Related Experiment Videos
Artificial Intelligence in Psychiatric Graduate Medical Education
Manal Khan1, Juliet Edgcomb1, Jonathan Heldt1
1David Geffen School of Medicine at University of California Los Angeles (UCLA) and UCLA Semel Institute for Neuroscience and Human Behavior, Los Angeles, CA, USA.
Objective:
Psychiatry graduate medical education (GME) faces converging pressures of increasing clinical demand, rising administrative burden, and workforce burnout. The emergence of artificial intelligence (AI), particularly large language models (LLMs), offers new possibilities for expanding educational capacity and personalizing training. However, the relational and narrative foundations of psychiatric practice create unique challenges for responsible integration. This review defines opportunities and risks of AI across key domains of psychiatry GME.
Methods:
Psychiatric education leaders applied the Josiah Macy Foundation's framework for AI in medical education to four domains of psychiatry GME: recruitment, didactic development and clinical learning, assessment and feedback, and program evaluation.
Results:
AI has potential to reduce administrative burden, augment clinical reasoning instruction, enable more consistent formative assessment, and support data-driven program improvement. Risks include overdependence, erosion of documentation and formulation skills, bias amplification, privacy vulnerabilities, and weakening of the therapeutic and supervisory relationships central to psychiatric training.
Conclusions:
Responsible AI integration in psychiatry GME requires staged introduction aligned with trainee developmental level, faculty engagement, human oversight in evaluation and decision-making, transparent communication of AI use, robust data governance, and prioritization of tools that deepen rather than replace human connection. Thoughtful implementation can support, rather than supplant, the relational and reflective practices at the heart of psychiatric education.
