EXPOSED
A large slice of the day — literature review, grant boilerplate, manuscript drafting, statistical analysis, protocol writing, sequence and image data processing — is exactly what current models do at usable quality, and AI-driven hypothesis generation is compressing the design tier too. What holds is the wet-bench and vivarium reality: running assays, handling cell lines and animal models, troubleshooting a failed experiment at the bench, and owning the call on whether a result is real. Licensure is not required for research itself (only clinical-trial PI roles lean on an MD), so the regulatory shield is thin — IRB and FDA accountability, not personal licensure.
Dipped in 2020, then grew past where it started.
Median pay $88,790 → $103,410 -6.8% 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
+8.7% 165,300 → 179,600 on the projections basis
Growing, and only partly exposed
The BLS expects +8.7% more of these jobs by 2034, and at 49/100 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.
~9,600 openings a year on average, including replacing people who leave.
AnatomistScientistCytologistResearcherSerologistHistologistToxicologistGerontologistImmunochemistChemotherapistNeuroscientistParasitologistPharmacologistStudy DirectorEndocrinologistPharmacognosistClinical AnalystHistopathologistCancer ResearcherMedical PhysicistMedical ScientistNeurophysiologistClinical ScientistMedical Researcher
Holding it up: judgment & accountability . Weakest point: liability shield .
Mixed — a routine tier and a judgment tier Roughly half your week — pulling PubMed, drafting R01 aims and specific-aims boilerplate, writing methods sections, running DESeq2/flow gating pipelines, batch-analyzing imaging data — is already reproducible by current models, while pipetting a finicky co-IP, salvaging a contaminated primary culture, and reading an unexpected Western are not, which is why this sits at 10 rather than in the resistant band.
Some physical or field component A 12 reflects that the bench is physical but climate-controlled: BSL-2 hoods, cryostorage, mouse colony handling, IACUC surgeries and perfusions, and hours at a confocal — real hands, real animals, but a fixed room with SOPs, not a field site or a patient's home.
Certification preferred, not legally required No licence gates the work — a PhD and institutional appointment are what let you run a lab — so the 6 comes entirely from the certification layer around you: IACUC/IRB approvals, BSL and radiation-safety training, CITI, and GLP/GCP sign-offs that attach to protocols and institutions rather than to your personal credential.
Exists to be accountable for ambiguous calls A 14 is earned by the calls nobody can procedure for you: whether an n=3 effect is real or a batch artifact, when to kill a two-year project line, which control the reviewer will demand, whether to report a finding that contradicts your own prior paper, and signing off that animal numbers and endpoints in the IACUC protocol are scientifically justified.
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 (10/20 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 (6/20) is whether the law requires a licensed human to sign. Trust premium (7/20) 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 27 of this occupation's 49 points (55%).
Embodiment (12/20) 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.
No occupation passed every test: close enough to medical scientists, except epidemiologists on skills and subject matter, at least 10 points more resistant, no big jump in training, no new licence, no pay cut, and not shrinking on its own. That happens for 223 of the 654 occupations here that aren't SAFE, and it is worth stating plainly rather than leaving the section off.
The usual reason is that exposure travels with the skill profile. The jobs most similar to yours tend to be exposed for the same reasons yours is, so the near neighbours don't clear the gap — and the ones that do are a different kind of work, not a transfer of what you already know. Read that as a limit of this method, not a verdict that you're stuck: it only compares whole occupations, and it cannot see specialisation, industry, or anything you'd bring that isn't in a federal skill survey.
Here is that claim on your own job rather than in the abstract. These are the three occupations closest to this one by skill and subject matter — the places the work would most naturally transfer — with what the register scores them:
That is the whole problem in three lines. The nearest work is not meaningfully safer, so there is no move here that trades a similar skill set for a better verdict. This is not us running out of ideas — it is what the neighbourhood looks like.
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 65/100, still EXPOSED.
Task-mix shift is genuine here: the occupation has a routine tier (lit review, boilerplate, standard stats, image segmentation) and a judgment tier (assay design under confounds, deciding a surprising result is real vs artifact, troubleshooting a failed prep). If the routine tier is fully automated and headcount contracts to the bench-plus-adjudication tier, measured resistance of the remaining role rises. Watch for reproducibility crises traced to unvalidated AI analyses, which push funders to require human wet-lab confirmation of computational findings.
FDA/OHRP or journal-side rules that require a named human investigator to attest personally to data provenance and AI involvement — e.g. an expansion of FDA 21 CFR Part 11 / ALCOA+ data-integrity attestation to cover AI-generated analyses, or an ICMJE/NIH rule making a specific author personally accountable for every AI-produced figure and statistic. Data Integrity Warning Letters already name individuals; formalizing a signature block for AI-assisted analysis would attach personal exposure.
State clinical-laboratory rules (CLIA high-complexity testing, NY State CLEP) already require a doctoral-level qualified person to sign out validated assays; extension of that signatory model to research-grade AI/ML-derived biomarker and sequencing pipelines used in translational work would pull a slice of 19-1042 under a named-signer regime.
If institutions respond to AI-generated research misconduct by naming a single accountable scientist per study (analogous to a study-record custodian), and if research-integrity offices begin sanctioning individuals for unverified AI outputs, the role's ownership of consequential calls under ambiguity becomes formally recognized rather than diffuse.
Rises only in relative terms — if automated liquid handlers and cloud labs stay confined to standardized protocols while novel-model, primary-tissue, and vivarium work resists standardization, the surviving job is disproportionately hands-on. AAALAC and IACUC requirements for trained human handling and daily animal-welfare judgment keep vivarium work human-attached.
The limit. Trust premium has no plausible route: grant reviewers and pharma buyers pay for validated results, not for human authorship, and no purchaser segment specifically demands a human-generated hypothesis. Liability gains are capped because research licensure does not exist and no professional body is credibly moving toward it; the shield can only grow through data-integrity attestation and CLIA-adjacent signoff, not personal licensure.
| Boston-Cambridge-Newton, MA-NH | 16,060 | $129,500 +25% |
| New York-Newark-Jersey City, NY-NJ | 10,090 | $103,360 +0% |
| Los Angeles-Long Beach-Anaheim, CA | 8,520 | $134,890 +30% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 8,020 | $102,950 +0% |
| San Francisco-Oakland-Fremont, CA | 7,310 | $160,180 +55% |
| Houston-Pasadena-The Woodlands, TX | 6,090 | $80,470 -22% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 5,980 | $107,110 +4% |
| Seattle-Tacoma-Bellevue, WA | 5,480 | $105,430 +2% |
| Vallejo, CA | 130 | $168,860 +63% |
| Topeka, KS | 80 | $162,810 +57% |
| Charleston-North Charleston, SC | 200 | $162,230 +57% |
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 49. 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.