Loading...
Clinical artificial intelligence (AI) tools retrieve knowledge from medical sources such as review articles, research studies, and guidelines when generating answers to clinical questions. This process, called “retrieval augmented generation,” aims to improve answer accuracy and provide citations for AI’s judgments. To assess their accuracy in answering clinical questions, investigators compared two clinical AI tools (OpenEvidence and UpToDate Expert AI) with three different general-purpose AI tools (OpenAI GPT-5.2, Google Gemini 3.1, and Anthropic Claude Opus 4.6).
The general-purpose AI tools significantly outperformed the clinical AI tools on answers to 1000 randomly-sampled “USMLE-style” board practice questions, with a…