Mahdi Naeim, Mohammad Narimani
The findings suggest that the responsible integration of generative AI into psychological practice requires balancing innovation with clinical oversight, ethical governance, and human judgment. The concepts of epistemic trust and the hybrid therapeutic alliance are proposed as novel conceptual frameworks for understanding clinicians' acceptance and responsible use of AI in mental health care. Future research should examine these concepts across diverse clinical and cultural contexts to inform implementation strategies, clinician training, and regulatory policy.
BACKGROUND: The rapid expansion of generative artificial intelligence (AI) in mental health care has raised critical questions regarding its acceptance, potential benefits, risks, and ethical implications for mental health professionals. This study explored in depth the perspectives of professionals from multiple countries on the integration of generative AI into psychological interventions.
METHODS: A qualitative design was employed, using written semi-structured interviews with 34 mental health professionals from the United States, Canada, the United Kingdom, Germany, Australia, New Zealand, Iran, and Türkiye. Data were analyzed through reflexive thematic analysis.
RESULTS: Four overarching themes were identified: (1) general perceptions characterized by enthusiasm alongside skepticism; (2) opportunities including expanded access to care, personalized support, and reduced clinical workload; (3) concerns related to response quality, the therapeutic alliance, and ethical and data security issues; and (4) future perspectives emphasizing accountability, complementary clinician-AI roles, and responsible innovation. Participants consistently highlighted the importance of algorithmic transparency, empirical validation, ethical governance, and clinician training. These findings informed the development of the concepts of epistemic trust and the hybrid therapeutic alliance as interpretive frameworks for understanding clinicians' evaluations of AI in psychological practice.
CONCLUSION: The findings suggest that the responsible integration of generative AI into psychological practice requires balancing innovation with clinical oversight, ethical governance, and human judgment. The concepts of epistemic trust and the hybrid therapeutic alliance are proposed as novel conceptual frameworks for understanding clinicians' acceptance and responsible use of AI in mental health care. Future research should examine these concepts across diverse clinical and cultural contexts to inform implementation strategies, clinician training, and regulatory policy.