SAFE
The paperwork half of this job — drafting IEP goal language, progress-report narratives, accommodation lists, data-tracking summaries — is squarely in AI's wheelhouse and will get faster. The other half is physically present work with 5-to-11-year-olds who have communication, behavioral, and motor needs: de-escalating a meltdown, hand-over-hand prompting, toileting and feeding support, reading a nonverbal child's frustration signals in real time. That part is not automatable with current robotics or software, and IDEA makes a certified human legally answerable for the plan.
Dipped in 2020, then grew past where it started.
Median pay $60,460 → $65,120 -13.8% 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
-1.8% 230,200 → 226,100 on the projections basis
Hard to automate, but shrinking anyway
The work resists current AI, yet the BLS projects -1.8% 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.
~15,400 openings a year on average, including replacing people who leave.
TeacherBlind TeacherBraille TeacherInterventionistSpecial EducatorInclusion TeacherElementary TeacherReading SpecialistBehavior SpecialistLearning SpecialistLip Reading TeacherResource SpecialistHome Therapy TeacherEarly InterventionistPublic School TeacherSign Language TeacherSpecial Needs TeacherAcademic InterventionistBehavior InterventionistHearing Impaired TeacherLearning Support TeacherPrimary Special EducatorResource Program TeacherEmotional Support Teacher
Holding it up: trust premium . Weakest point: liability shield .
Tasks largely resist digitisation A 14 reflects the split: goal-bank drafting, progress monitoring graphs, and prior-written-notice boilerplate are already being generated by IEP software, but discrete-trial instruction, physical prompting through a fine-motor task, and running a functional behavior assessment on a child who bolts when transitions are announced are the daily core and none of it survives being moved to a screen.
Hands-on in uncontrolled environments A 15 rather than a 19 because the work is unambiguously hands-on — blocking a head-banging episode, escorting a student to the sensory room, hand-over-hand tracing, diapering a nine-year-old — but it happens in a resource room and a school building you control, not on a roof or a roadside, so the environment is unpredictable in behavior rather than in terrain.
Licensed human required and personally liable State special education certification plus the IDEA requirement that a qualified teacher serve on the IEP team and deliver services as written puts this at 13: your signature is on the IEP and a due-process complaint names your implementation, but the district and the LEA representative absorb the financial judgment, so the personal exposure stops short of a physician's or an engineer's seal.
Exists to be accountable for ambiguous calls A 15 covers calls that no procedure resolves: deciding whether a behavior is a manifestation of disability before a removal, judging whether to recommend a more restrictive placement over a parent's objection, choosing when to restrain versus wait it out, and reporting suspected abuse when a child's disclosure is fragmented — each is high-stakes, contested, and yours in the moment.
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 (14/20 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 (13/20) is whether the law requires a licensed human to sign. Trust premium (18/20) is whether buyers specifically pay for a person. Judgment and accountability (15/20) 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 46 of this occupation's 75 points (61%).
Embodiment (15/20) 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 86/100, still SAFE.
As goal-drafting and progress-data summarization automate, the residual role concentrates in the genuinely contested calls: least-restrictive-environment placement, manifestation determinations in discipline cases, restraint/seclusion decisions, and eligibility disputes with parents' advocates. This is a real two-tier job and the judgment tier is the tier with legal exposure
Same task-mix shift: if narrative and data-tracking work is absorbed by district-adopted IEP software, the remaining day is behavior intervention, physical prompting, and crisis response — raising the resistant share of the job even as total hours fall
Already near ceiling — parent advocacy groups and special-education attorneys actively pressure districts against automation-mediated services, and caseload caps in union contracts (e.g., Chicago Teachers Union, UTLA special-ed provisions) price a human per student
IDEA/state-board rules explicitly requiring that IEP content generated with AI assistance be reviewed, edited, and signed by the certified case manager, with AI-drafted goals disallowed as the sole basis for placement decisions — several state education agencies (e.g., California, Ohio) have already issued AI guidance memos naming the IEP team's non-delegable authority; codifying this in state regulation or as an OSEP dear-colleague letter would harden it
Due-process hearing and OCR complaint outcomes that treat unreviewed AI-drafted IEPs as procedural denial of FAPE, making districts require a named certified teacher's attestation on every document — one published hearing decision on this point would propagate through district counsel guidance fast
The limit. At 75 the score is already near the practical ceiling for a public-sector role; the realistic risk is not displacement but headcount compression — paraprofessional substitution and larger caseloads justified by AI-assisted paperwork throughput. Dimension scores can hold while positions shrink.
| New York-Newark-Jersey City, NY-NJ | 22,270 | $91,610 +41% |
| Dallas-Fort Worth-Arlington, TX | 15,810 | $64,830 +0% |
| Houston-Pasadena-The Woodlands, TX | 12,680 | $65,080 +0% |
| Chicago-Naperville-Elgin, IL-IN | 11,210 | $79,690 +22% |
| Los Angeles-Long Beach-Anaheim, CA | 6,410 | $96,970 +49% |
| Austin-Round Rock-San Marcos, TX | 6,190 | $60,750 -7% |
| Atlanta-Sandy Springs-Roswell, GA | 5,860 | $77,960 +20% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 5,660 | $74,290 +14% |
| San Francisco-Oakland-Fremont, CA | 3,440 | $105,290 +62% |
| Oxnard-Thousand Oaks-Ventura, CA | 490 | $104,810 +61% |
| Seattle-Tacoma-Bellevue, WA | 1,740 | $100,150 +54% |
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 75. 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.