The FDA has cleared more than 1,000 AI-enabled medical devices since 2016, most of them in radiology and cardiology. For anyone choosing a health care career, the question is no longer whether AI will arrive—it is which jobs AI actually changes today, and which ones it cannot touch.
We score AI impact along two dimensions. Task displacement measures whether working AI solutions exist for a career’s core daily tasks—ranging from none (home health aide) to near-complete automation (medical transcription). Market deployment measures whether those tools are actually being purchased and used at health systems, or remain prototypes. A career can have high displacement potential on paper but near-zero deployment in practice.
The critical distinction: impact does not equal replacement. Cardiologists score High AI Impact because 600+ FDA-cleared cardiac AI tools augment their diagnostic work—but cardiologists use those tools to detect conditions earlier, see more patients, and improve outcomes. Their AI exposure makes them more valuable, not less. Home health aides score Very Low because their work is physical patient care in uncontrolled environments—no company has built an AI product for their core tasks.
Key Findings
- •Physical care resists automation: Careers built around hands-on work in uncontrolled environments—home health aides, dental assistants, phlebotomists—have the lowest AI impact scores. No commercial AI products exist for their core tasks because those tasks require physical presence, manual dexterity, and real-time adaptation to individual patients.
- •EKG and cardiovascular techs score highest on resilience: EKG/ECG technician (AI resilience 70/100) and cardiovascular technician (63/100) lead the resilience rankings because electrode placement, real-time patient monitoring, and equipment calibration involve physical precision that no AI product replicates at scale.
- •AI amplifies more than it replaces: Careers like sonographers, radiologic technologists, and cardiologists score Moderate AI Impact because AI augments their work without replacing it. BLS projects continued growth for most of these roles—professionals use AI to work faster and more accurately.
- •Only one career shows AI-driven job decline: Medical transcription is the only health care field where AI independently performs core tasks at production scale and BLS projects -6.6% employment decline. Every other career—including those with high AI impact—shows stable or growing projections.
AI Impact by Career
Every published health care career classified by evidence-based AI impact level, from Very Low to Very High.
The majority of health care careers—39 of 55—have low or very low AI impact. In these fields, AI either does not exist for core tasks or exists only as reference tools for one or two secondary workflows. The hands-on, relational, and physically demanding nature of most health care work remains beyond the reach of current AI products. EKG/ECG technicians (AI resilience 70/100) and cardiovascular technicians (63/100) exemplify this pattern—their equipment operation and real-time patient monitoring tasks have no commercial AI replacement.
The eight Moderate-tier careers represent the active frontier of AI augmentation. In these roles—sonographers, radiologic technologists, pharmacists—AI handles meaningful analytical work like image interpretation or drug-interaction screening, but professionals still perform the physical, relational, and judgment-intensive parts of the job. BLS projects continued growth for most of these careers, suggesting that AI amplification is making these workers more productive, not redundant.
Only one career, medical transcription, reaches Very High AI Impact—the only field where AI performs core tasks at production scale and BLS projects -6.6% employment decline. Every other health care career, including those with high AI exposure like cardiologists (AI resilience 55/100, $239,000 salary), shows stable or growing employment.
55 careers by AI impact level
Very Low AI Impact(10 careers)
Injury evaluation, taping, and rehabilitation are entirely hands-on; AI limited to wearable data dashboards.
NATA: Technology in Athletic Training · BLS: Athletic Trainers +14% (2023-2033)
Physical patient care (bathing, feeding, mobility) is irreducibly manual; AI only in back-office alerts and scheduling.
ANA: AI in Nursing Practice Position Statement · BLS: Nursing Assistants and Orderlies Outlook
Core work is community trust-building and in-person outreach; AI only automates adjacent logistics like scheduling and reminders.
Nature: AI Voice Agents in Medicine (2025) · NACHC and eClinicalWorks Partnership
No AI in chairside assisting tasks; robotic dentistry advances target the dentist's procedural role, not the assistant's support role.
New Atlas: Robot Dentist World First (2024) · Nature: Future of Dentistry Through Robotics (2025)
Dialysis machine monitoring has some automation but tech performs all cannulation, patient assessment, and emergency response.
BLS: Dialysis Technicians Outlook · Nephrology Nursing Journal: Technology in Dialysis Care
Zero AI in core physical patient care tasks; AI touches only back-office scheduling and documentation at the agency level.
McKnights Home Care: AI in Home Healthcare (2024) · Home Health Care News: Cautious AI Adoption (2025)
Hands-on functional training and adaptive equipment fitting are core tasks; AI only in documentation support.
AOTA: Technology in Occupational Therapy · BLS: Occupational Therapy Assistants Outlook
Hands-on therapeutic exercises and manual techniques are core work; AI limited to documentation and scheduling.
