Like much of modern life, medicine is rapidly being reshaped by AI. The fastest-growing -- and potentially most consequential -- application in healthcare is a technology known as "ambient intelligence."
Ambient intelligence is described as an invisible AI operation. It works in the background, activated when it receives an external cue, such as when someone speaks. Clinicians and large health systems are quickly embracing a host of competing ambient intelligence applications to "invisibly" record conversations during a patient visit.
The new tools transcribe the interactions into text and seamlessly integrate them into a patient's electronic health record (EHR), eliminating the need for a clinician to type in observations during a patient visit and subsequent manual charting.
These ambient intelligent tools use AI voice recognition technology to automatically write prescriptions, order lab tests, handle billing, and insurance claims. A few pilot programs can even suggest diagnosis and treatment options when the recorded visit is linked with lab results, in-office blood pressure and oxygen tests, and clinical observations.
"As applications of artificial intelligence (AI) spread rapidly, ambient scribes are poised to become one of the fastest technology adoptions in healthcare history," according to a new report from the Peterson Health Technology's AI Taskforce. "There is no technology in recent memory that has been adopted more enthusiastically by clinicians or has scaled up so uncharacteristically fast, absent a regulatory mandate."
But with the enthusiasm comes concern. The AI scribe field is unregulated, so an application's accuracy isn't being independently monitored. In one recent study, researchers reported scribes produced inaccuracies that the authors say will require "vigilance." Notes must be reviewed and should be viewed as an assistant, not a relied-upon replacement, the authors stated.
Hundreds of millions of investment dollars are backing 60 different doctor assistant products that have emerged in just the past few years, according to industry sources. The ones gaining widest use are Abridge, Athelas, Augmedix, DAX Copilot, DeepScribe, Heidi, and Suki. Microsoft is developing an ambient intelligence solution for hospitals that captures nursing care operations and converts them into patient orders and chart documentation, readily accessible to attending physicians and other medical staff.
Last year, Abridge raised $300 million from investors while Suki raised $70 million. In June, San Francisco-based Ambience raised a record $243 million from investment bankers. Its platform integrates directly with Athenahealth, Epic, Oracle Cerner, and other EHRs, according to the company.
The widely used Epic EHR has entered the field with its own ambient program. While Epic previously partnered with Microsoft's DAX Copilot and Abridge to create an ambient tool called Art, its AI Charting, introduced in March, is its own proprietary product.
"The idea that you can have software just listen to a normal everyday conversation in our offices and make it into a medical note (is)truly transformative," said Eric Boose, MD, associate chief medical information officer at Cleveland Clinic in Brecksville, Ohio. Cleveland Clinic onboarded about 1000 physicians within 8 days of launching the technology last year and currently has 4000 of 6000 eligible clinicians using it, Boose said.
For many physicians, the software eliminates the most tedious part of their job, Boose said.
The new programs improve upon earlier generation AI scribes that simply recorded conversations for medical records. Newer versions have the ability to weed out less relevant chit chat, such as discussions about the weather or how a patient's children are faring.
A much-quoted study published in the JAMA last fall found the new tools helped remedy clinician burnout by reducing documentation time. They also allowed doctors to focus their attention on patient interactions rather than typing in electronic notations during a visit, something patients also report liking, a few surveys found. Doctors in the study said the new AI scribes reduced so-called "pajama time" required for them to input notes from patient visits earlier in the day.
"The promise of these solutions to reduce burnout and improve workflows is driving the fast adoption among physician practices and health systems," the Peterson task force found.
"Burnout's a multipronged issue and there's no single solution, but we know that clinical documentation is a large contributor to burnout -- especially in outpatient arenas," Jason Misurac, MD, MS, a clinical associate professor with University of Iowa Health Care in Iowa City, Iowa, said in a recent American Medical Association webinar.
Kaiser Permanente is conducting one of the largest deployments, with Abridge's technology used by over 25,000 clinicians, including its primary care doctors, specialists, and pharmacists across all 40 of its hospitals and 616 medical offices.
Kaiser Permanente Medical Group's clinicians saved more than 15,700 hours in 1 year when they used an ambient scribe -- the equivalent of 1794 working days -- compared with nonusers, according to a study published in The New England Journal of Medicine in March 2025.
Mass General Brigham uses ambient AI for more than 2500 clinicians, it reports. UCSF Health & Memorial is using Ambience Healthcare, and Ochsner Health is offering DeepScribe to its 4700 clinicians.
In February 2025, Cleveland Clinic announced it would roll out Ambience Healthcare across its ambulatory practices after a successful pilot project covering more than 80 specialties. The organization onboarded about 1000 physicians within 8 days of launching the technology and currently has 4000 of 6000 eligible clinicians using it.
While the scribes are by far the most common AI tool gaining acceptance, other programs are being rolled out. For example, Mass General Brigham implemented CodaMetrix's AI-powered autonomous medical coding platform that has achieved a 74% automation rate for radiology test results. Its use has led to 58.7% reduction in claims denials, according to the institution. The estimated $750,000 in cost savings enabled the system to redeploy 12 full-time coders to other departments, while increasing annual growth in payments by 12%, according to CodaMetrix data.
Clinical Decision Support (CDS) programs are another AI tool being tested. They are a digital health tools integrated into EHRs that provide clinicians and patients with evidence-based information -- such as alerts, order sets, or diagnostic advice. Because CDS tools have a direct effect on patient care, the industry said they require more testing and alignment with an organization's clinical guidelines. UC San Diego Health developed an in-house AI algorithm called COMPOSE that reduced sepsis-related mortality by 17% in the emergency department, according to a 2024 study by university researchers.
"Despite these advancements and the clear potential of such tools, their widespread implementation remains a significant challenge," the study found. "A primary hurdle is the absence of standardized pathways for integrating these new technologies into most healthcare institutions."
Michael Waldholz is a veteran health and medical science writer.
