The purpose of this study is to empirically examine how generative A I literacy affects users' personal information disclosure intention, focusi ng on the mediating role of AI trust and the moderating effect of age. As generative AI services rapidly ...
The purpose of this study is to empirically examine how generative A I literacy affects users' personal information disclosure intention, focusi ng on the mediating role of AI trust and the moderating effect of age. As generative AI services rapidly expand, understanding the factors th at influence users' willingness to share personal information has becom e increasingly important for both service providers and policymakers. This study employs privacy calculus theory and trust theory as theor etical frameworks to explain the decision-making process regarding per sonal information disclosure in generative AI environments. A survey w as conducted with 304 respondents who have experience using gener ative AI services such as ChatGPT, with participants distributed across two age groups: digital natives (20s-30s) and digital immigrants (40s-5 The research hypotheses were tested using structural equation modeli ng (SEM) with AMOS. Multi-group analysis was employed to examine th e moderating effects of age across different pathways in the research mo The empirical findings reveal several key insights. First, generative AI literacy significantly influences information quality perception. Second, i nformation quality perception positively affects both AI trust and perso nal information disclosure intention, while privacy concern negatively i mpacts both AI trust and personal information disclosure intention. Third, AI trust significantly affects personal information disclosure intention(β = .154, p < .05), although AI trust did not significantly media te the relationship between antecedent variables and personal informati on disclosure intention in the full sample. Finally, multi-group analysis across the two age groups demonstrated that the effects of AI trust on personal information disclosure intention vary significantly by age grou p: AI trust significantly affects disclosure intention among digital immigra nts (40s-50s), whereas this effect is not significant among digital natives (20s-30s). Additionally, privacy concern significantly reduces AI trust only amon g digital natives. This study contributes to the literature by integrating privacy calculu s theory and trust theory to explain personal information disclosure beha vior in the context of generative AI. The findings provide practical implic ations for developing generation-specific AI literacy programs and desig ning privacy-preserving generative AI services that can address the different decision-making mechanisms across age groups. Key Words : Generative AI, AI Literacy, Privacy Calculus Theory, Trust, Personal Information Disclosure, Age Moderation