Diagnosis (general) · AI doctors and chatbots
AI note-takers are linked to doctors billing more
A Peterson Health Technology Institute report says AI scribes, now used across big US health systems, are pushing billing up. Two peer-reviewed studies of past records found a rise in clinician billing: about $3,044 a year more per doctor who adopted the scribe in one (at a single health system, UCSF, priced at 2025 Medicare rates), and about $167 a month in the other, which the authors called 'numerically minimal'. Neither shows the scribe caused the rise or measured what patients paid. Rating: Early research.
How far away is this?
AI scribes aren't a medical product that needs approval. The billing evidence comes from studies of past records plus a think-tank report, so we rate it Early research, the level we give studies of past records.
Get the weekly How Far
One email every Friday morning (Sydney time) on which medical AI actually reached patients, and what it means for you. Free, in plain English.
Where is it approved?
| Region | Regulator | Status | Notes |
|---|---|---|---|
| United States | FDA | Not applicable | Documentation software; generally not regulated as a device. |
| European Union | CE mark / EMA | Not applicable | No register record checked. |
| United Kingdom | MHRA | Not applicable | No register record checked. |
| Australia | TGA | Not applicable | No ARTG record checked. |
Admin software; not a regulated medical device in most places. "Register not searched yet" means we haven't searched that regulator's register; it doesn't mean the product is or isn't available. Approval in one country isn't approval in another.
What is it?
An AI scribe listens to your appointment and writes the doctor's notes. Fuller notes can justify billing a visit at a higher level.
What's the evidence?
The Peterson Health Technology Institute (PHTI) report says that in a recent survey of large health systems, all of them reported using AI tools for clinical documentation.[1]
Evidence stage: Announced only · #c1
The PHTI report says that after one multihospital system brought in an AI scribe, top-level (Level 5) visits rose 5% and Level 4 visits rose 7% for returning patients, adding about $1,004 in revenue per clinician each month. In plain words: the same visits were billed higher. The report doesn't name the system, and we haven't seen the analysis itself.[1]
Evidence stage: Announced only · #c2
PHTI says that under today's pay-per-activity system, AI is set to drive spending up rather than cut waste.[2]
Evidence stage: Announced only · #c3
A study at UCSF Health covering more than 1.2 million visits found that doctors with an AI scribe billed 1.81 more 'RVUs' a week than doctors without one. An RVU is the unit Medicare uses to price a doctor's work; priced at 2025 Medicare fee rates, that's about $3,044 more a year per doctor who adopted the scribe, at one health system (UCSF). Insurers didn't turn down any more claims. It's an analysis of past records, so it can't prove the scribe caused the change.[3]
Evidence stage: Early research · #c4
A study of 8,581 clinicians at 5 US academic health systems, published in JAMA in April 2026, found AI scribe users brought in about $167 more a month each in visit fees (priced at 2025 Medicare rates). They saw about half a visit more a week; the authors say the extra money may come from those extra visits or from visits being coded at higher levels. They also spent 16 fewer minutes on notes for every 8 hours of booked patient time. The authors call the extra money 'numerically minimal'. It's also an observational study, so it can't prove cause.[4]
Evidence stage: Early research · #c5
What does it mean for me?
If you're in the US, check your bills. If a visit is coded higher than you expected, you can ask the clinic to explain the code. You can also ask whether an AI scribe is recording your visit.
What could I ask my doctor?
- Is an AI scribe being used in my appointment?
- Can you explain the billing code for this visit?
These are general starting points, not a recommendation to choose any particular test or treatment.
Short videos, under a minute
We make short, plain-English videos about AI in medicine, including what it can't do yet. Follow How Far for the next one.
Change history
- : First published.
- : Editor pass before publication: labels aligned to site categories; sources and rating reviewed.
General information only, not medical advice. Talk to your doctor about your situation. AI-assisted source check. Not clinically reviewed.
References
- PHTI report: Driving Adoption of High-Value AI in Healthcare (Oct 2026). phti.org (2026). Service evaluation report. https://www.phti.org/wp-content/uploads/2026/10/PHTI-Driving-Adoption-of-High-Value-AI-in-Healthcare.pdf
- PHTI announcement (14 Oct 2026). phti.org (2026). Press release (claim). https://www.phti.org/announcement/new-phti-report-provides-roadmap-to-scale-high-value-ai-in-healthcare/
- Holmgren et al. Ambient AI scribes and physician financial productivity. JAMA Network Open (2026). doi.org (2026). Peer-reviewed observational study. https://doi.org/10.1001/jamanetworkopen.2025.53233
- Rotenstein, Holmgren et al. AI scribes, clinician time and visit volume: a multisite study. JAMA (2026). jamanetwork.com (2026). Peer-reviewed observational study. https://jamanetwork.com/journals/jama/fullarticle/2847319
Publisher shown as the website hosting each source. Dates are given where the source label includes them.