Is Transparency Always Good? Dual Cognitive Responses to Different Transparency Dimensions and Their Effects on Users’ Intention to Use

This study examines how the different dimensions of transparency (input transparency, output transparency, functionality transparency, and interaction transparency) influence individual intention to use AI digital assistants, and the dual cognitive mechanisms (privacy concerns as the affective-cognitive mechanism and perceived information overload as the information processing mechanism). The results suggest input transparency is negatively associated with intention to use, while the other three dimensions of transparency are positively related to intention to use. Privacy concerns and perceived information overload are both negatively related to intention to use. The mediating role of privacy concerns is only significant in the relationship between functionality transparency and intention to use. Perceived information overload significantly mediates the relationship between output transparency and intention to use, as well as the relationship between interaction transparency and intention to use. This research contributes to the advancement of transparency studies in AI-related applications and uncovers the cognitive mechanism of the effects of transparency on AI adoption.