AI Resilience Report forHospitalists

Good news first: AI in hospital medicine right now mostly looks like a smart assistant for hospitalists, not a replacement. Artificial intelligence is entering hospital medicine at a moment when the field needs it most, and for many hospitalists, interest in AI has grown out of day-to-day pressures: notes completed late at night, fragmented data streams, and a sense that cognitive work is being squeezed by administrative tasks. One of the biggest examples is "ambient AI scribes" that listen to doctor-patient conversations and write the notes automatically, which Optum Advisory experts say more health systems are adopting ^[1] to give physicians more time at the bedside.

AI is also helping with the behind-the-scenes work of running a hospital unit — a recent Frontiers in Medicine study described how Northwell Health used AI-driven patient assignment software ^[2] to cut lead hospitalists' morning assignment process from 2.5 hours to about 30 minutes. On the clinical side, an NPR-reported Science study found an OpenAI reasoning model matched and often outperformed two experienced physicians on real ER cases ^[3], though it was tested as a decision aid, not as a stand-alone doctor. Studies in JAMA Network Open and Nature Medicine show physicians using AI assistance improved diagnostic accuracy on complex cases, but the same studies identified automation bias — the tendency to over-rely on algorithmic suggestions, even when incorrect.

Adoption is moving fast because the tools are commercially available and the pain points are huge. The American Medical Association's 2026 survey found that 81% of physicians now use AI professionally — more than double the 2023 rate ^[4], with top uses including summarizing research, drafting discharge instructions, and documenting visits. OpenEvidence, an LLM-based system that references published medical literature, is now used by more than 40% of U.S. physicians for point-of-care decision support.

Hospitals are interested because labor is expensive and burnout is real — the American Hospital Association highlights health systems like Mayo Clinic using AI-enabled stethoscopes that helped diagnose twice as many cases of pregnancy-related heart failure ^[5]. But things will likely slow down where safety, ethics, and trust matter most. AMA CEO John Whyte said it is critical that augmented intelligence be designed to enhance — not replace — physicians, and that tools must be safe, effective, and used responsibly to truly improve patient care.

Hospital medicine still depends on uniquely human skills: bedside communication, ethical judgment, supervising residents, and managing competing priorities under pressure — things AI helps with but cannot truly own. For young people curious about this field, hospital medicine looks less like a career being replaced and more like one being upgraded, where doctors who know how to work with AI will be especially valuable.