UpToDate launches Expert GenAI to support clinicians with evidence-based answers, reducing risks from unreliable general-purpose LLMs.
UpToDate, part of Wolters Kluwer Health, has launched Expert AI, a GenAI tool designed to support clinical decision-making. Unlike general-purpose LLMs such as ChatGPT or Gemini, the system is trained on peer-reviewed medical literature. As well as the insights are from more than 7,600 medical experts. This aims to eliminate hallucinations and provide clinicians with accurate, citation-backed recommendations.
The challenge Expert AI addresses is critical: physicians increasingly rely on public LLMs, but those tools often deliver unsafe or inconsistent advice. Tests showed dangerous errors, including recommending unnecessary surgeries or improper drug withdrawal. Expert AI avoids these pitfalls by grounding responses in vetted content, ensuring reliability where mistakes can have life-or-death consequences.
The tool provides multi-step reasoning, clearly listing assumptions and linking every recommendation to original research. Clinicians can query it with natural language and receive structured, transparent answers. For example, it explains family cancer history scenarios step by step, or advises on sensitive vaccine hesitancy cases with contextual recommendations. This structured approach enhances trust and makes clinical reasoning visible to users.
Already tested with select institutions, Expert AI will roll out to 250,000 users in North America starting October 2025, before expanding globally. While latency remains a concern, accuracy has been rated high by early testers. For Wolters Kluwer, this marks a shift in GenAI’s role in healthcare: moving beyond administrative tasks to direct clinical support. The company believes Expert AI can reduce reliance on unsafe general models, helping doctors make faster, safer, and more informed decisions.