APTA: Technology in Physical Therapy (2024) · BLS: Physical Therapist Assistants Outlook
Ventilator AI modes assist but the RT makes all clinical decisions on settings, weaning, and airway management.
AARC: AI and Respiratory Care Position Statement · BLS: Respiratory Therapists +13% (2023-2033)
Hands in the sterile field passing instruments; robotic surgery assists surgeons, not scrub techs.
AST: Surgical Technology Scope of Practice · BLS: Surgical Technologists +5% (2023-2033)
Low AI Impact(29 careers)
AI monitors vitals and predicts adverse events; the anesthesiologist makes all dosing and airway management decisions.
ASA: AI in Anesthesiology Position Statement · FDA: 14 Cleared AI Devices for Anesthesia Monitoring
AI-driven hearing aid fitting algorithms assist; audiologist performs diagnosis, counseling, and vestibular assessment.
AAA: Technology in Audiology Practice · BLS: Audiologists +10% (2023-2033)
Prenatal, labor, delivery, and postpartum care are hands-on; AI limited to fetal monitoring alerts and documentation.
ACNM: Technology in Midwifery Practice · BLS: Nurse Midwives +6% (2023-2033)
AI monitors vitals and predicts adverse events; CRNA makes all anesthesia dosing, airway, and intraoperative decisions.
AANA: AI in Nurse Anesthesia Position Statement · FDA: AI Devices for Anesthesia Monitoring
AI detects periodontal disease on X-rays; all scaling, root planing, and prophylaxis remain manual.
ADHA: Technology in Dental Hygiene Practice · BLS: Dental Hygienists +7% (2023-2033)
AI assists dermoscopy image classification; dermatologist performs biopsy, diagnosis, and all treatment procedures.
AAD: AI in Dermatology Position Statement · FDA: AI Skin Lesion Analysis Devices
Field emergency care is entirely hands-on; AI only in dispatch triage and protocol reference tools.
NAEMSP: Technology in EMS Position Statement · BLS: EMTs and Paramedics +5% (2023-2033)
AI triage tools and imaging AI assist in ED; emergency physician makes all critical decisions and performs procedures.
ACEP: AI in Emergency Medicine Position Statement · BLS: Physicians and Surgeons +3% (2023-2033)
CDSS and AI scribes deployed in primary care; physician makes all clinical decisions, examinations, and treatment plans.
AAFP: AI in Family Medicine Position Paper · AI-based CDSS in Primary Care: Real-World Study (2025)
Robotic platforms (da Vinci) assist minimally invasive procedures; surgeon performs all decision-making, cutting, and tissue handling.
ACS: Robotic Surgery and AI Position Statement · BLS: Surgeons Outlook +3% (2023-2033)
AI variant classification tools are academic pilots, not production deployments; core counseling relationship remains irreducibly human.
PMC: Evaluating AI for Genetic Counseling in Rare Diseases · Journal of Genetic Counseling: Generative AI and the Profession · Trends in Genetics: AI as Partner for Genetic Counseling
AI virtual staining touches only post-analytical work; core tissue preparation (embedding, sectioning, staining) remains entirely manual and tactile.
PathAI: Virtual Staining Blog · PMC: Cutting-Edge Technology in Pathology Labs
CDSS and AI scribes deployed in internal medicine; physician makes all clinical decisions, examinations, and treatment plans.
ACP: AI in Internal Medicine Practice · AI-based CDSS in Primary Care: Real-World Study (2025)
Core bedside care (wound care, injections, vitals) is hands-on; AI only in documentation and scheduling tools.
ANA: AI in Nursing Practice Position Statement · BLS: Licensed Practical Nurses +3% (2023-2033)
AI ambient scribes serve MDs, not MAs; core MA tasks (vitals, injections, rooming patients) have zero AI automation.
Microsoft: Dragon Copilot (March 2025) · Chief Healthcare Executive: AI in Healthcare 2025
Automated analyzers handle routine tests; MLT troubleshoots instruments, runs QA, and handles complex specimens.
ASCP: AI in Clinical Laboratory Science · BLS: Clinical Laboratory Technologists +5% (2023-2033)
AI assists image reconstruction and quality control; tech positions patients, handles radiopharmaceuticals, and operates equipment.
SNMMI: AI in Nuclear Medicine Position Paper · BLS: Nuclear Medicine Technologists +1% (2023-2033)
Clinical decision support and ambient scribes provide reference information; the NP independently makes all clinical decisions and manages patient relationships.
OJIN: AI in Nursing Practice (May 2025) · AI-based CDSS in Primary Care: Real-World Study (2025)
AI assists documentation and outcomes tracking; core functional training and adaptive strategies are hands-on.
