User perception and self-disclosure towards an AI psychotherapy chatbot according to the anthropomorphism of its profile picture

Abstract

Studies have displayed that the favorable image of a psychotherapist increases the therapy outcome by encouraging the self-disclosure of clients. Recently, in a situation where AI psychotherapy chatbots were actively employed, this study examines how the degree of anthropomorphism in profile pictures of AI psychotherapy chatbots (company logo, robot character, human character, real human photo) affects user self-disclosure. Furthermore, this study verifies that the relationship between anthropomorphism and self-disclosure is mediated by the user's liking and mind perception towards the chatbot. The study found that the anthropomorphism of the psychotherapy chatbot profile picture did not directly affect the user's self-disclosure, and the higher the degree of anthropomorphism of the profile picture, the more self-disclosure when the user had a favorable feeling for the chatbot. On the other hand, when the user perceived the chatbot to have thoughts and emotions, the higher the degree of anthropomorphism of the profile picture, the lower the degree of self-disclosure. This study has significance in examining the user's perception according to the chatbot profile picture given the situation in which a psychotherapy chatbot will be used in the actual market. © 2023 Elsevier Ltd