Evaluation of the Therapist–Artificial Intelligence Integrated Approach in Mental Health Care: A Systematic Review

Abstract

Objective: This study aimed to evaluate the effectiveness, applications, benefits, and challenges of integrating artificial intelligence with therapist-delivered mental health care. Methods and Materials: This systematic review was conducted in accordance with the PRISMA guidelines. A systematic search of studies published between 2015 and 2026 was performed in PubMed, Google Scholar, ScienceDirect, SID, and Magiran using combinations of keywords related to artificial intelligence, psychotherapy, therapist–AI collaboration, therapist assistants, chatbots, co-therapists, and human–AI collaboration. Original Persian- or English-language studies examining artificial intelligence as a supportive tool for mental health professionals were eligible for inclusion. After duplicate removal, title and abstract screening, and full-text eligibility assessment, 11 studies were included. Data on study design, participants, interventions, artificial intelligence systems, assessment instruments, and outcomes were extracted, categorized, and narratively synthesized. Findings: Compared with conventional care, therapist-delivered interventions supported by artificial intelligence were associated with better clinical outcomes, greater treatment attendance and adherence, and lower dropout rates. Artificial intelligence-assisted approaches facilitated emotional and experiential expression, externalization of internal experiences, greater awareness of personal psychological states, and the application of coping skills and cognitive-behavioral techniques in everyday life. Among therapists and peer counselors, artificial intelligence support enhanced empathy, self-efficacy, confidence, and perceived competence; improved the length and quality of responses; accelerated clinical note preparation; and supported session planning, homework customization, and access to assignment histories. Artificial intelligence systems also assisted in identifying distress during sessions and recalling appropriate counseling strategies. Nevertheless, concerns were consistently raised regarding confidentiality, data security, algorithmic and cultural bias, potential misuse of client information, excessive reliance by inexperienced professionals, and the need for continuous human supervision. Conclusion: Artificial intelligence can improve the effectiveness, accessibility, and operational efficiency of evidence-based mental health care when used as a therapist-supervised assistant rather than a replacement for clinicians. Its responsible implementation requires professional oversight, ethical governance, transparent clinical protocols, and robust safeguards for privacy and data protection.