A Study on User Satisfaction With Generative Artificial Intelligence: From the Perspective of the Technology Acceptance Model

While generative artificial intelligence (GenAI) drives innovation across various domains, user satisfaction varies considerably under the influence of multiple factors, including ethical compliance and technical performance. Existing studies on user satisfaction with GenAI lack systematic investigation of multidimensional factors and provide limited analysis of the underlying influence mechanisms. To explore the key influencing factors and their underlying mechanisms, this paper develops a user satisfaction model for GenAI based on the Technology Acceptance Model (TAM), integrating three dimensions: legal-ethical considerations, technical attributes, and user capabilities. A total of 408 valid questionnaire responses were collected, and CB-SEM (Covariance-Based Structural Equation Modeling) analysis was performed using AMOS 26.0 software. The empirical results indicate that technical reliability exerts a significant positive effect on user trust ( β =0.43, P <0.001) and satisfaction ( β =0.305, P <0.05), and also significantly promotes perceived usefulness. Technical literacy has significant positive impacts on perceived ease-of-use ( β =0.356, P <0.001) and user trust ( β =0.168, P <0.05); it influences satisfaction through multiple pathways including independent mediation and chained mediation of “perceived ease-of-use → perceived usefulness” and trust. Perceived usefulness significantly and positively drives user satisfaction ( β =0.267, P <0.001). Ethical compliance directly improves user satisfaction ( β =0.127, P <0.01), yet shows no significant effect on user trust. User trust plays a partial mediating role between technical reliability and user satisfaction. This study broadens the analytical perspective on user behavior toward GenAI, and provides empirical evidence and practical guidance for enterprises seeking to optimize technical services, regulators to formulate policies, and interventions for improving user literacy.