SAFE
The modal worker here is clinical faculty in pharmacy, dentistry, PT/OT, medical, and allied health programs — lecture content, quiz banks, and case write-ups are exactly what generative AI now produces cheaply, but the job's center is supervising students on live patients, running skills labs and simulations, and signing off that a trainee is competent to practice. Accreditors (LCME, ACPE, CAPTE, CODA) and state boards require credentialed, often licensed, human faculty of record for clinical instruction and competency attestation, which is a regulatory moat that could narrow but not vanish. Didactic-only and adjunct lecturers are the exposed tier; clinical coordinators and preceptor-supervising faculty are the durable one.
Dipped in 2020, then grew past where it started.
Median pay $97,320 → $107,310 -11.8% in real terms
This line is counted by the Bureau of Labor Statistics — the one figure on this page that isn't a judgement of ours. Headcount moves on demand, offshoring, demographics and the business cycle, and automation is one term among several, often not the loudest.
So a falling line is not evidence that AI did it, and a rising one is not evidence that it won't. Both happen in this register: some occupations resist automation and shrink anyway, others are highly automatable and keep growing. The marked year is 2020.
BLS projection, 2024–2034
+17.3%
Percentage only. The projection counts a different population from the 221,270 above — it includes self-employed workers, which for this occupation is most of them, so the two headcounts are not comparable.
Hard to automate, and growing
The work resists current AI and the BLS projects +17.3% more of these jobs by 2034. Note that safe does not mean well paid — several of the fastest-growing resistant occupations are among the lowest paid on the register.
Different clocks. The score is what current AI could do to this work today. The projection is how many of these jobs will exist in 2034. Everything between the two — how fast employers actually adopt, whether demand grows in the meantime — is why they can point opposite ways without either being wrong.
~27,400 openings a year on average, including replacing people who leave.
TeacherLecturerProfessorInstructorFaculty MemberHealth TeacherAnatomy TeacherCoding EducatorHygiene TeacherSurgery TeacherTherapy TeacherUrology TeacherMedicine TeacherPharmacy TeacherPodiatry TeacherSerology TeacherVirology TeacherCollege ProfessorDentistry TeacherDietetics TeacherDietitian TeacherFirst Aid TeacherHistology TeacherNeurology Teacher
Holding it up: trust premium . Weakest point: liability shield .
Mixed — a routine tier and a judgment tier Slide decks, item-banked exams, and syllabus updates are already drafted by machine, but watching a second-year student's hands during a scaling procedure or a transfer from bed to wheelchair, correcting technique in real time, and debriefing a high-fidelity sim scenario are the tasks that hold the 13 rather than pushing higher — a meaningful share of contact hours is still lecture and grading that a model can shoulder.
Hands-on in uncontrolled environments Clinical faculty teach in dental operatories, PT gyms, hospital wards, and cadaver labs where they physically position students, demonstrate palpation and injection technique, and stand at the chairside during live patient care — it sits at 13 rather than 17 because the same person also holds office hours, writes curriculum, and sits on accreditation committees from a desk.
Licensed human required and personally liable Most clinical faculty hold an active state license (DDS, PharmD, PT, MD) and are personally on the hook for patient care rendered under their supervision, plus their signature on a competency attestation carries weight with ACPE, CODA, and CAPTE site visitors — the 13 rather than 18 reflects that the teaching credential itself is institutional, and basic-science and didactic-only faculty may hold no license at all.
Exists to be accountable for ambiguous calls Deciding that a student is not safe to continue on a clinical rotation, remediating a near-miss with a real patient, and defending that call through a program's due-process appeal are the ambiguous high-stakes decisions here — 14 rather than 17 because progression standards and accreditation competency frameworks give the call a documented structure to lean on.
The verdict above describes this occupation as a whole. Almost nobody does the typical version of a job — tick what's actually in your week and see how your own mix sits.
Your task mix speaks to task resistance (13/20 here) — how much of the day's work current AI already does. That is the dimension the boxes above are about.
It cannot move the other three. Liability shield (13/20) is whether the law requires a licensed human to sign. Trust premium (16/20) is whether buyers specifically pay for a person. Judgment and accountability (14/20) is whether the role exists to own consequential calls. Those are facts about the occupation's standing, not about which tasks are in your week — a paralegal who does only trial exhibits still holds no licence. Together they are 43 of this occupation's 69 points (62%).
