Industry · Medical practices and outpatient clinics · 2026

AI for medical practices: shorter notes, answered phones, fewer evenings charting

Most physicians now use some AI, but far fewer practices see a lighter workload. The difference is choosing one measurable task, checking every output and telling patients what you use.

The short answer

In one paragraph

Start with the work that keeps clinicians charting at night: visit notes, the phone line and the patient inbox. An ambient scribe that drafts notes for review, plus automated reminders and call handling, gives the clearest time back, and large programs have measured hours saved. Measure before and after, because in one 2026 poll only 19% of practice leaders said AI had reduced workload. Caution: tell patients before you record a visit, sign a business associate agreement with every vendor, and have a clinician read each AI note before signing. AI drafts; clinicians decide and keep the conversation.

By Ian Luca Calcavechia, AI consultant, Kraków · Last updated 2026-11-13

AI in medical practices and outpatient clinics: where things stand in 2026

  • 81% of US physicians used AI in their practice in 2026, up from 38% in 2023, with an average of 2.3 uses each. (American Medical Association, March 2026)
  • 88% of physicians were concerned about potential skill loss from AI, and 86% said data privacy is critical for wider AI adoption. (American Medical Association, March 2026)
  • 83% of US medical groups used AI in some patient visits in August 2026, up from 71% a year earlier. (MGMA Stat, August 2026)
  • Asked about workload, 19% of practice leaders said AI had reduced it, 27% said work had shifted to other tasks and 30% saw no real change. (MGMA Stat, August 2026)
  • 57% of physicians named cutting administrative burden through automation as AI's biggest opportunity. Documentation of billing codes, charts and visit notes was the top use, at 21%. (American Medical Association, February 2025)

Six uses that pay off first

01 · Reminders and no-shows

Automated texts confirm visits, offer easy rescheduling and fill freed slots from a waitlist, so fewer appointments go empty.

02 · Phones and booking

An AI agent answers routine calls about hours, booking and refills, and hands clinical or upset callers to staff with a short summary.

03 · Visit notes

An ambient scribe listens, with the patient's permission, and drafts the note for the clinician to edit and sign.

04 · Patient inbox

AI sorts portal messages by urgency and drafts replies to routine questions. A clinician reviews anything clinical before it goes out.

05 · Prior authorizations and eligibility

AI pulls the needed details from the chart into payer forms and checks coverage before the visit, leaving staff to fix the exceptions.

06 · Coding and claim checks

AI suggests codes from the signed note and flags likely denials before claims go out, with a biller making the final call.

How leading companies use it

The Permanente Medical Group

Gave ambient AI scribes to 7,260 physicians, covering 2,576,627 patient encounters from October 2023 to December 2024. An estimated 15,791 hours of documentation saved over 63 weeks. 56% of patients reported a positive effect on visit quality and none reported a negative one. (American Medical Association, June 2025)

Canada Health Infoway

Ran a national AI scribe program for primary care that enrolled 12,019 clinicians between June 2025 and February 2026. Clinicians reported saving about 1.11 hours a week during work hours and nearly 1 hour after hours, and same-day note completion rose from 55.9% to nearly 75%. Most figures are self-reported. (Canada Health Infoway, June 2026)

AssociatesMD

A 54-provider medical group with nine locations in South Florida that moved from manual documentation to the Sunoh.ai scribe across all sites. Fewer clicks and more reliable order capture, according to the vendor's press release; no figures were published. (Sunoh.ai (vendor press release), December 2025)

What’s working now: the boldest AI already in use

The boldest AI that is actually running, not concept videos: named organisations, dated, with the source. Figures marked company- or vendor-reported have not been independently audited.

Live · April 2025 (peer-reviewed)

The Permanente Medical Group (Kaiser Permanente) · USA

An ambient AI scribe turns the visit conversation into a transcript and suggested visit summary; patients are told it is in use and the physician reviews and edits the note. Result: 7,260 physicians used it in about 2.58 million encounters (October 2023 to December 2024), saving nearly 16,000 hours of documentation, published in NEJM Catalyst. (Kaiser Permanente Division of Research)

Pilot · October 2025 to October 2026

Doctronic, under Utah's AI regulatory sandbox · USA

An AI agent handles 30-, 60- and 90-day renewals of medications a licensed provider already prescribed; it issues no new prescriptions and no controlled substances, and in this phase a licensed practitioner authorises every request. Result: The state's May 2026 public report notes that a third-party red-teaming report found possible weaknesses under adversarial use; Utah's Medical Licensing Board wrote to the state about the pilot in April 2026. (Utah Department of Commerce)

Copy it small: An off-the-shelf ambient scribe with a business associate agreement, measuring after-hours charting time before and after.

