The Dominant Position of Perceived Usefulness and the Limited Role of Artificial Intelligence Anxiety: Determinants of Generative Artificial Intelligence Adoption Intention among University Students in Learning

The integration of generative artificial intelligence (GenAI) into the educational landscape of China is advancing at a pace marked by unprecedented speed, which requires researchers to re-examine the psychological mechanisms underlying students' technology adoption. Despite the established role of Perceived Usefulness (PU) as a core cognitive driver, the interplay between this factor and the emerging emotional barrier of Artificial Intelligence Anxiety (AIA) has not been sufficiently examined. This study theoretically bridges the Technology Acceptance Model (TAM) and the emotional cognitive perspective. Through a questionnaire survey of 192 Chinese university students, the impact of PU, AIA and their interaction on the intention of learning behavior using GenAI was systematically tested. Hierarchical regression analysis shows that PU emerged as a robust predictor of Behavioral Intention (BI) ( = 0.697, p < .01); however, AIA exerted no significant direct or interactive (with PU) effect on intention. Further exploratory analysis shows that even if AIA is deconstructed into three sub-dimensions of learning anxiety, sociotechnical blindness anxiety and AI configuration anxiety, its predictive effect on BI is still not significant, and the dominant role of PU remains robust ( = 0.720, p < .001). The research results show that under the present learning context, the adoption decision-making of university students shows the phenomenon of "intention-emotion decoupling", that is, it is mainly driven by the rational evaluation of the value of the tool, and different types of AIA do not substantially intervene in the process of intention formation.