"Do you mind if I record this on my phone?"

Ambient AI is the ability for a clinician to turn on an audible recording, and the natural language processing model then turns that person-to-person conversation into a generated note. Not just transcribing but condensing and even queuing up orders for providers. Ambient documentation has real momentum on the physician side. A large multisite study published in JAMA this year tracked more than 8,500 ambulatory clinicians across five academic health systems and found AI scribe adoption saved about 16 minutes of documentation time per eight-hour clinical day.[1]

Nursing is a different problem, and to their credit, some organizations are actually treating it that way instead of just repackaging the physician tool.

Mayo Clinic is the example I point to most often. Working with Abridge, Mayo built an ambient documentation tool specifically for inpatient nursing, co-designed by nurses rather than adapted from a physician product.[2] It captures nurse-patient conversations during care and converts them into structured draft documentation that nurses review and finalize. Involving nurses in the design instead of handing over a provider-created tool and calling it done is the right answer.

Here is where I get skeptical.

Almost every ambient nursing tool on the market right now, including Mayo's, is built around the same core use case: narrating a head-to-toe assessment out loud so it converts into flowsheet documentation. Ambience Healthcare's flowsheet product describes its function plainly as capturing the narrated patient assessment into structured flowsheets.[3] That is the industry's answer to nursing documentation burden right now, and I do not think it is the right first answer.

A head-to-toe assessment is not just data collection. It is a clinical exam that most nurses do through touch, observation, and pattern recognition as much as conversation. Done in their heads. Using critical thinking skills to mentally note what is present, what has changed while quietly making other observations. Asking a nurse to narrate that assessment out loud, in real time, for an AI to capture, changes what the assessment actually is. It is not a documentation shortcut. It is a different clinical task wearing a documentation shortcut's clothes.

Two use cases make more sense to me.

Neither is where most vendor energy is going.

Home health. Home health nurses have some of the worst documentation burden in nursing, often finishing OASIS assessments and visit notes from a car or a kitchen table, with no support staff and no shift handoff to lean on. A small number of vendors are starting to build specifically for this, generating OASIS assessments and visit notes directly from the encounter.[4] That is a setting where ambient capture solves a real, acute problem instead of changing how a clinical exam gets performed.

Admission assessments. An admission is already a long, conversation-heavy encounter, and it happens once. It is naturally suited to ambient capture in a way ongoing shift assessments are not, because it is not competing with a task the nurse needs to perform differently just to make the technology work.

There is also a basic usability problem nobody has solved yet.

Every ambient nursing tool I am aware of still requires a manual step to start recording: a button, an app, a session that has to be opened before the conversation begins. That is a meaningful gap between where the technology is and where it needs to be for inpatient nursing specifically, where hands are rarely free, and workflow interruption is exactly what this technology is supposed to eliminate.

The version I actually want is a smart room that is always listening in context, not a tool I have to remember to activate. I want to say, "I just emptied 350 mls of clear yellow urine from the Foley," while I am doing it, and have that land in the right flowsheet field without touching a screen or pressing a button first. That is not close to available yet. But it is the actual target, and it is worth naming even though the industry is not there.

None of this is a case against ambient AI for nursing. It is a case for being honest about where the technology actually is right now, versus where the vendor marketing says it is, and making sure nurses, not just vendors, are the ones deciding which problem gets solved first.

---

References

1. Rotenstein L, et al. Changes in Clinician Time Expenditure and Visit Quantity With Adoption of Artificial Intelligence-Powered Scribes: A Multisite Study. JAMA. Published online April 1, 2026. doi:10.1001/jama.2026.2253
2. Abridge. Mayo Clinic and Abridge Reimagine Nursing with Ambient AI. Published 2026. Accessed August 2026. https://www.abridge.com/blog/mayo-clinic-ambient-ai-nursing
3. Ambience Healthcare Introduces Nursing Suite to Extend System of Intelligence Across Inpatient Nursing. BusinessWire. Published June 23, 2026. https://www.businesswire.com/news/home/20260623636774/en/Ambience-Healthcare-Introduces-Nursing-Suite-to-Extend-System-of-Intelligence-Across-Inpatient-Nursing
4. Lime Health AI. Ambient AI Scribe for Home Health Nurses. Accessed August 2026. https://getlimeai.com/ambient-ai-scribe-home-health/