EXPOSED
Sequencing arrivals, issuing routine clearances, and flight-strip/data handoffs are exactly the kind of constrained optimization automation handles well, and FAA decision-support tools (TBFM, DataComm, remote-tower video) already absorb pieces of the job. What resists is real-time deconfliction when weather, emergencies, equipment failures, and non-compliant pilots collide — a controller owns separation and the outcome, under FAA certification and a strong union. Modal worker is a certified professional controller at a tower or en route center, not a supervisor.
Roughly flat across the period, with year-to-year wobble.
Median pay $122,990 → $148,080 -3.7% 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
+1.2% 24,100 → 24,400 on the projections basis
Growing, and only partly exposed
The BLS expects +1.2% more of these jobs by 2034, and at 63/100 the work is only partly exposed — some tasks are automatable, the core of the job is not. Nothing here is in tension.
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.
~2,200 openings a year on average, including replacing people who leave.
DispatcherControl OperatorAirway ControllerFlight ControllerFlight DispatcherAirline DispatcherControl SpecialistEnroute ControllerTraffic SupervisorAir Traffic ManagerAir Route ControllerFlight Radio OfficerAircraft CommunicatorFlight Radio OperatorSignal Tower OperatorCommunications OfficerControl Tower OperatorTechnical CommunicatorTower Control OperatorAir Traffic CoordinatorCommunications OperatorFlight Tower DispatcherTraffic Control ManagerAirport Tower Controller
Holding it up: liability shield . Weakest point: trust premium .
Mixed — a routine tier and a judgment tier Metering arrival streams and reading standard clearances is bounded optimization that TBFM and DataComm already partially execute, but a 12 rather than a 6 reflects the shifts where a thunderstorm cell parks over the final, a Cessna goes NORDO, and you have to invent a plan for eight aircraft by voice in ninety seconds — work no current system sequences unaided.
Some physical or field component The job is a scope, a landline, and a headset, which alone would put it near 3, but a 9 recognizes tower controllers still work visually out the cab windows in weather, physically pass strips, and cannot be remote — plus the medical, sleep, and rotating-shift constraints that tie the function to a specific body in a specific facility.
Licensed human required and personally liable A 17 tracks the FAA CTO certificate and facility rating you cannot work a position without, the Class II medical, the fact that separation losses are logged against your operating initials and investigated by the NTSB and FAA, and that no algorithm can hold the certificate — the reason it isn't 20 is that ATO is a federal employer and the government, not you personally, absorbs civil damages under the FTCA.
Exists to be accountable for ambiguous calls When two aircraft are converging and one crew reads back the wrong altitude, the call on who turns and how far is yours in the seconds you have, made under 7110.65 rules that tell you the minima but not the resolution — a 17 rather than 20 because those minima, LOAs, and standard emergency procedures do constrain the option set more than an ICU physician's.
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 (17/20) 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 (17/20) 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 42 of this occupation's 63 points (67%).
Embodiment (9/20) 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 72/100 — SAFE.
Post-incident regime shift: if an NTSB report on a runway incursion or loss-of-separation event involving decision-support automation assigns the failure to over-reliance on the tool, and the FAA responds by formally naming the controller as the accountable authority for accepting/rejecting machine advisories (as happened after the 2023-25 incursion cluster and the Reagan National midair review), the role's ownership of the ambiguous call is tightened rather than diluted.
Codification in 14 CFR / FAA Order 7110.65 of an explicit rule that separation authority cannot be delegated to an automated system — i.e. any AI-generated clearance must be issued or affirmatively approved by a CPC holding a facility rating. NATCA has already pushed 'human-in-the-loop' language into remote/contract tower and TBFM deployment discussions, and ICAO Annex 11 amendments on automated ATS are in play.
Genuine two-tier structure: if TBFM/DataComm/ATD-2 absorb routine sequencing, strip management and standard clearance issuance, the residual day becomes weather reroutes, emergency handling, equipment-outage degraded-mode operations and non-compliant aircraft — the tier automation handles worst. Watch for FAA staffing standards that redefine positions around 'exception handling' rather than traffic count.
No realistic route upward; the work is already screen- and voice-mediated, and remote-tower video pushes it further off-site. Omitted as a lever except insofar as facility-presence requirements (FAA restrictions on remote operation of Class B towers) keep the controller physically at a specific console.
The limit. Trust premium has no plausible route: passengers and airlines do not choose or pay for a specific human controller, and there is no market signal to attach a preference to. The whole defense here is statutory and institutional — FAA certification plus NATCA — not customer demand. That makes the score durable but brittle: a single FAA reauthorization cycle that authorizes automated separation in low-density airspace, or contract-tower expansion under remote-tower certification, moves liability_shield and task_resistance together and fast.
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 1,800 | $174,870 +18% |
| New York-Newark-Jersey City, NY-NJ | 1,010 | — |
| Chicago-Naperville-Elgin, IL-IN | 850 | $189,780 +28% |
| Atlanta-Sandy Springs-Roswell, GA | 810 | $186,640 +26% |
| Dallas-Fort Worth-Arlington, TX | 800 | $186,510 +26% |
| Miami-Fort Lauderdale-West Palm Beach, FL | 720 | $162,140 +9% |
| Los Angeles-Long Beach-Anaheim, CA | 690 | $173,020 +17% |
| Minneapolis-St. Paul-Bloomington, MN-WI | 620 | $179,670 +21% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 220 | $198,590 +34% |
| Sacramento-Roseville-Folsom, CA | 240 | $196,290 +33% |
| San Francisco-Oakland-Fremont, CA | 430 | $193,360 +31% |
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 63. 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.