When citizens support AI policies: the moderating roles of AI efficacy on AI news, discussion, and literacy

This research revisits the communication mediation model to understand when citizens support AI policies. Using data from a national online survey (N = 1252), our research delineates the mechanism by which AI-related news consumption enhances AI literacy through AI-related discussions, subsequently fostering enhanced support for AI policies and regulations. In this process, AI efficacy emerges as a critical boundary condition. Specifically, the relationship between AI-related news consumption and support for AI policy and regulation via AI discussions and literacy varies contingent on levels of AI efficacy. The findings highlight a moderated indirect effect, whereby the positive influence from AI-related news consumption to support for AI policy and regulation through AI discussions and AI literacy only occurs when citizens have high levels of AI efficacy. This research offers a deeper understanding of the factors encouraging citizen engagement in AI policy and regulation activities and provides valuable insights for AI literacy and education from theoretical, practical, and policy perspectives.