From promise to practice: insights into ChatGPT-4o use in child and adolescent mental health from professionals.

Background: Despite growing interest, empirical studies on ChatGPT-4o's clinical role in child and adolescent mental health remain scarce. This study explored child and adolescent mental health professionals' attitudes toward ChatGPT-4o, focusing on its clinical applications, ethical implications, and integration challenges. Methods: A sequential exploratory design was used, beginning with interviews to inform item generation. Finalized surveys were distributed online to 96 child and adolescent psychiatrists and 70 psychologists between April and May 2025. The instrument measured views across seven subscales and demonstrated strong internal consistency (α = 0.887 for child and adolescent psychiatrists; α = 0.903 for psychologists). Results: Overall, 47.9% of psychiatrists and 40% of psychologists reported prior use of ChatGPT-4o. Child and adolescent psychiatrists rated "Clinician-Facing Tool" and "Acting as a Therapist" most favorably, while psychologists expressed the most positive views toward "Bias" and "Profession". Both groups viewed "Ethical Issues" least favorably. Comparative analyses revealed that psychiatrists scored significantly higher than psychologists on the profession (d = 0.46), psychoeducational use within treatment (d = 0.43), patient-facing tool (d = 0.68), digital access and personalization (d = 0.55), and crisis prevention and safety planning (d = 0.69). Psychiatrists also showed greater positive views toward self-help and behavior change interventions (U = 2649.5, Z = -2.41, p = 0.016, r = 0.19). In contrast, psychologists rated bias more favorably, representing the largest observed difference (d = 1.56). Development priorities differed slightly: child and adolescent psychiatrists emphasized software support for diagnostic & treatment, system oversight, and ethics, while psychologists also prioritized system oversight and ethics. Conclusion: Mental health professionals show cautious optimism toward ChatGPT-4o, with discipline-specific emphases. While a measured interest in ChatGPT-4o's clinical integration, shared concerns around ethics highlight the need for role-specific guidelines and human oversight.