Understanding user trust in AI-generated content: an elaboration likelihood model perspective
Tao Zhou, Xiaoqian Fang
Abstract
Purpose Based on the elaboration likelihood model (ELM), this research identifies the effect of both central and peripheral factors on user trust in AI-generated content (AIGC). Design/methodology/approach We adopted a mixed method of structural equation modeling and fuzzy-set qualitative comparative analysis to conduct data analysis. Findings The results indicate that central factors (perceived accuracy, perceived personalization and content explainability) and peripheral factors (perceived anthropomorphism, perceived bias and privacy risk) significantly affect user trust in AIGC. In addition, algorithm literacy has a moderating effect on trust. Originality/value Previous studies have primarily examined user adoption and continuous use of AIGC, and have seldom investigated the formation process of AIGC user trust. Based on the ELM, this study explores the mechanism shaping user trust in AIGC.