Large Language Models in Clinical Medicine: Toward Safe and Responsible Deployment

Abstract

Large language models (LLMs) are rapidly entering clinical practice, offering opportunities to improve healthcare delivery and clinical decision support. However, concerns regarding hallucinations, patient safety, regulatory oversight, and equitable implementation remain significant. This letter highlights key challenges and proposes safeguards to support the safe, transparent, and responsible integration of LLMs into healthcare.