Künstliche Intelligenz in der Psychotherapie – Einstellungen und Kompetenzen ärztlicher und psychologischer Psychotherapeuten

BACKGROUND: Arificial Intelligence (AI) is increasingly becoming established in medicine, such as therapeutic chatbots in mental health. Despite this, knowledge about clinicians' views and skills is lacking. This study examines the attitudes toward AI and competencies of German physicians (ÄPT), psychotherapists (PPT) and psychotherapists in training (PiA) for the first time. METHOD: Cross-sectional study with online survey (June-July 2025) of 335 participants, recruited through professional associations and chambers. Attitudes were assessed using the AIAS‑4 scale, competencies using the MAILS scale. Statistical analysis using ANCOVA with age as covariate (Bonferroni-corrected). RESULTS: All professional groups showed moderate AI literacy (M = 4.9-5.9/11), e.g. recognizing and evaluating the use of AI and moderately positive attitudes toward AI (AIAS-Total; M = 5.6-6.3/10). After controlling for age (ANCOVA), professional group effects emerged only for attitudes (p = 0.012), with physicians showing more positive attitudes than both psychologist groups (d = 0.40-0.49) but not for AI literacy (p = 0.267). There was a strong generational effect: younger participants demonstrated higher AI literacy (r = -0.28) and more positive attitudes (items 1-3: r = -0.12 to -0.17). Development competency for AI was almost nonexistent. CONCLUSION: The competence of reflective use of AI appears primarily generation-dependent, not profession-specific. Age-adapted, cross-professional training programs are necessary to systematically integrate reflective competencies in education and continuing professional development.