SAFE
The job is operating a 40-foot vehicle full of unsupervised children through residential streets, railroad crossings, snow, and double-parked cars — plus managing behavior, verifying students board and exit safely, and handling medical or behavioral incidents in transit. Autonomous driving is nowhere near unsupervised child transport, and even a fully self-driving bus would still need an adult aboard for supervision and liability, which is the expensive part. Route optimization and dispatch paperwork are already software; the seat behind the wheel is not.
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
+0.2% 387,300 → 388,200 on the projections basis
Hard to automate, and growing
The work resists current AI and the BLS projects +0.2% 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.
~61,000 openings a year on average, including replacing people who leave.
Bus DriverBus MonitorCampus DriverStudent DriverSchool Bus DriverSchool Van DriverShuttle Bus DriverSchool Bus OperatorBus Drive CoordinatorSubstitute Bus DriverPublic School Bus DriverSPED Driver (Special Education Driver)SPED Bus Driver (Special Education Bus Driver)CDL Bus Driver (Commercial Driver's License Bus Driver)SPED School Bus Driver (Special Education School Bus Driver)CDL School Bus Driver (Commercial Driver's License School Bus Driver)
Holding it up: embodiment . Weakest point: judgment & accountability .
Tasks largely resist digitisation Route sheets and dispatch logs are already digitized, but the daily core — sweeping the bus at end of route for a sleeping kindergartener, activating the crossing arm and physically escorting a child across a live traffic lane, stopping and opening the door at every rail crossing, securing wheelchair tiedowns — is a sequence of physical judgment calls no vendor is piloting without a driver, which is why this sits at 15 rather than into the high teens: the navigation itself is the one part that is being automated.
Hands-on in uncontrolled environments You are in the vehicle in whatever weather the district doesn't call a snow day for, doing pre-trip inspections on brakes, mirrors, and emergency exits in a cold yard at 5:30am, backing a 40-footer in a cul-de-sac, and walking the aisle to break up a fight — 19 because the entire shift is uncontrolled outdoor and in-vehicle physical work, short of the top only because you are seated for most of it.
Licensed human required and personally liable A CDL with the S (school bus) and P (passenger) endorsements, DOT medical card, state fingerprint and background clearance, and an FMCSA-regulated drug and alcohol program mean a positive test or a preventable crash ends your ability to hold the job anywhere — 14 rather than 18 because the district and its insurer absorb the civil damages, and your exposure is licence-level and administrative rather than personal malpractice liability.
Meaningful discretion The written framework is heavy — stop-arm procedure, railroad crossings, evacuation drills, discipline referral forms, when to radio dispatch rather than decide — but you personally call whether an unfamiliar adult at the stop gets the child, whether a hill is too icy to attempt, and whether a diabetic or seizure incident means pulling over and calling 911 or continuing to the school nurse, which is real unscripted authority and lands it at 10.
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 (15/20 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 (13/20) is whether buyers specifically pay for a person. Judgment and accountability (10/20) 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 37 of this occupation's 71 points (52%).
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 83/100, still SAFE.
Formalization of the driver's non-driving duties into named, auditable responsibility: state rules making the driver the accountable party for student-count reconciliation at each stop (post-'child left on bus' fatality statutes, several states now mandate end-of-route sweep and electronic child-check systems with driver sign-off), for railroad-crossing and evacuation decisions, and as a Title IX / mandated child-abuse reporter for on-bus incidents. If districts add crisis-response and de-escalation certification requirements after behavioral incidents, the ambiguity-under-consequence portion of the role is documented rather than invisible.
Explicit state statutes barring driverless operation of school buses and requiring a CDL-holding, S-and-P-endorsed adult in the driver's seat regardless of automation level — the pattern in existing state AV bills that carve school buses out of driverless authorization (e.g., bills in NY, Georgia and Texas that exempt or prohibit school-pupil transport from AV pilots), plus NHTSA FMVSS 222/school bus rules being amended to name a responsible operator. Watch also for state pupil-transportation regs making the driver the mandated reporter and the person who signs the pre-trip inspection log with personal liability.
School board policy or district RFP language requiring a human driver plus, for routes with pre-K, special-needs, or IEP-transport students, a second credentialed adult aide — already contractual in many districts under IDEA-related transportation plans. If parent groups convert AV pilots into explicit board resolutions banning unmanned pupil transport, the human-in-seat requirement becomes a purchased good, not just a technical default.
Task-mix shift only weakly applies: routing, dispatch, and attendance logging are already software, so what remains is the driving-plus-supervision tier. If automation handles the highway/arterial portion of routes but the driver retains stop-side loading, wheelchair securement, and behavior management, the residual job is entirely the tier machines cannot do — but this raises the score only marginally because the seat is already the job.
The limit. The binding constraint is district budgets, not capability: driver shortages have already pushed districts to consolidate routes, shorten service areas, and shift to parent-transport stipends or contracted rideshare-style vans for small loads. Headcount can fall sharply while every dimension score stays high — the role is displacement-resistant but not headcount-resistant.
| New York-Newark-Jersey City, NY-NJ | 29,280 | $59,790 +25% |
| Chicago-Naperville-Elgin, IL-IN | 15,210 | $49,180 +3% |
| Atlanta-Sandy Springs-Roswell, GA | 11,190 | $47,310 -1% |
| Houston-Pasadena-The Woodlands, TX | 9,730 | $43,570 -9% |
| Dallas-Fort Worth-Arlington, TX | 9,030 | $47,030 -2% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 8,980 | $51,960 +8% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 8,620 | $57,610 +20% |
| Minneapolis-St. Paul-Bloomington, MN-WI | 8,070 | $50,240 +5% |
| San Jose-Sunnyvale-Santa Clara, CA | 610 | $75,230 +57% |
| San Francisco-Oakland-Fremont, CA | 2,350 | $73,680 +54% |
| Mount Vernon-Anacortes, WA | 200 | $72,070 +50% |
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 71. 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.