Your field-guide to AI — what it means for your job and what to do about it
Physicians & Doctors
AI scribes and diagnostic tools are transforming medicine, cutting documentation time by hours per day and catching conditions that human eyes miss.
Current AI Tools
Nuance DAX Copilot (Dragon Copilot) is an ambient AI that listens to patient visits and drafts clinical notes automatically. It holds roughly 33% of the ambient scribe market [1]. In March 2025, Microsoft merged it with Dragon Medical One under the “Dragon Copilot” brand. It is live in the U.S., Canada, and U.K. Pricing runs $369 to $830+ per provider per month on enterprise contracts [2].
Abridge is an enterprise-grade ambient AI for clinical documentation, named #1 Best in KLAS for Ambient AI two years running [3]. It is deployed enterprise-wide at Duke Health, Johns Hopkins, Mayo Clinic, and UNC Health, trusted by over 200 major health systems.
Freed is popular with solo and small-group practices, generating notes in about 40 seconds per encounter. It is HIPAA-compliant with SOC 2 Type II certification and costs $90-99 per clinician per month.
Epic AI Charting is Epic’s native ambient scribe. Epic reports 85% of its healthcare customers are live with generative AI features across its Art, Emmie, and Penny copilot tools, with 160+ AI projects underway [4].
For diagnostics, Viz.ai is FDA-cleared for stroke detection, deployed in 1,700+ hospitals with 50+ FDA-cleared algorithms [5]. Studies show it reduces treatment delays by 30-52% [6]. Aidoc provides FDA-cleared radiology triage for CT and MRI scans across 900+ hospitals [7]. PathAI received FDA clearance in June 2025 for its digital pathology platform for primary diagnosis [8].
Essential Skills Today
You should be proficient with at least one ambient AI scribe tool. Understanding how to review, edit, and approve AI-generated clinical notes efficiently is now a daily workflow requirement. Physicians at Mass General Brigham reported saving roughly 4 hours per week on documentation with AI scribes [9].
Baseline AI literacy is expected – understanding what AI can and cannot do diagnostically. Familiarity with AI-assisted clinical decision support within your EHR (Epic, Oracle Health) is important, as these systems increasingly flag potential diagnoses and treatment considerations.
You also need to be comfortable communicating with patients about AI use during visits, including consent for ambient recording. This is both a practical and ethical skill.
12-24 Month Outlook
Expect AI-generated diagnostic suggestions to become routine in clinical workflows. Interpreting these suggestions and integrating them into your clinical reasoning – rather than blindly accepting or ignoring them – is the key skill to develop.
AI-driven predictive analytics for patient risk stratification are maturing rapidly. Systems that identify patients at risk for deterioration, readmission, or chronic disease progression will become standard tools.
Training programs are multiplying: Harvard, Johns Hopkins, Stanford, Mayo Clinic, and the University of Illinois all offer AI in Healthcare certificates or specializations. CMS is expected to accept AI-generated progress notes for billing by mid-2026, which will accelerate adoption [10].
5-Year Outlook
The physician role will not shrink but will transform significantly. Documentation burden – the leading cause of physician burnout – will be largely automated. AI will handle most note-writing, flag diagnostic findings for review, and identify at-risk patients. Physicians will focus on clinical judgment, patient relationships, and complex cases.
Despite 2016 predictions that AI would replace radiologists, radiology residency positions hit a record 1,208 in 2025, a 4% increase year-over-year, with vacancy rates at all-time highs [11]. AI handles repetitive tasks while new subspecialties like radiomics and radiogenomics emerge.
The BLS projects 3% growth from 2024 to 2034 with approximately 23,600 annual openings [12]. There are over 750,000 physicians in the U.S. with a median wage above $239,200 [12]. AI adoption among physicians jumped from 38% in 2023 to 66% in 2024 [13]. There are now 1,451 cumulative FDA-cleared AI medical devices, with 295 new authorizations in 2025 alone [14].
Displacement risk is low. The physician shortage is the workforce concern, not AI replacement.
Action Items
Try an AI scribe this week. If your practice does not already use one, sign up for a trial of Freed ($99/month) or ask your health system about DAX Copilot or Abridge access. The documentation time savings are immediate and significant.
Learn to review AI-generated notes critically. Spend time comparing AI-generated notes to what you would have written manually. Identify where the AI captures information well and where it misses nuance or makes errors. Building this review skill is essential.
Explore AI diagnostic tools in your specialty. Whether it is Viz.ai for stroke, PathAI for pathology, or AI-powered imaging tools in your EHR, understand what AI diagnostic assistance is available in your specific area of practice.
Take an AI in Healthcare course. Stanford, Johns Hopkins, and Harvard all offer online certificates. Even a short course gives you the vocabulary and framework to evaluate AI tools critically rather than adopting them blindly.
Talk to your patients about AI. Start having conversations about how AI is used in your practice. Patient comfort with AI in healthcare varies widely, and building trust through transparency is both good medicine and good practice.
Sources
- Menlo Ventures — 2025: The State of AI in Healthcare — DAX Copilot market share analysis
- Vero Scribe — Nuance DAX Copilot Review 2026 — pricing details for DAX Copilot
- Abridge — Best in KLAS 2026 for Ambient AI — KLAS award recognition
- Fierce Healthcare — Epic’s AI scribe goes live — Epic AI adoption statistics
- Viz.ai — Official Website — hospital deployment and FDA-cleared algorithm count
- Viz.ai — New Studies Demonstrate Impact of Stroke Solution — treatment delay reduction data
- Aidoc — Official Website — hospital deployment count
- PathAI — FDA Clearance for AISight Dx Platform — FDA clearance announcement
- Mass General Brigham — Hybrid Ambient Documentation Decreases After-Hours Work — AI scribe time savings
- CMS — Improving Access to Technology-Supported Care — CMS AI documentation framework
- Radiology Business — Match Day 2025 — radiology residency position data
- Bureau of Labor Statistics — Physicians and Surgeons — employment projections and wage data
- American Medical Association — 2 in 3 Physicians Using Health AI — physician AI adoption survey
- FDA — AI/ML-Enabled Medical Devices — cumulative FDA-cleared AI device count