AOTA: Technology and Occupational Therapy Practice · BLS: Occupational Therapists +12% (2023-2033)
Robotic platforms (Mako) assist joint replacement at ~10% adoption; the surgeon performs all cutting, implantation, and decision-making.
AAOS: Robotic-Assisted Surgery Position Statement · Stryker Mako: ~10% Joint Replacement Adoption
CAD/CAM assists design but fitting, alignment, and patient interaction are hands-on; 3D printing is a fabrication tool, not AI.
AOPA: Technology in Orthotics and Prosthetics · BLS: Orthotists and Prosthetists +10% (2023-2033)
Advanced field emergency care (intubation, IV, cardiac drugs) is entirely hands-on; AI assists dispatch and protocol reference.
NAEMSP: Technology in EMS Position Statement · BLS: EMTs and Paramedics +5% (2023-2033)
CDSS and growth chart AI assist screening; pediatrician performs all examinations, diagnoses, and family counseling.
AAP: AI in Pediatric Practice · BLS: Physicians and Surgeons +3% (2023-2033)
Vein-finders are passive reference tools; Vitestro's robotic phlebotomy device is still in EU trials — years from routine US adoption.
Phlebotomy.com: Vitestro Aletta Update · The Pathologist: Robot Blood Draws (2025) · Clinical Trials Arena: Vitestro Trial Endpoints
Motion-capture AI assists movement assessment; all manual therapy and exercise prescription remain human.
APTA: AI and Physical Therapy Practice · BLS: Physical Therapists +14% (2023-2033)
Same AI tools as NPs (CDSS, ambient scribes); PA makes all clinical decisions and manages patient relationships.
AAPA: AI in PA Practice Survey (2024) · AI-based CDSS in Primary Care: Real-World Study (2025)
AI chatbots handle low-acuity screening only; psychiatric diagnosis, medication management, and therapy require human clinician.
APA: AI in Psychiatry Practice Guidelines · Nature Mental Health: AI Chatbots vs Therapists (2024)
AI meal planners exist but medical nutrition therapy counseling requires human relationship and clinical judgment.
Academy of Nutrition and Dietetics: AI Position Paper · BLS: Dietitians and Nutritionists +7% (2023-2033)
AI apps provide speech practice exercises; the SLP does all assessment, diagnosis, and hands-on therapy.
ASHA: AI in Speech-Language Pathology · BLS: Speech-Language Pathologists +4% (2023-2033)
Moderate AI Impact(8 careers)
AI auto-codes routine medical records; tech handles complex cases, auditing, and compliance reviews.
AHIMA: AI in Health Information Management · BLS: Medical Records Specialists +7% (2023-2033)
AI handles a growing share of routine scheduling, insurance verification, and patient communications; role evolving from task executor to AI supervisor and exception handler.
HealthTech Magazine: AI in Healthcare Administration (2026) · Sprypt: AI Scheduling in Healthcare 2025 · GlobeNewsWire: AI Medical Scheduling Market (2026)
AI auto-generates post-call notes, creates tasks, and drafts care plans; coordinator reviews and personalizes everything, maintaining the patient relationship.
Aidoc: AI Care Coordination · ThoroughCare: AI Co-pilot · HealthEdge: GuidingCare with AI
Robotic dispensing at ~10% adoption handles high-volume fills; pharmacist performs clinical review, counseling, and medication therapy management.
ASHP: AI in Pharmacy Practice Survey (2024) · BLS: Pharmacists +3% (2023-2033)
AI auto-contours tumors and optimizes treatment plans; therapist positions patients, delivers treatment, and monitors reactions.
ASTRO: AI in Radiation Oncology White Paper · BLS: Radiation Therapists +1% (2023-2033)
AI augments quality checks, dose optimization, and auto-positioning; tech still physically positions patients and operates equipment (6% BLS growth projected).
Philips: SmartSpeed Precise MRI AI (2025) · IntuitionLabs: AI in Radiology 2025 · JAMA Network Open: FDA AI Devices in Radiology
AI auto-scores a majority of sleep stages with high accuracy; tech monitors patients overnight, troubleshoots equipment, and handles emergencies.
AASM: AI in Sleep Medicine Position Statement · BLS: Sleep Technologists Outlook
AI guides probe placement and auto-measures but does not displace scanning; sonographer performs all scans with real-time clinical judgment (13% BLS growth).
Exo: Handheld Ultrasound with AI · Diagnostic Imaging: Sonio FDA Clearance · Mount Sinai: BrightHeart AI
High AI Impact(7 careers)
AI significantly augments diagnostics with 600+ FDA-cleared cardiac AI tools; cardiologist makes all treatment decisions and performs procedures.
