PERCEIVED THERAPEUTIC BOND IN INTERACTIONS WITH MENTAL HEALTH CHATBOTS: A NARRATIVE REVIEW

Abstract

Background: The rapid development of digital mental health interventions has led to the increasing use of AI-based chatbots as accessible alternatives or complements to traditional psychotherapy. While the therapeutic alliance is a well-established factor influencing treatment outcomes in conventional settings, its applicability to human–chatbot interactions remains unclear. Objective: This narrative review aims to examine existing evidence regarding the subjective experience of therapeutic bond in interactions with mental health chatbots and to explore its characteristics and potential clinical relevance. Methods: A narrative review of selected literature was conducted using databases such as PubMed and Google Scholar. Studies focusing on users’ subjective experiences and relational aspects of interaction with conversational agents were included. Results: Available evidence suggests that users can develop a form of therapeutic bond with mental health chatbots, characterized by perceived empathy, support, and emotional engagement. Both quantitative and qualitative studies indicate that users may attribute human-like qualities to chatbots and describe interactions in relational terms. However, the intensity and nature of this bond vary across individuals, with some users perceiving the interaction as purely instrumental. The bond appears to be limited in depth and may lack key features of traditional therapeutic relationships, such as reciprocity and genuine emotional responsiveness. Conclusions: Therapeutic bond in chatbot-based interventions may represent a distinct and simplified form of relational engagement shaped by the characteristics of human–AI interaction. Further research is needed to assess its long-term development and its relationship to clinical outcomes.