Online Information Exposure, Knowledge Development, and AI Anxiety: Extending the Cognitive Mediation Model to Explain GenAI Adoption in China
Lunrui Fu, Mulin Jiang, Bo Hu, Yong Li · Social Science Computer Review · 2026
As Generative AI (GenAI) becomes increasingly integrated into everyday information environments, many individuals remain uncertain or reluctant to engage with these technologies despite their growing potentials. By extending the Cognitive Mediation Model (CMM), the study emphasizes the cognitive and emotional processes that underpin public engagement with emerging digital information technologies. A survey of 527 Chinese participants was conducted using an online questionnaire. Structural equation modeling was employed to test the extended CMM, examining the relationships between exposure to traditional and social media, elaboration, interpersonal communication, knowledge acquisition, anxiety, and GenAI usage intention. Results indicate that exposure to both traditional and online social media were positively associated with elaboration and interpersonal communication, which were in turn negatively associated with anxiety and positively associated with perceived knowledge. However, elaboration does not significantly relate to factual knowledge, while interpersonal communication was negatively associated with factual knowledge. Anxiety was negatively associated with GenAI usage intention, whereas factual and perceived knowledge were positively associated with adoption. Theoretical and practical contributions are discussed as well.