PMC: AI for Cardiovascular Care Part 2 · AHA: AI in Heart Disease Outcomes · Mount Sinai: AI Algorithm for Heart Patients
AI auto-measures guideline-recommended echo parameters and generates draft reports in minutes, significantly reducing analysis time; tech still performs physical scanning.
PMC: AI in Echocardiography · Nature Reviews Cardiology: AI-Enhanced Echocardiography · ScienceDirect: AI vs Human Echo Measurements
AI performs initial slide screening and flags regions of interest; cytotechnologist role transforming from exhaustive visual screener to AI output verifier.
Clinical Lab Products: Hologic Digital Cytology FDA Clearance · Nature: AI Transforming Pathology (2025) · Stanford: AI Tool for Pathologists (2025)
AI independently interprets ECGs for 35+ conditions with cardiologist-level accuracy; tech role shifting from preliminary interpretation to equipment operation and QA.
Healio: FDA Clears AI-Powered ECG Device · Medical Economics: AccurKardia FDA Clearance (2026) · Cardiovascular Business: Cardiology #2 in FDA AI
AI auto-codes routine encounters but adoption is ~46%, not 90%; BLS projects +7% job growth; human coders handle complex cases, auditing, and denials.
Fathom: 90%+ autonomous coding (vendor claim; actual adoption ~46%) · BLS: Medical Records Specialists +7% projected growth 2023-2033 · AMBCI: 80% automation target by 2030 (aspirational, not current)
Total lab automation independently handles routine analysis 24/7; lab scientist essential for complex specimens, troubleshooting, and quality assurance.
WEF: AI, Robotics and Automation in Diagnostics (2024) · Roche: AI in Laboratory Workflows · PMC: Total Laboratory Automation Trends
Robotic dispensing handles high-volume prescription filling; tech role shifting from counting pills to clinical support and AI oversight.
ASHP: Pharmacy Automation Survey (2024) · BLS: Pharmacy Technicians -6% (2023-2033)
Very High AI Impact(1 career)
AI ambient scribes bypass dictation entirely; employment declined 46% since 2012; market adoption ~38% across US health systems.
BLS: Medical Transcriptionists Outlook (-5% projected) · Menlo Ventures: State of AI in Healthcare 2025 · PHTI: AI Adoption in Healthcare Delivery Systems
Sources
- •Bureau of Labor Statistics — Occupational Outlook Handbook, 2024–34
- •Bureau of Labor Statistics — OEWS, May 2024
- •FDA: Artificial Intelligence and Machine Learning (AI/ML)-Enabled Medical Devices
- •Menlo Ventures: The State of Generative AI in Healthcare 2025
- •Peterson Health Technology Institute: AI Adoption in Healthcare Delivery Systems
- •American Medical Association: How AI Is Helping Doctors Improve Care
Methodology
Two-dimensional rubric. Each career is scored on two dimensions: task displacement (0–50) and market deployment (0–50). Task displacement measures whether working AI solutions exist for a career’s core daily tasks. Market deployment measures whether AI products are actually being purchased and used at health systems. The total (0–100) determines the AI impact tier.
Task displacement brackets. 0–10: No AI exists for core tasks — it touches only admin or adjacent work. 11–20: AI reference tools exist for one or two tasks. 21–30: AI augments several core tasks but a human reviews everything. 31–40: AI independently performs one or more major task components. 41–50: AI performs most core tasks autonomously and the human role is minimal or declining.
Market deployment brackets. 0–10: No commercial products for core tasks. 11–20: Early-stage pilots at limited sites. 21–30: Named vendors with paying customers; adoption still limited. 31–40: Major health systems deploying; broad adoption underway. 41–50: Industry standard; widespread adoption.
Evidence requirements. Every classification requires at least two cited sources: named AI products or vendors with identified customers, FDA clearance records, peer-reviewed clinical deployment studies, BLS employment projections, or verified health system adoption data (e.g., Kaiser Permanente, Cleveland Clinic). Theoretical capabilities, research-only papers, and unverified vendor claims do not qualify.
Tier boundaries. 0–19: Very Low AI Impact (no AI in core tasks). 20–34: Low AI Impact (AI reference tools for 1–2 tasks). 35–54: Moderate AI Impact (AI augments several core tasks, human reviews). 55–74: High AI Impact (AI independently performs major task components). 75–100: Very High AI Impact (AI performs most core tasks autonomously).
Complete coverage. All published careers have been individually researched and scored with the evidence-based rubric. Each entry includes task displacement and market deployment subscores, a one-sentence summary, and 2–3 cited sources.
Limitations. This methodology measures demonstrated AI deployment, not prediction of future displacement. A career with high AI impact may benefit enormously from adoption (as with cardiologists and diagnostic AI) or face genuine role restructuring (as with medical transcriptionists). Scores reflect current deployment evidence and will evolve as AI products mature and adoption patterns change.