EXPOSED
Roughly half this job — building schedules, tracking productivity metrics, writing status reports, reviewing clerical output for errors, drafting procedure documents — is exactly the screen work current AI handles at usable quality, and it shrinks further as the clerical headcount being supervised shrinks. What persists is people management: hiring and firing, coaching a struggling records clerk, absorbing an escalated customer or a payroll error, and being the human who answers to management when a process fails. There is no license or signature requirement here, so the moat is purely relational and organizational, and span-of-control expansion means fewer supervisors covering larger teams.
Roughly flat across the period, with year-to-year wobble.
Median pay $56,620 → $69,500 -1.8% 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
-0.3% 1,558,400 → 1,554,400 on the projections basis
Exposed, and shrinking
Both signals point the same way: the tasks are largely automatable and the BLS projects -0.3% by 2034. This is the case where the score and the forecast agree, and it is the one worth taking seriously.
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.
~144,500 openings a year on average, including replacing people who leave.
Weigh BossOffice ManagerPayroll MasterMail SupervisorMailing ManagerPayroll ManagerProperty MasterRate SupervisorService ManagerSupport ManagerBusiness ManagerCargo SupervisorClerk SupervisorCustomer ManagerDispatch ManagerFiles SupervisorPractice ManagerRoute SupervisorStock SupervisorWarranty ManagerAdmitting OfficerClaims SupervisorFront End ManagerOffice Supervisor
Holding it up: judgment & accountability . Weakest point: liability shield .
Mixed — a routine tier and a judgment tier Shift rosters, queue-volume dashboards, error-rate audits of clerical output and SOP write-ups are already generated end-to-end by workforce-management and LLM tools, which is why this sits at 9 rather than mid-teens — only the disciplinary conversation, the performance-improvement plan, and the interview panel genuinely resist, and those are perhaps a third of the week.
Some physical or field component The 5 covers walking the floor: checking that the front desk is actually covered at 8am, retrieving a misfiled paper record, standing at a jammed high-volume copier or mail meter, and being physically present in a cubicle bank — real but incidental duties in a climate-controlled office, not fieldwork.
No licence, no signature requirement Nothing here is licensed or credentialed; a supervisor can be promoted from a records clerk seat on Monday with no exam, and when a payroll batch or a records-retention lapse turns into an FLSA or HIPAA problem, it is the employer and the HR or compliance officer named, not the supervisor's signature.
Meaningful discretion Calls on whether an absence pattern becomes a write-up, how to handle a clerk who mishandled a confidential file, and which escalated customer complaint gets a credit are real discretion exercised without a script — capped at 11 because HR policy, the union contract, and the manager above set the boundaries and sign off on termination and any spend.
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 (9/20 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 (1/20) 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 (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 21 of this occupation's 35 points (60%).
Embodiment (5/20) 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.
Gambling Managers EXPOSED
Sales Managers EXPOSED
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 47/100, still EXPOSED.
Employment-law exposure concentrating in the supervisor role: if EEOC/state enforcement and case law continue to hold that an identified human decision-maker must make and document adverse employment actions (discipline, termination, accommodation denials) rather than an algorithmic scoring system — as NYC Local Law 144 and the Illinois AI Video Interview Act already gesture at for hiring — the supervisor becomes the mandatory named decider on every personnel call, and that tier of the job cannot be delegated to software.
Genuine two-tier structure: if scheduling, metric dashboards, status reporting and error QC are fully absorbed by workflow software, the residual day is coaching, conflict, escalated-customer absorption, and hiring/firing conversations — none of which current systems do at usable quality. Watch for job postings that drop 'produce reports' and lead with 'coach, develop, retain'.
A narrow route only in regulated back offices: if FINRA branch-supervision rules, HIPAA-covered records units, or state notary/vital-records offices extend named-principal-supervisor sign-off to AI-produced clerical output (e.g. FINRA Rule 3110 supervisory review of communications applied to AI-drafted correspondence), a subset of these supervisors gains a real personal-accountability signature.
Union or works-council contract language requiring a named human supervisor to review any automated productivity-based discipline before it takes effect — the pattern in recent warehouse and public-sector contracts and in California SB 7 style 'automated decision systems' bills covering notice and human review.
The limit. Trust premium has no realistic route up: buyers of administrative work are internal management, not clients choosing a human, and span-of-control expansion cuts the other way. Even with every lever, span consolidation means fewer people holding a somewhat more defensible job.
| New York-Newark-Jersey City, NY-NJ | 95,180 | $80,760 +16% |
| Los Angeles-Long Beach-Anaheim, CA | 54,340 | $77,530 +12% |
| Dallas-Fort Worth-Arlington, TX | 45,710 | $73,070 +5% |
| Houston-Pasadena-The Woodlands, TX | 36,940 | $65,990 -5% |
| Chicago-Naperville-Elgin, IL-IN | 33,180 | $71,730 +3% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 28,900 | $72,700 +5% |
| Miami-Fort Lauderdale-West Palm Beach, FL | 26,510 | $67,550 -3% |
| Atlanta-Sandy Springs-Roswell, GA | 25,800 | $69,990 +1% |
| San Jose-Sunnyvale-Santa Clara, CA | 7,630 | $95,010 +37% |
| San Francisco-Oakland-Fremont, CA | 19,480 | $88,560 +27% |
| Longview-Kelso, WA | 410 | $81,980 +18% |
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 35. 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.