Main risk: Smaller real savings than claimed: a February 2026 review found savings of under a minute per note in two studies, against vendor claims of about five minutes per encounter. (source)

AI playbook by budget: 15 moves for medical practices and outpatient clinics

Fifteen moves for medical practices and outpatient clinics, from what any business can switch on this week to the bets that need serious money and development. Costs and timelines are typical ranges (estimates), not quotes. Named results link to their source; company and vendor figures are labelled. Every move keeps a person approving anything that goes to a customer, a payment or a legal or clinical decision.

Start this week

Under US$100/month or US$1,000 onceDays to 2 weeks

  1. Ambient AI scribeWith patient consent and a business associate agreement; the clinician edits every note. The Permanente Medical Group had 3,442 physicians using AI scribes in up to 303,266 encounters in the first 10 weeks (NEJM Catalyst, 2024). (source)Tools: Freed, Heidi, Suki, DAX Copilot, Abridge · Measure: after-hours EHR time
  2. Patient-message reply draftsMeasure: inbox time
  3. Prior-authorisation lettersMeasure: time per prior authorisation
  4. AI phones for scheduling and refillsMeasure: abandoned calls
  5. Handouts at the right reading levelMeasure: patient comprehension
Start on Fiverr →

Invest for a few months

About US$5,000–50,0001–6 months

  1. EHR-integrated scribe with coding suggestionsMeasure: charge capture
  2. Revenue-cycle automationTools: Waystar, Infinitus · Measure: clean-claim rate
  3. Referral and fax intakeTools: Notable, Tennr · Measure: referral-to-appointment time
  4. No-show predictionMeasure: fill rate
  5. Care-gap outreachOnly to consented patients.Measure: quality scores
Build it on Fiverr →

Bet big

US$100,000 and up6–24 months

  1. Asynchronous virtual careWith clinician sign-off.Measure: visits per clinician
  2. Value-based-care risk modelsMeasure: shared savings
  3. Clinical decision supportFDA software-as-a-medical-device rules apply; in the EU, AI in regulated medical devices is high-risk from 2 August 2028.Measure: guideline adherence
  4. Chronic-care voice follow-upOnly to consented patients.Measure: readmissions
  5. De-identified data productsMeasure: licensing revenue
Scope the bet on Fiverr →

What not to do

  • Recording visits without telling patients. A proposed class action reported in December 2025 alleges Sharp HealthCare recorded visits with an AI scribe without telling patients, estimates more than 100,000 patients may have been recorded, and says notes falsely stated patients had consented. The claims are unproven, but consent has to be real and documented. (source)
  • Signing AI notes without reading them. Texas has required practitioners who use AI for diagnosis to review every AI-generated record since September 2025. Reading before signing is good practice everywhere. (source)
  • Trusting vendor accuracy numbers. In 2024 the Texas attorney general settled with a clinical AI vendor that had advertised hallucination rates below 0.001%, alleging the claims broke state consumer protection law. Ask for independent evidence and test on your own notes. (source)
  • Overlooking bias in decision-support tools. Since May 2025, US practices covered by Section 1557, which includes those receiving Medicare Part B funds, must make reasonable efforts to find and reduce discrimination risks in patient care decision-support tools, including AI. (source)
  • Buying without a baseline. MGMA found that 27% of practice leaders say AI just shifted work to other tasks. Without before-and-after numbers you cannot tell which group you are in. (source)

Rules to know

  • HIPAA (US): a cloud service provider that creates, receives, maintains or transmits electronic patient data for a practice is a business associate, even if it only stores encrypted data without the key, and the practice must have a business associate agreement with it. (US Department of Health and Human Services, December 2022)
  • California AB 3030: since 1 January 2025, clinics and physician practices that send patients generative AI messages about clinical information must say the message was AI-generated and explain how to reach a human, unless a licensed clinician reviewed it first. (ArentFox Schiff, October 2024)
  • EU AI Act: AI medical devices that need a notified body's assessment under the MDR or IVDR count as high-risk AI systems and must meet both sets of rules. (OpenRegulatory (MDCG 2025-6 explained), August 2026)

Practical guidance from an AI consultant, not legal advice; check each build with your adviser or compliance team.

