SAFE
The heavy paperwork layer — scoring protocols, drafting psychoeducational evaluation reports, summarizing observation notes, tracking IEP compliance deadlines — is exactly what language models do well, and that is a real share of the week. But the core act of sitting with an 8-year-old to administer a WISC, watching a child in a classroom, running a crisis response after a suicide threat, and defending an eligibility determination in a contentious IEP meeting requires a credentialed human in the room who owns the call. State credentialing (NASP/NCSP or state licensure) makes the signature on an evaluation legally non-transferable.
Headcount grew steadily across the period.
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.
BLS projection, 2024–2034
+0.7% 67,200 → 67,700 on the projections basis
Hard to automate, and growing
The work resists current AI and the BLS projects +0.7% 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.
~3,800 openings a year on average, including replacing people who leave.
PsychologistPsychometristAutism ConsultantBehavioral AnalystChild PsychologistBehavior SpecialistChild PsychometristLearning ConsultantSchool PsychologistSchool PsychometristAssessment SpecialistBehavioral SpecialistConsulting PsychologistCounseling PsychologistArea School PsychologistEducational PsychologistK-12 School PsychologistEducational DiagnosticianContract School PsychologistSchool Psychology SpecialistBilingual School PsychologistCertified School PsychologistSchool Psychological ExaminerElementary School Psychologist
Holding it up: judgment & accountability . Weakest point: embodiment .
Mixed — a routine tier and a judgment tier Scoring a WISC-V or WIAT protocol, generating the boilerplate of a psychoeducational report, and tracking 60-day evaluation timelines are already substantially machine-doable, which pulls it down from the high teens, but standardized administration with a squirming first-grader, classroom observations of function, and threat/risk assessment interviews still cannot be delegated to software — hence mixed rather than resistant.
Some physical or field component You are physically in buildings — pushing a testing cart between three campuses, sitting on the floor with a kindergartener for a DAS, watching recess behavior, and physically responding when a student is in crisis in a hallway — but it is a school, not a rooftop or a roadside, so the environment is uncontrolled in behavior rather than in hazard, which is what puts this at 12 and not 17.
Licensed human required and personally liable State education agencies require a school psychologist credential (or licensure) for a psychoeducational evaluation to be legally valid under IDEA, and your name on that eligibility report is what a district defends in due process; the 14 rather than 18 reflects that you practice inside a district's legal umbrella and typically cannot bill independently or prescribe.
Exists to be accountable for ambiguous calls You decide whether a child's profile meets SLD or emotional disturbance criteria on data that rarely aligns cleanly, whether a threat is transient or substantive, and when to break confidentiality under duty-to-warn — calls with no algorithm behind them, made under IDEA and FERPA, that get litigated.
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 (12/20 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 (14/20) is whether the law requires a licensed human to sign. Trust premium (15/20) is whether buyers specifically pay for a person. Judgment and accountability (16/20) 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 45 of this occupation's 69 points (65%).
Embodiment (12/20) 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.
State education agencies or IDEA guidance explicitly requiring that any AI-assisted component of a psychoeducational evaluation be disclosed in the report and personally attested to by the credentialed school psychologist — mirroring the disclosure/attestation language already appearing in state board rules for telehealth and in APA's 2025 guidance on AI in psychological assessment. Also: test publishers (Pearson, WPS, Riverside) writing licensing terms that void score validity if protocols are administered or interpreted without a qualified examiner (Level C qualification already does much of this).
Genuine two-tier occupation: if report drafting, protocol scoring and IEP deadline tracking are absorbed by district-adopted platforms, the residual week concentrates into standardized test administration with rapport-dependent children, classroom observation, threat and suicide risk assessment, and adversarial IEP meetings — none of which current systems perform. Watch for districts raising caseload counts while cutting clerical support, which is the visible signature of this shift.
Due-process hearing decisions and OCR findings that hold the district's psychologist personally responsible for eligibility determinations where AI-generated report language was found to have produced an inadequate evaluation — making the named signer the accountable party in litigation rather than the vendor. Contentious eligibility and manifestation-determination reviews are already the pressure point.
Parent advocacy groups and special-education attorneys demanding independent educational evaluations (IEEs, already a parent right under IDEA at district expense) specifically on the grounds that the district evaluation was AI-generated. This creates paid demand for a named human evaluator rather than a cheaper automated one.
No realistic route above current level beyond what already exists — in-person test administration and classroom observation are already counted. Any rise would come from mandates for in-person rather than tele-assessment delivery, e.g. state rules restricting remote WISC administration, which some state boards tightened post-2021.
The limit. Shortage is the dominant force here, not displacement: NASP reports roughly a 1:1,065 ratio against its recommended 1:500, so tools that clear paperwork are likely absorbed as caseload relief rather than headcount cuts. The realistic ceiling is mid-to-high 70s; the residual risk is not the psychologist being replaced but districts substituting lower-credentialed staff plus software for portions of the assessment role, which would erode the liability shield rather than strengthen it.
| New York-Newark-Jersey City, NY-NJ | 7,890 | $117,540 +22% |
| Los Angeles-Long Beach-Anaheim, CA | 3,060 | $122,390 +28% |
| Chicago-Naperville-Elgin, IL-IN | 2,770 | $90,620 -6% |
| Boston-Cambridge-Newton, MA-NH | 2,340 | $98,670 +3% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 1,690 | $82,190 -14% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 1,590 | $106,550 +11% |
| Atlanta-Sandy Springs-Roswell, GA | 1,300 | $99,110 +3% |
| Houston-Pasadena-The Woodlands, TX | 1,300 | $84,850 -12% |
| El Centro, CA | 50 | $145,400 +51% |
| Boulder, CO | 110 | $143,020 +49% |
| Riverside-San Bernardino-Ontario, CA | 760 | $139,400 +45% |
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 69. 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.