SAFE
The core of this job — conducting cell and body searches, breaking up fights, escorting inmates, counting heads, responding to medical emergencies and self-harm attempts — happens with hands and bodies in one of the least predictable physical environments that exists, and no robot is doing it. AI is already eating the paperwork tier: incident report drafting, log entries, visitor screening, and camera monitoring with behavior-flagging analytics, which will thin out control-room and booking posts. Employment risk here comes far more from incarceration policy, budget cuts, and chronic understaffing than from automation.
This fall is concentrated in 2020 and has not recovered since.
Median pay $45,180 → $58,940 +4.4% 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
-7.8% 387,500 → 357,400 on the projections basis
Hard to automate, but shrinking anyway
The work resists current AI, yet the BLS projects -7.8% by 2034. Whatever is shrinking this occupation, the evidence does not point to automation — demand, demographics, offshoring and industry decline all shrink jobs that no machine could do.
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.
~30,100 openings a year on average, including replacing people who leave.
GuardDeputyJailerJailorMatronGate GuardJail GuardJailkeeperPoint GuardJail OfficerPrison GuardConvict GuardDeputy JailerGate WatchmanPenal OfficerPolice MatronCottage MasterCottage ParentPrison OfficerStation JailerBooking OfficerDetention DeputyPatrol ConductorResident Monitor
Holding it up: embodiment . Weakest point: trust premium .
Tasks largely resist digitisation Pat-downs and strip searches, cell extractions, restraint application, suicide cut-downs, and pill-line supervision cannot be delegated to software, and the 17 rather than 20 reflects that formal counts, log entries, visitor screening and incident-report drafting are already being absorbed by RFID/biometric counting and video analytics.
Hands-on in uncontrolled environments You physically move through housing units, control tight corridors, take punches during fights, and go hands-on with combative people who are actively resisting — the environment includes weapons improvised from anything, bodily fluids, and no ability to pause the situation, which is about as uncontrolled as work gets.
Licensed human required and personally liable Most states require academy certification and POST-style licensure with revocation power, and 8th Amendment deliberate-indifference and excessive-force suits name officers personally, but the 11 rather than 16 reflects qualified immunity, union representation, and the fact that the agency and warden absorb most of the exposure for policy-driven outcomes.
Meaningful discretion You decide alone at 3 a.m. whether a threat justifies force, whether a suicide watch report is credible, when to call a code, and whether to write up or de-escalate, all under policy manuals and use-of-force continuums that constrain but do not resolve the call — discretion inside a rulebook rather than the open-ended authority of a warden or a shift commander.
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 (17/20 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 (11/20) is whether the law requires a licensed human to sign. Trust premium (8/20) is whether buyers specifically pay for a person. Judgment and accountability (13/20) 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 32 of this occupation's 68 points (47%).
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.
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 80/100, still SAFE.
Task-mix shift: as report drafting, log entry, and head-count verification are automated, the residual post is almost entirely the ambiguous calls — when a fight becomes a force incident, whether self-harm talk is imminent, which cell moves prevent a gang assault. This tier already exists and is separately reviewed; thinning the clerical tier concentrates it.
State corrections statutes or consent-decree terms (e.g. DOJ CRIPA settlements, California's AB 2632-style restrictions on solitary) that require a named, certified officer to personally authorize and sign each use-of-force, restraint, strip-search, or suicide-watch placement decision — explicitly barring reliance on AI risk scores as the authorizing record. Several state POST-equivalent boards already mandate officer certification; extending personal-signature liability to AI-flagged incidents is the concrete step.
PREA (Prison Rape Elimination Act) audit standards being amended to require that camera-analytics alerts be reviewed and dispositioned by a certified officer of record, with that officer's name in the audit trail — turning camera monitoring from an automatable post into a signature post.
Union contracts (e.g. AFSCME, CCPOA, NYSCOPBA) adding language that AI behavioral flags are advisory only and cannot substitute for officer discretion or be used as the sole basis for discipline — codifying that the officer owns the call.
Minimum staffing-ratio legislation or court-ordered floors (as in ongoing Alabama and Mississippi DOJ actions) that set officers-per-inmate independent of surveillance technology, blocking substitution of remote monitoring for physical posts.
The limit. trust_premium has no realistic route: the buyer is a government agency under budget pressure, and incarcerated people and their families are not paying customers who can express a preference for human staffing. embodiment is near ceiling at 19. The real employment risk remains decarceration policy and facility closures, which no dimension here captures.
| New York-Newark-Jersey City, NY-NJ | 14,730 | $93,220 +58% |
| Phoenix-Mesa-Chandler, AZ | 8,920 | $57,280 -3% |
| Houston-Pasadena-The Woodlands, TX | 8,650 | $53,810 -9% |
| Los Angeles-Long Beach-Anaheim, CA | 6,070 | $78,770 +34% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 5,570 | $67,200 +14% |
| Dallas-Fort Worth-Arlington, TX | 5,170 | $58,930 +0% |
| Miami-Fort Lauderdale-West Palm Beach, FL | 5,010 | $66,770 +13% |
| Chicago-Naperville-Elgin, IL-IN | 4,940 | $84,550 +43% |
| San Jose-Sunnyvale-Santa Clara, CA | 1,050 | $132,670 +125% |
| Salinas, CA | 1,510 | $109,620 +86% |
| Bakersfield-Delano, CA | 3,280 | $107,220 +82% |
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 68. 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.