How to start in 30 days

  1. Week 1: Pick one problem and measure it: after-hours charting time, missed calls, no-show rate or inbox backlog.
  2. Week 2: Shortlist two vendors. Get a signed BAA, ask how long audio and data are kept and whether they are used for training, and check the link to your EHR.
  3. Week 3: Write a short patient notice and consent script, then pilot with two or three willing clinicians who review every output.
  4. Week 4: Compare with week 1 and ask clinicians, staff and a few patients what changed.
  5. Week 5: Expand or stop, and write a one-page AI policy covering approved tools, who reviews outputs and how to report errors.

Tools in use in this sector

Abridge · Sunoh.ai · eClinicalWorks · OpenEvidence · n8n · Make · Zapier. The right one is usually the one the team already uses; automations in n8n, Make or Zapier connect them.

Work with an AI consultant for medical practices and outpatient clinics

Implement it

Set up the first moves for medical practices and outpatient clinics in days, or build the connected system (n8n, Make or Zapier agents, CRM, integrations) over a few months.

AI automation on Fiverr →
Plan it

An AI roadmap for your business: which moves pay off first, tools, costs, risks, and an EU AI Act check.

AI roadmap on Fiverr →
Review a specific case

Send the process, the vendor proposal or the AI tool you are considering; get a written second opinion on fit, cost, risks and what to ask.

Quick scan on Fiverr →
Bet big

For a Tier 3 bet: a PRD, roadmap and backlog, and someone running the build with your developers and vendors.

Product owner on Fiverr →

Ian Luca Calcavechia is an independent AI consultant in Kraków working in English, Spanish, French and Russian, AI product owner and delivery lead for a US healthcare company since 2022, freelancing since 2019 with 100+ Fiverr orders. All services and prices · All 30 industries

Who wrote this

Ian Luca Calcavechia · AI consultant, Kraków

Independent AI consultant who builds AI roadmaps, AI automations and agents (n8n, Make, Zapier) and custom CRMs for small and mid-sized companies in the EU, the US and Latin America. Since 2022 AI product owner and delivery lead for a US healthcare company, where he rebuilt a multi-party operation as a custom CRM with five portals and an AI layer that turns calls, texts and emails into case notes. Services and prices · All industries

Tools named in this guide

Get it done

Start with a fixed package

AI roadmap

Where AI pays off first in medical practices and outpatient clinics, tools, costs and risks

Start on Fiverr →

AI automation

One workflow for medical practices and outpatient clinics, with an AI step and a walkthrough video

Start on Fiverr →

AI project manager

Run the AI rollout with your team and vendors

Start on Fiverr →

Questions

Frequently asked

Does an AI scribe pay off for a small practice?

Large programs report real time savings: The Permanente Medical Group estimated 15,791 hours saved over 63 weeks, and Canadian primary care clinicians reported about two hours a week. Results vary by clinician, so run a one-month pilot and measure note time before and after.

Do patients need to agree before a visit is recorded?

Tell patients and give them a way to say no. California requires all parties to consent before confidential conversations are recorded, and Texas has required providers to disclose AI use in treatment since January 2026. Check your own state's rules with counsel.

Will AI replace my front desk or medical assistants?

Not in most practices. MGMA's 2026 poll found AI more often shifted work than removed it. Use it for repetitive calls, reminders and paperwork so staff have more time for the patients who need a person.

Which AI tools do medical practices use?

Ambient scribes such as Abridge and Sunoh.ai draft visit notes, and OpenEvidence answers clinical questions from medical literature. Automation tools such as n8n, Make or Zapier can connect reminders, forms and practice systems, as long as patient data stays covered by a BAA.

Where should a medical practice start with AI?

Start with one measurable admin task, such as no-shows, phone overflow or visit notes. Get the contract and patient notice right first, pilot for a month and keep only what saves time you can count.

Who can implement AI for medical practices and outpatient clinics?

Ian Luca Calcavechia, an independent AI consultant, implements AI for medical practices and outpatient clinics: quick setups that take days, connected systems with AI agents, CRMs and integrations built over a few months, and product ownership for bigger bets. Fixed packages start on Fiverr (fiverr.com/ianlucac).

Can someone review an AI vendor's proposal or a specific case for medical practices and outpatient clinics?

Yes. Ian Luca Calcavechia reviews a specific process, vendor proposal or AI tool and gives a written second opinion on fit, cost, risks and the questions to ask, starting with the AI quick scan package on Fiverr.

What is the cheapest AI to start with in medical practices and outpatient clinics?

The moves in the Start this week column: AI answering and booking for calls, chats and emails, and AI drafts of documents the business already writes, with a person approving what goes out. They usually cost under US$100 a month (estimate).