mechanism demonstrated by the present data. In addition, the study introduces a dual-dimensional “value–dynamics” analytical framework as a conceptual extension intended to guide future research, since the dynamics-orientation dimension was not directly assessed in the present design. Overall, the findings provide preliminary evidence for the context-dependent nature of psychological trait effects and offer tentative implications for AI governance and occupational training in high-responsibility professional settings: AI integration programs and attitude-related interventions may need to be tailored to the distinct responsibility structures and psychological orientations of different occupational groups, rather than assuming universal effects derived from general-population research.