Speakers from academia, industry, government and nonprofits explored developing and implementing artificial intelligence in medicine responsibly, effectively and reliably.

A new JAMA Network Open study found that an AI-powered workflow for generating hospital discharge summaries was associated with reduced physician burnout and minimal reported safety risk during real-world clinical deployment. Authored by colleagues across DoM’s Divisions of Hospital Medicine and Computational Medicine, the Clinical Excellence Research Center, and Stanford colleagues.

An AI program put to the most rigorous tests in modern medicine aced its exams, and in fact performed better than human doctors on reasoning tasks such as making emergency room decisions.

One of the first studies conducted on AI’s ability to perform complex medical reasoning tasks has been published in “Science” magazine on Thursday. It has found that AI is often performing better than human physicians in complex medical diagnosis and reasoning.

In a new Science study, Ethan Goh, Evelyn Bin Ling, Jason Hom, Jonathan Chen, and colleagues from Harvard Medical School and Beth Israel Deaconess Medical Center evaluated the OpenAI o1 model against hundreds of physicians across real-world clinical scenarios. The model outperformed both clinicians and prior systems in diagnosis and management, underscoring its growing potential in clinical care.

Congratulations to Nigam Shah (center) on his election to the Association of American Physicians (AAP). He was recognized among colleagues at the AAP New Member Induction Dinner over the weekend — an honor that highlights his outstanding contributions to academic medicine and research.

In conference rooms and clinics across the country, leaders are asking the same question: How do we move from excitement about artificial intelligence to systems that actually work — safely, responsibly, and at scale? At the Stanford Division of Computational Medicine, a new program aims to answer that question.