Exploring self-disclosure in conversations with AI Chatbots vs human mental health support among Young Adults

Abstract

The increasing processes of artificial intelligence integration into mental health services result in the paradigm shift in the manner in which individuals seek emotional support and share their personal thoughts. Mental health chatbots powered by artificial intelligence provide anonymity, availability, and immediate conversation, either as an adjunct to human therapists or as an alternative to human therapists. However, there are no empirical studies on the process of self-disclosure in AI-mediated communication and traditional human therapy. In this qualitative study, the researchers examine the self-disclosure experiences of young adults when they interact with AI-powered mental health chatbots compared to human mental health professionals. It used a phenomenological research approach, and semi-structured interviews were conducted with the participants who had experience with both. To identify the patterns based on emotional experiences, depth of disclosure, and factors influencing the willingness of people to disclose, thematic analysis proposed by Braun and Clarke (2006) was employed for analyzing the data. The findings show a very clear set of differences between the two forms of relationship. The idea of AI chatbots was perceived as being very unpersonal and unemotional, a facilitator of superficial self-disclosure with very little emotional spilling. The main aim of the AI chatbots for the participants was to cope with stress, gain information, or demonstrate curiosity rather than emotional support. On the other hand, the relationships with human mental health professionals were characterized by warmth, empathy, validation, and psychological safety, which facilitated the disclosure of more and more authentic information about self, family, and emotional topics. The obstacles to self-disclosure were stigma and privacy concerns, and empathy and non-judgmental response were the elements that helped to overcome these obstacles. Overall, the study illustrates the complementary role of AI chatbots in mental health and the inability of the latter to substitute human interaction as it contributes to meaningful self-disclosure.