Understanding psychiatrist readiness for AI: a study of access, self-efficacy, trust, and design expectations

Abstract

BACKGROUND: Artificial intelligence (AI) is receiving growing attention in psychiatric practice, yet psychiatrists vary considerably in how they perceive its benefits, risks, and clinical usefulness. Successful implementation depends not only on technological performance but also on clinicians' readiness, including their access to AI, confidence in using it, trust in its reliability, and expectations for its design. Evidence on these dimensions remains limited in China. OBJECTIVE: This study examined Chinese psychiatrists' readiness for AI across four dimensions-access, self-efficacy, trust, and design expectations-and explored variation across demographic and professional subgroups. METHOD: A cross-sectional online survey was distributed through the WeChat platform from March 20 to 22, 2025. Eligible participants were licensed psychiatrists engaged in clinical practice. A total of 134 valid responses were obtained from clinicians across diverse provinces, hospital tiers, and professional roles. Descriptive analyses and group comparisons were conducted. RESULTS: Psychiatrists reported broad exposure to AI, though knowledge was largely acquired through informal rather than structured sources. Overall self-efficacy was moderate, with higher confidence observed among younger clinicians, male clinicians, and those who had received AI-related training. Trust in AI was generally positive, and department heads expressed especially strong confidence in its future role. Across subgroups, psychiatrists consistently prioritized AI applications that reduce administrative and documentation burden, while expressing lower expectations for AI use in communication, assessment, or psychotherapy. CONCLUSION: Chinese psychiatrists demonstrated cautious optimism and moderate readiness for AI. Limited access to formal training and subgroup differences in confidence highlight the need for targeted capacity-building. The strong preference for documentation-support tools underscores the importance of designing AI systems that integrate smoothly into clinical workflows while preserving the human-centered nature of psychiatric care.

Ähnliche Arbeiten