Understanding Generative AI Adoption: An Integrated Model of Emotional, Cognitive, and Motivational Drivers
Van Kien Pham, Ai-Van Huynh, Linh Le Phuong Giao, Oanh Hoang · Human Behavior and Emerging Technologies · 2026
This study develops and empirically tests an associational model of generative artificial intelligence (GenAI) adoption by focusing on emotional resistance, social‐cognitive appraisal, and motivational agency. Rather than claiming that the breadth of constructs alone is novel, the study specifies how perceived AI competence functions as a GenAI‐specific capability cue that may be interpreted through anthropomorphic and social‐categorization processes, and how this competence cue is translated into adoption through trust, perceived control, and attitude. Drawing on survey data from 215 respondents with experience or awareness of GenAI in academic and research‐related contexts, the study examines perceived anxiety, perceived risk, perceived AI competence, perceived trust, perceived control, attitude toward GenAI use, and behavioral intention. The model was assessed using a nonexperimental cross‐sectional design and PLS‐SEM‐oriented reporting procedures; therefore, the results are interpreted as theoretically ordered associations rather than causal effects. The findings show that attitude is the strongest predictor of behavioral intention, whereas perceived control and perceived AI competence also have positive direct associations with intention. Perceived trust mainly operates through attitude rather than directly predicting intention. Perceived anxiety has a small negative direct association with behavioral intention, whereas perceived risk shows a capability‐trust dissociation; it reduces trust but does not significantly reduce perceived AI competence. Reverse‐outcome robustness checks further support treating the model as an associational explanation rather than as evidence of temporal mediation. The model explains 71.9% of the variance in behavioral intention. The study contributes to GenAI adoption research by clarifying how perceived competence, trust calibration, perceived agency, anxiety, and ethical risk appraisal jointly shape adoption among informed users.