Exploring user characteristics, motives, and expectations and the therapeutic alliance in the mental health conversational AI Clare®: a baseline study.

Abstract

This study examined the characteristics, motives, expectations, and attitudes of users interested in artificial intelligence (AI) self-help provided by the bot Clare®, a conversational AI for mental health support, and explored the development of a working alliance. A cross-sectional survey of 527 English-speaking self-referred users revealed high levels of anxiety (69%), depression (59%), severe stress (32%), and loneliness (86%). The participants expressed positive attitudes toward digital mental health solutions, with key motives including avoiding embarrassment (36%) and concerns about appearance in face-to-face consultations (35%). Expectations focused on emotional support (35%) and expressing feelings (32%). A strong working alliance was established within 3-5 days (Working Alliance Inventory-Short Report, M = 3.76, SD = .72). These findings highlight the potential of conversational AI in providing accessible and stigma-free support, informing the design of human-centric AI in mental health. Future research should explore long-term user outcomes and clinical large language model integration with traditional mental health services.