EXPOSED
Almost nothing a janitor does is text or screen work: mopping stairwells, gathering and hauling trash, scrubbing restrooms, restocking supplies, unclogging a toilet, salting an icy entryway, and responding to a spill in a hallway full of people are physical tasks in messy, unpredictable buildings that today's robots handle only in wide, flat, obstacle-free zones. What leaves this occupation exposed rather than safe is not language AI but the absence of any credential, sign-off requirement, or client relationship that a buyer pays a premium for — the work is bid on price, so autonomous floor scrubbers and route-optimization software reduce headcount per square foot even where they can't replace a person. The modal worker is a contracted or in-house cleaner in a school, hospital, office tower, or warehouse.
Roughly flat across the period, with year-to-year wobble.
Median pay $27,430 → $36,840 +7.4% 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
+2% 2,447,700 → 2,495,500 on the projections basis
Growing, and only partly exposed
The BLS expects +2% more of these jobs by 2034, and at 52/100 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.
~351,300 openings a year on average, including replacing people who leave.
CleanerJanitorSweeperScrubberCustodianHired ManDay PorterCamp TenderFloor WaxerScrub WomanWall WasherChore WorkerDuct CleanerFloor BufferFloor PersonHall CleanerHired WorkerPatch WorkerPool CleanerPower WasherWall CleanerAlley CleanerBlinds WasherBrass Cleaner
Holding it up: embodiment . Weakest point: liability shield .
Tasks largely resist digitisation Stripping and waxing a stairwell landing, chasing a coffee spill through a lobby, dumping desk-side bins that are placed differently on every floor, and snaking a plugged floor drain are all one-off physical judgments in cluttered space — the only piece that has genuinely automated is the open-floor scrub, which is why this sits at 18 and not 20.
Hands-on in uncontrolled environments Every shift is on foot with a cart: kneeling to clean toilet bases, reaching overhead vents, hauling 40-pound trash bags to a compactor, working around wet floors, bloodborne-pathogen cleanups, and salting outdoor entries in winter — 19 rather than 20 only because the environment is an indoor building rather than a road or roof.
No licence, no signature requirement No state licence gates this work; a hospital or school may require a bloodborne-pathogen or floor-safety orientation and OSHA HazCom training on the chemicals you mix, but the building owner and the contractor carry the liability if someone slips — a 3 reflects those training cards, not any credential that protects the job.
Executes defined procedures on defined inputs The work runs on written cleaning specs, chemical dilution labels, and route sheets; the discretion is real but small — deciding to wet-vac and cone off a leak before it reaches carpet, or flagging a broken lock or biohazard up to the supervisor rather than resolving it — hence 5.
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 (18/20 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 (3/20) 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 (5/20) 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 15 of this occupation's 52 points (29%).
Embodiment (19/20) 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 janitors and cleaners, except maids and housekeeping cleaners 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 69/100 — SAFE.
Union contracts and municipal responsible-contractor ordinances that price out low-bid competition: SEIU 32BJ and Local 26 master agreements, plus building-service prevailing-wage laws (e.g. Los Angeles and Minneapolis service-worker ordinances) and green-cleaning mandates that require trained named staff, shift procurement away from pure price-per-square-foot. This raises the paid-for-a-specific-human component only where such coverage expands to schools, hospitals, and Class-A office in more metros.
Genuine two-tier structure exists: routine wide-floor scrubbing and trash-route work versus restroom detail, fixture repair, spill triage, and cluttered/occupied-space cleaning. Fleet deployment of Tennant/Avidbots/Softbank machines removes the routine tier first, so the residual job is denser in what robots cannot do — but this raises resistance per remaining worker while cutting headcount, so it does not protect the occupation's size.
Healthcare-associated infection rules that name a credentialed individual: if CMS Conditions of Participation or a state hospital licensing rule required terminal-clean rooms to be signed off by a certified environmental services technician (AHE/CBIC CHEST or CSCT credential) with ATP or fluorescent-marker verification logged per room, the EVS tier of this occupation acquires a named human attestor. Precedent exists in Joint Commission surveys and in state C. diff/CRE reporting rules; the missing piece is individual rather than facility accountability.
Task-mix shift plus first-responder role formalization: as autonomous scrubbers take open-floor mopping, the remaining human day is exception handling — deciding whether a hallway needs closure, identifying a mold or asbestos disturbance and stopping work, escalating a suspected outbreak, sequencing an isolation-room clean. If facility protocols name the on-shift cleaner as the person who calls a wet-floor closure or halts a disturbance, ambiguity ownership becomes explicit rather than implicit.
OSHA/state hazmat sign-off for bloodborne-pathogen and chemical spill cleanup: if a licensed spill responder had to certify remediation of body-fluid or hazardous spills in public buildings (analogous to existing lead- and asbestos-abatement worker certification under EPA RRP), the incident-response portion becomes credentialed work.
The limit. Even with every lever, the binding constraint is that displacement here happens through headcount-per-square-foot reduction rather than task replacement. Credentialing and union coverage would plausibly lift the score into the mid-60s for the hospital-EVS and unionized-commercial segments, but the large contracted low-bid segment — warehouses, strip retail, small offices — has no realistic route to any of these, and no buyer there pays for a specific human.
| New York-Newark-Jersey City, NY-NJ | 183,030 | $42,840 +16% |
| Los Angeles-Long Beach-Anaheim, CA | 83,160 | $38,720 +5% |
| Chicago-Naperville-Elgin, IL-IN | 70,000 | $38,470 +4% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 59,090 | $37,840 +3% |
| Houston-Pasadena-The Woodlands, TX | 48,880 | $30,650 -17% |
| Dallas-Fort Worth-Arlington, TX | 48,360 | $34,860 -5% |
| Boston-Cambridge-Newton, MA-NH | 43,160 | $46,550 +26% |
| Miami-Fort Lauderdale-West Palm Beach, FL | 41,890 | $34,180 -7% |
| Longview-Kelso, WA | 510 | $46,590 +26% |
| Boston-Cambridge-Newton, MA-NH | 43,160 | $46,550 +26% |
| Seattle-Tacoma-Bellevue, WA | 26,820 | $46,490 +26% |
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 52. 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.