Loading
Comparison · dataset August 2026

Nurse Practitioner vs Physician Assistant: which is more exposed to AI?

Effectively tied: 1 point separates Nurse Practitioners from Physician Assistants.

Nurse Practitioner scores 31% time-weighted AI exposure and Physician Assistant scores 32% — close enough that the headline number tells you almost nothing. The difference lives underneath it: Nurse Practitioners lose the most ground on "draft visit notes and documentation" (82%), while for Physician Assistants it is "draft clinical documentation" (82%). Both roles sit inside Healthcare, so the exposure difference reflects task design rather than a change of field.

1PP GAP

Seven dimensions, side by side.

METRICNURSE PRACTITIONERPHYSICIAN ASSISTANTDELTA
AI exposure31%32%1pp gap
Resilience score80/10078/1002pt gap
Substitutable work time15%15%Fully automatable today
Human-critical work time57%58%Models score poorly here
Median salary$126k$130k$4k apart
10-year growth45%28%Nurse Practitioner
US workforce280k148kBLS OEWS
Task level

What actually creates the gap.

Both roles hand a similar slice of the week to substitutable work — 15% for Nurse Practitioners, 15% for Physician Assistants — but it is different work. Nurse Practitioner exposure concentrates in "draft visit notes and documentation"; Physician Assistant exposure concentrates in "draft clinical documentation". Two roles can share a score and face completely different disruption timelines.

Nurse Practitioner
MOST EXPOSED TASKS
  • Draft visit notes and documentation82% · 8% time
  • Generate patient instructions78% · 4% time
  • Code visits for billing74% · 3% time
HUMAN-CRITICAL CORE
  • Counsel patients and build trust8% · 15% time
  • Perform physical examinations10% · 20% time
  • Coordinate care with physicians16% · 8% time
Physician Assistant
MOST EXPOSED TASKS
  • Draft clinical documentation82% · 8% time
  • Generate discharge instructions78% · 4% time
  • Prepare billing and coding74% · 3% time
HUMAN-CRITICAL CORE
  • Communicate hard news to patients6% · 10% time
  • Perform clinical procedures8% · 14% time
  • Examine and assess patients10% · 22% time
What transfers

Both roles lean on judgement, social, cognitive — that is the part of your experience that travels intact. Beyond that, the two capability profiles are unusually close: no dimension separates them by more than 15 points, which is why the switch difficulty below reads the way it does.

Switching between them
LowDIFFICULTY

Physician Assistant appears in our dataset as a mapped adjacent career for Nurse Practitioners: the move raises exposure by 1 points, landing at 32%. Switch difficulty reads low — capability profiles are 4 points apart on average and both sit in the same family.

Score your own exposure in 8 questions →

Common questions.

Is Nurse Practitioner or Physician Assistant more at risk from AI?

Physician Assistant. It scores 32% time-weighted AI exposure against 31% for Nurse Practitioner — a 1-point gap. 15% of physician assistant work time is already fully substitutable by current models, versus 15% for Nurse Practitioners.

Which pays more, Nurse Practitioner or Physician Assistant?

Physician Assistant, by roughly $4k at the median ($130k versus $126k). Note that the higher-paying role here is also the more AI-exposed one, which matters if you are weighing pay against durability.

Can a nurse practitioner switch to being a physician assistant?

Physician Assistant appears in our dataset as a mapped adjacent career for Nurse Practitioners: the move raises exposure by 1 points, landing at 32%. Switch difficulty reads low — capability profiles are 4 points apart on average and both sit in the same family.

Which role is growing faster, Nurse Practitioner or Physician Assistant?

Nurse Practitioner, at 45% projected ten-year growth versus 28% — a 17-point difference. Growth and AI exposure are separate signals: a role can grow in headcount while the content of the work is substantially rewritten.

Career families
Healthcare
Methodology
Scores are time-weighted across each role's canonical O*NET tasks, graded against current frontier-model capability. How we score.