Embodiment (13/20) is also a property of the work rather than the worker, but we don't tag individual tasks as physical or not, so the picker can't tell you anything about it. That's a limit of this tool, not a claim.
Did we get the list right? Tell us what's missing — the tasks are written from the outside, and you're reading this from the inside.
The moves above are yours to make. This is the other half: what would have to change in the world for the occupation itself to score higher. None of it is in any one person's gift, but it is where the floor actually comes from. Scores here are not a one-way ratchet. Only two of the five dimensions — task resistance and embodiment — track what machines can do. The other three track law, what buyers will pay for, and who is answerable, and those move in both directions, often in response to the same pressure AI creates. If every lever below landed, this occupation would score around 82/100, still SAFE.
Task-mix shift: this occupation has a genuine two-tier structure. If lecture authoring, quiz banks, and case write-ups are substantially automated and the surviving role is simulation debriefing, direct patient-side supervision, and competency judgment, the residual day is the judgment tier. Watch for job postings shifting from 'didactic instructor' to 'clinical coordinator/simulation faculty'.
Accreditor rules that explicitly name a licensed, credentialed faculty of record as the sole permissible signer of clinical competency attestations and entrustable professional activity (EPA) sign-offs, with AI-generated assessment barred as sole evidence — e.g., LCME Element 9.x guidance, ACPE Standards revisions, or CAPTE clarifying that AI-assisted evaluation cannot substitute for licensed supervisor judgment. Also state board rules on preceptor-to-student ratios (real, e.g. nursing board caps) extending to allied health.
Formalized remediation and dismissal due-process regimes where the faculty member personally owns the 'unsafe to progress' call and is deposed in student litigation — already happening as courts hear more academic-dismissal suits; if institutions respond by naming an individual attesting faculty rather than a committee, the personal ownership of ambiguous calls rises.
Expansion of hands-on skills-lab and simulation contact-hour minimums in accreditation standards (post-COVID reversals of distance-learning flexibilities are already visible in CODA and CAPTE guidance), plus growth of in-person OSCE proctoring as remote assessment integrity fails.
Limited route: students and residency programs pay for named clinician-educators with active practice, and reputational preceptor networks matter for match placement. But tuition is paid to institutions, not individuals, and adjunct rates show no willingness to pay for human didactic instruction — this dimension is near its realistic ceiling at 16.
The limit. The moat is accreditation, not consumer preference. If accreditors loosen contact-hour and faculty-of-record definitions to permit AI-assisted or asynchronous competency assessment — pressure that already exists from cost-driven online health programs — liability_shield and embodiment fall together and the didactic tier collapses fast. Realistic band is roughly 65-80.
| New York-Newark-Jersey City, NY-NJ | 15,040 | $128,090 +19% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 11,210 | $108,150 +1% |
| Los Angeles-Long Beach-Anaheim, CA | 8,410 | $134,590 +25% |
| Boston-Cambridge-Newton, MA-NH | 7,830 | $164,450 +53% |
| Denver-Aurora-Centennial, CO | 6,710 | $168,720 +57% |
| Chicago-Naperville-Elgin, IL-IN | 6,390 | $101,010 -6% |
| San Francisco-Oakland-Fremont, CA | 4,340 | $357,220 +233% |
| Miami-Fort Lauderdale-West Palm Beach, FL | 4,090 | $128,960 +20% |
| San Francisco-Oakland-Fremont, CA | 4,340 | $357,220 +233% |
| Little Rock-North Little Rock-Conway, AR | 1,560 | $214,790 +100% |
| Greenville, NC | 700 | $181,290 +69% |
We have no reported case of a named organisation automating this occupation. Not one deployment, not one announcement.
That is worth saying out loud next to a score of 69. The verdict above is about what the work exposes — what current AI could do to these tasks. It is not a claim that anyone has done it. For this occupation those two things have come apart completely: the capability argument is on this page, and the evidence column is empty.
Has AI actually changed your work? One tap, anonymous, and the running tally is public. Nothing else is asked of you.
Rather than check back: get the digest and we'll tell you what changed — or watch a single occupation from its own page.