Linking incidental AI information exposure to AI-generated information corrective intention: testing an extended cognitive mediation model among Chinese AI users

Abstract

Purpose The rapid proliferation of generative AI technologies has heightened concerns about the spread of misinformation generated by large language models. Despite the significance of this issue, limited research has examined the psychological factors that motivate individuals to correct such misinformation. This study investigates the cognitive and social antecedents of AI information correction intention (AICI) within the extended Cognitive Mediation Model (CMM). Design/methodology/approach Drawing on survey data from 611 Chinese AI users, this study proposes a conceptual framework that links IAIE, elaboration, perceived AI hallucination, AI literacy, and perceived norms to AICI, and employed structural equation modeling to examine the proposed relationships. Findings The findings show that AI literacy and perceived norms significantly enhance individuals’ intentions to correct AI-generated misinformation. Furthermore, IAIE, elaboration, and perceived AI hallucination are positively associated with AI literacy, and both elaboration and perceived AI hallucination positively predict perceived norms. In addition, elaboration and perceived AI hallucination are significantly influenced by IAIE. Originality/value This study extends the CMM to the context of generative AI by highlighting the role of cognitive processing and social norms in motivating individuals to correct AI-generated misinformation. The findings contribute to the theoretical understanding of human–AI interaction and offer practical insights for developing interventions to improve AI literacy and foster responsible engagement with AI technologies.

Ähnliche Arbeiten