Investigating psychotherapists’ attitudes towards artificial intelligence in psychotherapy

BACKGROUND: The increasing prevalence of mental health disorders, compounded by a global shortage of psychotherapists, highlights the need for innovative solutions such as Artificial Intelligence (AI) or Machine Learning (ML) applications. These technologies have demonstrated potential in diagnostics, treatment personalization, and therapy optimization. However, their integration into psychotherapeutic practice requires understanding psychotherapists' attitudes toward AI/ML, which remains underexplored. This study aims to investigate these attitudes, focusing on factors influencing AI acceptance and perceived usefulness. METHODS: A cross-sectional survey was conducted among 181 licensed psychotherapists in Germany, recruited via the German Psychotherapeutical Association's online directory. The survey assessed attitudes toward AI/ML, technical affinity, and perceptions of AI's utility across psychotherapeutic tasks. Hierarchical regression analyses were used to identify predictors of AI acceptance. RESULTS: Positive attitudes toward AI/ML were significantly predicted by its perceived usefulness in conducting diagnoses and creating personalized treatment plans. Empathic support, while rated lower in terms of enhancing therapy, was still a significant predictor across all groups. Technically affine therapists associated AI with benefits in diagnostics, whereas non-affine therapists emphasized empathic support and relapse prediction. CONCLUSION: Negative attitudes toward AI/ML application are discussed in a frame of fears of professional replacement and limited understanding of AI/ML technologies. Overall, 40% of the sample self-identified as not technically inclined, suggesting a knowledge gap in AI/ML that might influence attitudes. Education could emerge as a critical factor in addressing fears and uncertainties surrounding AI/ML. Emphasizing the irreplaceable human qualities of psychotherapists may also alleviate fears of obsolescence.