EXPOSED
Pathology is the medical specialty most exposed to current AI: whole-slide image interpretation, Gleason and Nottingham grading, mitotic counting, IHC quantification, and lymph node metastasis detection are exactly the pattern-recognition tasks deep learning already does at or near expert level, and reports are structured text. What holds is the physical and legal core — gross dissection of resection specimens, intraoperative frozen sections with the surgeon waiting, autopsies, FNAs, and the fact that a licensed MD must sign every diagnostic report and serve as CLIA laboratory director. Expect fewer pathologists reading more cases with AI pre-screening, not an empty specialty.
Roughly flat across the period, with year-to-year wobble.
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.
BLS projection, 2024–2034
+4.2% 12,600 → 13,100 on the projections basis
Growing, and only partly exposed
The BLS expects +4.2% more of these jobs by 2034, and at 63/100 the work is only partly exposed — some tasks are automatable, the core of the job is not. Nothing here is in tension.
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.
~400 openings a year on average, including replacing people who leave.
PhysicianCytologistPathologistCytopathologistHistopathologistNeuropathologistOral PathologistHematopathologistImmunopathologistAnimal PathologistDermatopathologistOcular PathologistAutopsy PathologistMedical PathologistPathology PhysicianPoultry PathologistAnatomic PathologistChemical PathologistClinical PathologistForensic PathologistSurgical PathologistMolecular PathologistPathologist PhysicianPediatric Pathologist
Holding it up: liability shield . Weakest point: task resistance .
Mixed — a routine tier and a judgment tier An 8 reflects that the bulk of a surgical pathologist's day — screening H&E slides, grading prostate and breast carcinoma, counting mitoses, scoring ER/PR/HER2 and Ki-67, hunting micrometastases in sentinel nodes — is now matched by commercial whole-slide algorithms, and synoptic CAP cancer templates make the report itself a fillable structure; it isn't lower because gross dissection, margin orientation, frozen-section triage under time pressure, and correlating an odd immunoprofile with clinical history and molecular results still need you at the scope.
Some physical or field component An 11 comes from the hours you actually spend with your hands on tissue and needles — grossing a Whipple or colectomy, inking and sectioning margins, performing FNAs and bone marrow aspirates, cutting frozen sections at the cryostat, doing autopsy evisceration — but it stays out of the 13+ band because all of it happens in a fixed, ventilated grossing room or morgue you control, not in unpredictable field conditions.
Licensed human required and personally liable A 19 is warranted because nothing leaves the lab without an MD signature on the diagnostic report, board certification in anatomic and/or clinical pathology gates the job, and under CLIA '88 the laboratory director is a named individual personally answerable to CMS for the entire lab's proficiency testing, validation, and QA — plus you are the defendant of record when a missed melanoma or mis-graded biopsy becomes a malpractice claim.
Exists to be accountable for ambiguous calls A 16 recognises that you make irreversible, ambiguous calls with no procedure to hide behind — benign versus malignant on a scant atypical core, whether a melanocytic lesion crosses into melanoma, calling a margin positive when re-excision means the surgeon reopens the patient, ruling on cause of death — and you must decide when to defer, order more IHC, or overrule an algorithm's output while the clinical clock runs.
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 (8/20 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 (19/20) is whether the law requires a licensed human to sign. Trust premium (9/20) is whether buyers specifically pay for a person. Judgment and accountability (16/20) 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 44 of this occupation's 63 points (70%).
Embodiment (11/20) 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.
Anesthesiologists SAFE
Dermatologists SAFE
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 73/100 — SAFE.
CMS/CLIA rulemaking or a CAP checklist requirement that explicitly names AI-assisted whole-slide diagnosis as requiring a named board-certified pathologist's attestation per case (not per-batch validation), plus a CLIA lab-director sign-off on each algorithm version deployed — CAP's AI committee and the FDA's Paige Prostate de novo authorization already condition use on pathologist confirmation
Two genuine tiers exist. If routine screening-tier reads are automated, the residual job is grossing complex resections, frozen-section margin calls with the surgeon in the room, cytology adequacy judgment at the FNA needle, and integrating molecular/genomic results into a single diagnostic narrative — none of which current systems produce end-to-end. Watch for whether AI vendors ship gross-room and frozen-section workflow products or stay in the WSI viewer
Task-mix concentration: as AI pre-screens negatives and grades routine prostate/breast cases, the signed-out workload shifts toward discordance adjudication, rare-tumor and hematopathology sign-out, molecular tumor board correlation, and being the accountable human when the algorithm and the clinical picture conflict. Also raised if malpractice carriers begin writing policies that name the pathologist as the responsible reviewer of AI-flagged cases
Rises only if health systems reverse specimen-handling centralization — e.g. state licensure or CAP rules restricting digital-only remote sign-out across state lines, forcing on-site pathologists for grossing and intraoperative consultation at each hospital. Interstate telepathology licensure fights are real but currently trend the other way
The limit. Trust premium has no plausible route: patients almost never know their pathologist's name, do not choose one, and cannot pay for a human reader — the buyer is the hospital or lab, optimizing cost per slide. Any protection here is legal and institutional, not consumer preference. Even with the liability shield near maximum, a shield determines who signs, not how many signers are needed; headcount can fall sharply while every report remains physician-signed.
| New York-Newark-Jersey City, NY-NJ | 1,140 | $327,770 +5% |
| Dallas-Fort Worth-Arlington, TX | 480 | $316,150 +1% |
| Los Angeles-Long Beach-Anaheim, CA | 330 | $319,340 +2% |
| Minneapolis-St. Paul-Bloomington, MN-WI | 280 | $239,200 -23% |
| Indianapolis-Carmel-Greenwood, IN | 260 | $342,090 +10% |
| Boston-Cambridge-Newton, MA-NH | 240 | $285,240 -9% |
| Houston-Pasadena-The Woodlands, TX | 230 | $291,510 -7% |
| Phoenix-Mesa-Chandler, AZ | 220 | $349,470 +12% |
| Las Vegas-Henderson-North Las Vegas, NV | 30 | $384,550 +23% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 90 | $365,830 +17% |
| Tampa-St. Petersburg-Clearwater, FL | 100 | $362,470 +16% |
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 63. 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.