Medical AI Moves From Research to Hospital Use
Health-system leaders say clinical AI now has the deployment capacity its decades of research promised, with outcome measurement as the test.
Updated on Nov. 2, 2026 in Artificial Intelligence
Live Poll
Do you trust AI to be used safely in your own medical care?
Philip Payne, chief health AI officer at BJC HealthCare and WashU Medicine, warned health systems against treating medical AI as a new field, speaking on a panel at Newsweek's AI Health Summit in New York City. The remarks describe current AI deployment as a maturing practice built on research dating to the 1950s and 1960s, not an announced product launch.
Why it matters
Payne argues the ability to deploy AI at scale has finally caught up with decades of research, a shift that moves clinical AI from experiment toward routine operations. The point signals where health technology is heading: integration into hospital workflows rather than standalone novelty.
The evidence cited is qualitative expert testimony rather than a measured trial result, with the 2018 heart rhythm example and late-2022 adoption timeline as the only dated anchors. The panel did not supply benchmark figures, and the claims about deployment readiness remain unquantified.
The players
Philip Payne
Chief health AI officer at BJC HealthCare and WashU Medicine, a PhD-trained informatician with decades of AI experience.
Rohit Chandra
Executive vice president and chief digital officer at Cleveland Clinic, who described current AI as the most immature version health systems will see.
Bilal Muhsin
Executive vice president and president of the connected care segment at BD, a medical technology company.
The details
Payne described clinical AI as systems that process patient data to flag patterns clinicians might miss, such as computerized detection of abnormal heart rhythms. At Houston Methodist Hospital, continuous vital-sign monitoring feeds a virtual operations center, where staff detected a concerning signal and alerted a bedside nurse. Payne said his program will not start an AI project without a clear pathway toward measuring outcomes.
Timeline
The 1950s and 1960s saw early AI research that Payne says health systems risk overlooking.
In 2018, Payne cited computerized identification of abnormal heart rhythms as AI operating in clinics.
Most health systems began seriously considering organization-wide AI deployment in late 2022.
The Tech Race
The St. Louis-based BJC and WashU Medicine Center for Health AI is positioned to move academic AI research into hospital settings. Rival health systems such as Cleveland Clinic and Houston Methodist are pursuing their own continuous-monitoring and digital-operations approaches.
Patients at health systems adopting AI monitoring may see alerts routed through virtual operations staff to bedside nurses. The source does not give availability timelines or pricing for these tools at any St. Louis hospital.
The takeaway
Watch whether BJC and WashU Medicine's Center for Health AI publishes outcome measurements for any deployed system, since Payne tied programs to measurable results. Health systems that report outcomes against that standard will show whether decades-old research is translating into clinical benefit.
Further reading
For more on how hospitals are putting these tools to work, see Artificial Intelligence coverage across our tech desk.
Source note: This article includes information reported by Newsweek.
Live Poll
Do you trust AI to be used safely in your own medical care?
