AI Chatbot Use for Emotional Support Among Young Adults in Japan: Psychosocial Correlates and Dependence Tendencies in a Cross-Sectional Study (Preprint)

BACKGROUND Generative artificial intelligence (AI) chatbots are increasingly used for mental health–related purposes, yet little is known about factors associated with such use or AI chatbot dependence. OBJECTIVE This study examined factors associated with young adults’ use of AI chatbots for emotional self-disclosure and mental health–related support, as well as associations between AI chatbot use and dependence. METHODS Participants were 820 individuals aged 18 to 29 years recruited through a Japanese online survey panel. Self-report measures assessed psychosocial factors (self-esteem and loneliness), mental health symptoms (anxiety and depression), AI chatbot use (emotional self-disclosure and mental health–related support), and AI chatbot dependence (emotional dependence and social anxiety–related dependence). RESULTS Overall, 46.5% of participants reported using AI chatbots for emotional self-disclosure at least once in the past month, including 7.9% who reported doing so almost daily. In addition, 41.0% reported using AI chatbots for mental health–related support, including 6.5% who reported almost daily use. Anxiety and depressive symptoms were combined into a mental health symptoms composite because of their high correlation. Multivariable logistic regression showed that mental health symptoms, but not self-esteem or loneliness, were associated with both emotional self-disclosure and mental health–related support. Among emotional self-disclosure users, emotional self-disclosure frequency was positively associated with both dependence dimensions. Path analyses indicated that a model specifying psychosocial factors and mental health symptoms as antecedent variables to AI chatbot use and dependence fit the data better than an AI-initiated alternative. In this model, lower self-esteem and greater loneliness were associated with higher mental health symptoms, which were associated with greater AI chatbot dependence largely through direct pathways, with a smaller additional pathway through more frequent emotional self-disclosure. Among emotional self-disclosure users, loneliness was the only significant psychosocial correlate of emotional dependence. CONCLUSIONS Findings suggest that young adults experiencing mental health symptoms may be more likely to use AI chatbots for emotional self-disclosure and mental health–related support, while loneliness may be particularly relevant to emotional dependence among users. These findings highlight the need to understand AI chatbot use not only as a technological behavior, but also as part of young adults’ broader emotional and help-seeking landscape.