By Alex Smith |
Brown University Health has scaled Microsoft Dragon Copilot to more than 400 clinicians and used Microsoft Copilot Studio to build 24 AI agents supporting emergency department guidance, patient routing, translation, scheduling and back-office operations.
The Rhode Island-based health system serves a population of approximately a million people. Since 2024 it has acquired two Massachusetts hospitals and merged with a large academic physician practice, expanding its footprint while facing growing patient demand, labour shortages, financial constraints and clinician burnout.
“We are under constant financial pressure, and we want to meet and exceed expectations for how we deliver high-quality care,” said Adam Landman, senior vice president and chief digital information officer at Brown Health. “We see AI as a key enabler, one part of a broader set of changes needed to transform how we deliver care.”
Documentation has been identified internally as one of the leading contributors to burnout, particularly in emergency medicine, with physician burnout rates in some specialties above 40 per cent.
To address this challenge, Brown Health adopted Dragon Copilot, an ambient clinical documentation tool that records patient encounters with permission and drafts structured clinical notes, as the first major application under a newly formed AI Center of Excellence. The team spent its initial six months on governance, developing a risk-based framework for evaluating new AI tools before expanding deployment.
An early pilot involving 420 clinicians across multiple specialties found reduced after-hours documentation time and lower reported cognitive burden among participants. Physician Anthony Napoli, executive vice chair of emergency medicine, helped lead adoption within the emergency department, which the organisation says became one of the first fully ambient-enabled emergency physician groups in the country.
“Now, I can focus on what I think is going on with you and hearing you and being with you,” said Napoli. “The documentation is happening in parallel with engagement with the patient.”
Alongside Dragon Copilot, Brown Health deployed Microsoft 365 Copilot across departments for tasks including inbox management, meeting summarisation and contract analysis. The organisation’s AI Governance Council used the Researcher agent within Microsoft 365 Copilot to develop its responsible AI policy, which was largely finalised within four months.
New AI use cases are evaluated through a risk-based process before deployment, weighing clinical and operational risk against potential value. “It’s not just about whether a tool works,” said Ali Chambers, a data scientist at Brown Health’s AI Center of Excellence. “It’s about what happens if it doesn’t.” Lower-risk tools, such as internal knowledge assistants, move through prototyping and small pilots more quickly, while tools that could influence clinical decisions undergo more extensive validation. The centre continues to track adoption, usage patterns and clinician trust after deployment.
Brown Health is now applying Copilot Studio to build additional AI agents, with current use cases spanning emergency department guidance, care routing, translation, scheduling and operational processes.
“It’s going to bring back the joy to the practice of medicine,” said Landman. “Clinicians, including me, went into this because they want to help people. Most did not go into it because they want to interact with computers.”