An Interpretive Description Study of Clients’ Experiences of Integrating Artificial Intelligence-Based Mental Health Support With Face-to-Face Counseling

Abstract

Objective: This study aimed to explore and interpret clients’ experiences of integrating artificial intelligence (AI)-based mental health support with face-to-face counseling in England.

Methods and Materials: This qualitative study employed an Interpretive Description design and included 24 adult clients in England who had attended at least four face-to-face counseling sessions and had concurrently used an AI-based mental health support tool for a minimum of four weeks. Participants were recruited purposively to achieve variation in demographic characteristics, counseling experiences, and patterns of AI use. Data were collected through individual semi-structured, in-depth interviews focused on motivations for using AI, perceived benefits and limitations, disclosure, trust, emotional safety, continuity of support, and the relationship between AI use and counseling. Interviews were audio-recorded and transcribed verbatim. Data collection and analysis proceeded concurrently using iterative coding, constant comparison, reflexive memoing, and progressive development of clinically meaningful themes.

Findings: Analysis generated six interrelated themes: AI as an in-between-session holding space; AI as a rehearsal and translation space for counseling; greater control and reduced interpersonal exposure; recognition of the emotional and contextual limitations of AI; negotiation of trust, accuracy, and safety; and reshaping rather than replacing the therapeutic relationship. Four patterns of integration were also identified: complementary integration, selective parallel use, substitutive drift, and conflict-and-repair. AI was experienced as most beneficial when it supported reflection, preparation, disclosure, and continuity between counseling sessions, whereas less constructive use emerged when convenience facilitated avoidance, AI-generated guidance was accepted uncritically, or important AI-mediated experiences remained disconnected from counseling.

Conclusion: AI-based mental health support may function as a useful intermediary psychological space between private self-reflection and interpersonal therapy, but its therapeutic value depends on critical evaluation, appropriate boundaries, transparency, and continued reliance on human counseling for contextual understanding, relational depth, accountability, and complex clinical judgment.