AI Health Counselor’s Self-Disclosure: Its Impact on User Responses Based on Individual and Cultural Variation

Abstract

This research extends the literature by showing that the positive effects of self-disclosure by an AI voice bot (AVB) vary depending on the identity perception of the bot (human or machine) and the cultural characteristics of the users: when an AVB is perceived as human and engages in personal self-disclosure, users feel more intimate with and trust the bot, and when the bot is perceived as machine and engages in factual self-disclosure, users feel more intimate with and trust the AVB. We also found that South Korean (vs. U.S.) users felt more intimate with and trusted AVBs when they perceived the bot as human and when it engaged in personal self-disclosure. The theoretical and practical implications of the study are discussed.