Dive Brief:
- Microsoft is upgrading its artificial intelligence assistant for providers to combine voice dictation, ambient listening and generative AI capabilities, the tech giant announced Monday.
- The tool, branded as Dragon Copilot, will be able to help clinicians with a variety of tasks, like creating a structured note after a patient visit, automating referral letters or helping doctors find medical information from their own records and outside sources, according to the company.
- Microsoft plans to launch Dragon Copilot for the U.S. and Canada in May before rolling it out in other countries. Dragon Copilot will be available to health systems currently using Microsoft’s voice documentation products as an upgrade. The company declined to comment what clients plan to adopt Copilot at this time and did not answer questions on Copilot’s cost.
Dive Insight:
Since acquiring clinical documentation company Nuance for nearly $20 billion in 2022, Microsoft has steadily built the business into a suite of clinical assistant tools, including Dragon Medical One, which uses speech recognition software to document patient information, and AI-backed clinical assistant DAX Copilot, an AI-backed clinical assistant and decision support tool.
Microsoft says its AI-backed products have been received warmly by a healthcare industry struggling with heavy administrative burden.
Dragon Medical One has been used by more than 600,000 clinicians to document billions of patient records, while DAX Copilot has generated more than 3 million doctor-patient conversations for 600 healthcare organizations over the past month, according to Ken Harper, Dragon and DAX Copilot’s general manager.
But Dragon Copilot yokes those products together — and backstops them with generative AI — in a move Harper called “by far the most exciting advancement that we’ve made to date” during a call with press.
The goal with Copilot was to “unify everything into one experience,” Harper said.
Before Copilot, doctors using Microsoft tools had to toggle between different task-specific applications. Now, clinicians can dictate notes and automate tasks in one place, according to the GM.

Dragon Copilot’s welcome page.
Courtesy of Microsoft
Copilot is also different because it uses AI to allow clinicians to query notes and other information from outside sources, like the Centers for Disease Control and Prevention or the Food and Drug Administration.
For example, a doctor could ask Copilot whether a particular patient should be screened for lung cancer, and Copilot’s answer will include links to those vetted sources to help doctors verify accuracy, Harper said.
Generative AI products, which can create original text and images, can occasionally make things up, one reason driving concern behind increasing uptake of such products in healthcare. Another reason is a lack of standardized oversight, a gap that private sector groups — the largest of which include or are led by Microsoft — say they’re working to fix.
Copilot’s ability to query medical sources is similar to a product Google launched last year called Vertex AI Search, which searches reams of medical text and images to answer questions from doctors and healthcare workers.
That’s not to mention a bevy of other AI-backed clinical assistants that have entered the market in the last few years from companies like Suki, Abridge and Nabla.
But Dragon Copilot is distinct, according to Harper.
“The biggest thing that’s really different here, that is unique to Dragon Copilot relative to other assistants, is just the scope of what’s available,” Harper said, listing dictation, ambient listening, natural language processing, custom template creation and information search.
“So many of these other capabilities are out there [but] are specific to a single skill,” he said.
Microsoft is one of many tech giants looking to turn a profit by helping healthcare workers drowning in administrative work.
The company in October also launched a number of other tools for the industry, including help for organizations wanting to build their own AI assistants, foundation models for medical imaging and a healthcare data analysis platform.
Google has also been focused on beefing up Vertex AI Search in advance of a major health IT conference that kicks off Monday, announcing that it can now pull information from images in addition to text.
And on Friday, Salesforce launched a library of prebuilt AI agents meant to automate a variety of healthcare tasks.