EXPOSED
The work is almost entirely physical and interpersonal: monitoring patients on a psychiatric unit, assisting with bathing, dressing and meals, escorting patients between activities, and physically intervening during agitation or restraint situations. None of that is text-on-a-screen work, and current robotics cannot manage an unpredictable human body in crisis. The real exposure is not AI substitution but the thin paper trail around the role — no license protects it, documentation and shift-report writing will be AI-assisted, and staffing decisions are driven by budgets and monitoring tech rather than by capability.
This fall is concentrated in 2020 and has not recovered since.
Median pay $31,110 → $44,910 +15.5% 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
-0.4% 38,500 → 38,300 on the projections basis
Shrinking, but not obviously because of AI
The BLS projects -0.4% by 2034, but at 64/100 this work is only moderately exposed — not the profile of a job current AI can simply do. Occupations shrink for many reasons, and the score does not point at automation as this one's cause.
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.
~5,300 openings a year on average, including replacing people who leave.
Charge AideBehavior AideBehavioral AideCharge AttendantPsychiatric AideBehavior AssistantClinical AssistantDevelopmental AideDirect Care WorkerPsychiatric OrderlyBehavioral AssistantHealth Service WorkerMental Health OrderlyNeuropsychiatric AidePsychiatric AssistantPsychiatric AttendantMental Health AssistantMental Health AssociateMental Health Aide (MHA)Psychiatric Nursing AideBehavior Support AssistantMental Health Worker (MHW)Behavioral Health AssistantBHA (Behavioral Health Aide)
Holding it up: embodiment . Weakest point: liability shield .
Tasks largely resist digitisation Fifteen-minute safety checks, one-to-one observation of a suicidal patient, hands-on assistance with toileting and showering, and de-escalating or holding a patient during a takedown are tasks no software performs; the 3 points off reflect that shift notes, behavior logs and incident documentation are already being templated and voice-dictated.
Hands-on in uncontrolled environments You are on a locked unit in physical contact with unpredictable bodies — blocking a swing, sitting on a mattress on the floor with someone in seclusion, walking a group to the courtyard, cleaning up after incontinence — and the environment is defined by the fact that it cannot be controlled.
No licence, no signature requirement Most states require only a high school diploma plus employer-run CPI/crisis-intervention and CPR training; the RN or attending physician signs the orders for restraint and medication, so when something goes wrong the investigation lands on the nurse's judgment and your employer's staffing plan, not a license you hold and could lose.
Meaningful discretion You decide in real time whether pacing is escalating and needs a nurse now, whether to close observation distance, whether contraband is worth a search — real calls with real consequences, made inside written unit protocols and a chain of command you are expected to escalate to rather than substitute for.
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 (4/20) 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 27 of this occupation's 64 points (42%).
Embodiment (20/20) 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.
Massage Therapists SAFE
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 85/100 — SAFE.
State minimum-staffing-ratio statutes for behavioral health inpatient units that count direct-care aides by name (following Oregon HB 2697, 2023, which set enforceable ratios including CNAs/techs on behavioral health units) — making the headcount legally mandated rather than budget-discretionary
Deployment of AI patient-monitoring (e.g., Oxehealth, Care.ai vision systems on psych units) where the aide is designated the human who must clear or escalate each algorithmic alert, and that disposition is logged against their credential in incident review — ownership of the false-negative shifts explicitly onto the aide
Task-mix shift as shift reports, rounding logs and observation charting move to ambient AI documentation: the residual role concentrates in crisis response, restraint-alternative de-escalation and bodily assistance — genuine two-tier structure here, with the clerical tier being the automatable one
Extension of California-style Psychiatric Technician licensure (BVNPT, Business & Professions Code) to other states, or a state scope-of-practice rule making a certified psych tech the person who must be physically present and personally attest to restraint/seclusion monitoring checks under CMS Conditions of Participation §482.13(e) — turning an aide's initials into a licensed, individually accountable signature
Joint Commission or state survey findings that make the aide's real-time decision to initiate hands-on intervention versus continue verbal de-escalation a documented, reviewable clinical call rather than an undocumented reflex
Private psychiatric and residential treatment facilities marketing guaranteed human 1:1 observation coverage (rather than camera-based watch) as a purchased differentiator for suicide-risk patients and their families
The limit. Embodiment is already maxed at 20 and cannot rise. The score gap is almost entirely institutional: without licensure or a named-signature requirement, this role stays a budget line, and the main real-world threat is headcount reduction justified by monitoring tech, not AI doing the work. Trust premium is structurally capped because the buyer is a hospital administrator, not the patient.
| New York-Newark-Jersey City, NY-NJ | 4,310 | $49,840 +11% |
| Los Angeles-Long Beach-Anaheim, CA | 1,450 | $47,050 +5% |
| Jackson, MS | 1,040 | $31,560 -30% |
| Salem, OR | 910 | $50,580 +13% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 870 | $48,560 +8% |
| Indianapolis-Carmel-Greenwood, IN | 780 | $45,080 +0% |
| Detroit-Warren-Dearborn, MI | 580 | $45,410 +1% |
| Boston-Cambridge-Newton, MA-NH | 570 | $48,100 +7% |
| Kiryas Joel-Poughkeepsie-Newburgh, NY | 210 | $75,670 +68% |
| Trenton-Princeton, NJ | 370 | $60,110 +34% |
| Utica-Rome, NY | 220 | $58,600 +30% |
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 64. 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.