Risk Perception and Trust-Building in AIGC Applications: A Bayesian Structural Equation Model Analysis

The rapid diffusion of generative artificial intelligence (AIGC) technologies is accompanied by multiple risks, which profoundly impact public acceptance and trust in the technology. This study integrates the Unified Theory of Acceptance and Use of Technology 2 (UTAUT2) and the Social Amplification of Risk Framework (SARF) to construct a theoretical model encompassing Risk Perception, System Trust, Risk Trust, Behavioral Intention, and Risk Prevention Sensitivity. Based on 696 valid survey responses from Jiangsu Province, a Bayesian Structural Equation Model (BSEM) is employed to empirically analyze the complex interactions among these variables. The results reveal that both Risk Perception and System Trust significantly and positively influence Risk Trust, with System Trust exerting a stronger effect. Furthermore, Risk Trust positively affects Behavioral Intention, while Risk Prevention Sensitivity demonstrates a significant negative inhibitory effect. Based on these findings, the study proposes policy recommendations such as enhancing algorithmic transparency, improving multi-stakeholder governance mechanisms, and strengthening public digital risk literacy to promote responsible innovation and effective governance of AIGC technologies.