SAFE
The job is physical: screening patients for ferromagnetic implants, positioning bodies and coils to millimeter tolerance in a bore, starting IVs for gadolinium contrast, and talking anxious or claustrophobic patients through a 40-minute scan. AI is already eating the adjacent screen work — protocol selection, reconstruction and denoising, artifact flagging, scan-time reduction — which will raise throughput per tech and compress the number of techs needed per scanner, but it does not touch the bedside half of the shift. Certification (ARRT-MR or ARMRIT) plus state licensure in most states keeps a credentialed human on the hook for magnet-room safety, where a mistake is a projectile injury.
Headcount grew steadily across the period.
Median pay $73,410 → $95,480 +4.1% 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
+7.1% 44,100 → 47,200 on the projections basis
Hard to automate, and growing
The work resists current AI and the BLS projects +7.1% 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.
~2,600 openings a year on average, including replacing people who leave.
Imaging TechnologistMRI Tech (Magnetic Resonance Imaging Technician)MRI Specialist (Magnetic Resonance Imaging Specialist)MRI Coordinator (Magnetic Resonance Imaging Coordinator)MRI Radiographer (Magnetic Resonance Imaging Radiographer)MRI Technologist (Magnetic Resonance Imaging Technologist)Travel MRI Tech (Travel Magnetic Resonance Imaging Technician)Research MRI Technologist (Research Magnetic Resonance Imaging Technologist)MRI QA Coordinator (Magnetic Resonance Imaging Quality Assurance Coordinator)MRI Special Procedures Technologist (Magnetic Resonance Imaging Special Procedures Technologist)
Holding it up: embodiment . Weakest point: judgment & accountability .
Tasks largely resist digitisation Zone IV screening, coil placement and pad-and-strap immobilization on a patient who can't hold still, IV gadolinium starts, and real-time decisions to re-run a motion-corrupted sequence are the bulk of the shift and none of them survive without hands on the patient — it's 15 rather than 18 because protocol selection, sequence parameter tuning, reconstruction, denoising and artifact detection are already being automated by vendor software.
Hands-on in uncontrolled environments You spend the shift transferring patients onto the table, positioning heads and shoulders to millimeter tolerance inside a bore, wanding for shrapnel and pacemakers, and working in a 1.5T–3T fringe field where a dropped oxygen cylinder becomes a projectile; it's 17 not 20 only because the scan room is a fixed, controlled suite rather than a roadside or a home.
Licensed human required and personally liable ARRT-MR or ARMRIT credentialing plus licensure in the majority of states, with your name on the MR safety screening form and the contrast administration record, means a ferromagnetic screening failure or a gadolinium extravasation lands on you personally — but it's 12, not 17, because a radiologist signs the diagnostic interpretation and the MR Medical Director owns the safety program above you.
Meaningful discretion You decide in the moment whether an implant with an unverifiable MR conditional label goes in the room, whether to abort a sequence for motion, and whether an unexpected finding warrants pulling the radiologist off the reading list — real calls with real stakes, but scored 11 because ACR safety guidance, protocol libraries and departmental contrast protocols pre-answer most of them.
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 (15/20 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 (12/20) is whether the law requires a licensed human to sign. Trust premium (12/20) is whether buyers specifically pay for a person. Judgment and accountability (11/20) 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 35 of this occupation's 67 points (52%).
Embodiment (17/20) 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 79/100, still SAFE.
CMS conditions of participation or a state DOH rule requiring a credentialed human to countersign AI-generated protocol selection, implant-conditionality determinations, and gadolinium dose calculations before the sequence runs — the same countersignature pattern already emerging in AI triage device labeling (FDA-cleared imaging AI is nearly all cleared as adjunct-with-human-review).
Task-mix shift is genuinely two-tiered here: if AI absorbs protocol selection, reconstruction, denoising and artifact flagging, the remaining shift is disproportionately screening interrogation, IV access on hard sticks, pediatric and claustrophobic coaching, and non-standard positioning — none of which current systems touch. Note this raises resistance per remaining worker while reducing headcount per scanner.
Universal state licensure for MRI personnel: currently only ~10-15 states license MR techs, and the federal CARE Bill (Consistent, Accuracy, Responsible Care for Everyone / RadCARE) has been reintroduced across multiple Congresses to require credentialing for personnel performing medical imaging reimbursed by Medicare. If enacted, or if The Joint Commission converts its MRI safety recommendations into a survey-scored requirement that a named ARRT-MR-credentialed MR Safety Officer/MR Safety Expert (per ACR Manual on MR Safety roles) personally sign off zone-IV access and implant clearance for every scan, the sign-and-be-liable layer hardens.
Formal absorption of MR Safety Officer duties into the staff tech role rather than a separate manager: as scanner throughput rises and implant density in the population grows (cardiac devices, neurostimulators, retained fragments), the residual job becomes conditional-implant adjudication, contrast-reaction response, and abort/continue calls on incidental findings and motion. If ABMRS certification (MRSO/MRSE) becomes an employer or accreditor requirement for the bedside tech, the role formally owns ambiguous calls.
The limit. No plausible route to a higher trust premium: patients do not select or pay extra for a specific MRI tech, and referral flows through radiologist and payer, not the tech. Also, every liability and judgment gain here is per-worker, not per-headcount — AI-driven throughput gains compress techs per scanner even as the surviving role gets harder to automate. A safer job is compatible with fewer of them.
| New York-Newark-Jersey City, NY-NJ | 3,020 | $118,360 +24% |
| Chicago-Naperville-Elgin, IL-IN | 1,930 | $100,730 +5% |
| Los Angeles-Long Beach-Anaheim, CA | 1,290 | $109,800 +15% |
| Boston-Cambridge-Newton, MA-NH | 1,160 | $123,910 +30% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 1,160 | $98,620 +3% |
| Atlanta-Sandy Springs-Roswell, GA | 1,010 | $100,180 +5% |
| Dallas-Fort Worth-Arlington, TX | 950 | $102,490 +7% |
| Houston-Pasadena-The Woodlands, TX | 820 | $102,570 +7% |
| San Jose-Sunnyvale-Santa Clara, CA | 220 | $176,950 +85% |
| Vallejo, CA | 30 | $176,650 +85% |
| San Francisco-Oakland-Fremont, CA | 420 | $146,260 +53% |
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 67. 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.