SAFE
Autopilot already flies most of the cruise phase, and flight planning, fuel calculation, weight-and-balance and weather briefing are increasingly software-generated — but the pilot exists to handle the off-nominal case: engine failure, windshear, a diversion with a medical emergency and a closing weather window. Single-pilot and remote operations are technically plausible and are being studied, yet FAA certification, ALPA, insurers and passenger tolerance make crew reduction a decade-plus regulatory fight, not a software update. Median worker here is a regional or narrowbody first officer/captain holding an ATP with type ratings and personal legal responsibility for the aircraft.
Dipped in 2020, then grew past where it started.
Median pay $147,220 → $232,140 +26.1% 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
+3.9% 100,000 → 103,900 on the projections basis
Hard to automate, and growing
The work resists current AI and the BLS projects +3.9% 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.
~11,700 openings a year on average, including replacing people who leave.
PilotCaptainCopilotCo-PilotJet PilotLine PilotCheck AirmanAirline PilotFighter PilotAirbus CaptainAircraft PilotAirplane PilotMilitary PilotAir Force PilotAirline CaptainAircraft CaptainHelicopter PilotFirst Officer (FO)Navy Fighter PilotArmy Helicopter PilotRegional Airline PilotAirline Transport PilotCommercial Airline PilotFacilities Flight Check Pilot
Holding it up: liability shield . Weakest point: trust premium .
Mixed — a routine tier and a judgment tier Autoland and FMS-coupled RNAV approaches have automated the repeatable parts of the flight, but the manual, un-proceduralised residue is real work — hand-flying a raw-data circling approach into a non-precision field, negotiating a re-route with ATC while running a QRH abnormal, deciding to take contaminated-runway braking action reports at face value or not — which is why this sits at 13 and not down with dispatch paperwork.
Hands-on in uncontrolled environments You are strapped into the machine you are operating, working rudder and thrust levers in turbulence, doing exterior walkarounds in ramp weather, and physically commanding an evacuation — but the flight deck is a climate-controlled, standardised cockpit with everything within arm's reach, which keeps this at 14 rather than up with linemen on poles.
Licensed human required and personally liable An ATP certificate with type rating, a first-class medical, 14 CFR 91.3 giving the PIC final authority over the aircraft, and personal exposure to FAA certificate action or NTSB findings after any deviation — the captain's name is on the airworthiness release acceptance and the logbook, and no software can hold that certificate.
Exists to be accountable for ambiguous calls Continue or divert with 45 minutes of fuel and a thunderstorm cell over the destination, accept or refuse an aircraft with a deferred MEL item, reject a takeoff at 130 knots — these are unrepeatable calls made in seconds with incomplete information and hundreds of lives attached, and they are yours alone under 91.3(b).
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 (13/20 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 (18/20) 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 (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 48 of this occupation's 75 points (64%).
Embodiment (14/20) 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 82/100, still SAFE.
Codification of two-pilot minimum crew in statute rather than operations specifications — e.g. an FAA reauthorization amendment barring certification of reduced-crew (eMCO/single-pilot) Part 121 operations absent a rulemaking, of the kind ALPA has pushed since the 2023 reauthorization cycle; also EASA's eMCO workstream being closed out rather than advanced. Would lock the licensed-signatory requirement (PIC signs the dispatch release, holds 14 CFR 91.3 final authority) against erosion.
Genuine two-tier structure: as EFB software, ATC datalink and predictive dispatch absorb the routine tier (fuel, W&B, route, weather), the residual role concentrates in off-nominal handling, checklist arbitration under conflicting failures, and crew resource management. No law required; the measured task mix shifts toward the tier current systems cannot do.
Post-incident regulatory reaction that formally expands PIC authority over automation — e.g. an NTSB recommendation after an automation-related event requiring documented pilot override authority and pilot sign-off on any AI-generated diversion, fuel or dispatch recommendation, mirroring the post-MAX MCAS rulemaking on flight-crew alerting.
Insurer/lessor covenants: hull and liability underwriters (Lloyd's aviation syndicates, AIG) or aircraft lessors writing two-crew and minimum-experience clauses into policies and lease agreements, making reduced crew uninsurable regardless of certification status.
Contractual passenger-facing guarantees — corporate travel policies or charter/fractional (NetJets, Flexjet) contracts specifying two ATP-rated humans aboard, plus any high-profile carrier marketing crew qualification. Narrow route: leisure passengers do not shop on crew size.
The limit. Liability shield is near its practical maximum at 18; the realistic contest is defending it, not raising it. Trust premium is capped low because the buyer is the airline, not the passenger — fare-driven demand has never rewarded crew staffing, and cargo (where no passenger opinion exists) is where reduced crew will be tested first. Total realistic headroom is a few points, and the register is already SAFE.
| New York-Newark-Jersey City, NY-NJ | 10,030 | $335,270 +44% |
| Atlanta-Sandy Springs-Roswell, GA | 8,700 | $618,090 +166% |
| Los Angeles-Long Beach-Anaheim, CA | 6,710 | $364,730 +57% |
| Chicago-Naperville-Elgin, IL-IN | 6,360 | $293,490 +26% |
| Dallas-Fort Worth-Arlington, TX | 5,980 | $212,810 -8% |
| Denver-Aurora-Centennial, CO | 5,600 | $307,490 +32% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 5,100 | $209,750 -10% |
| Miami-Fort Lauderdale-West Palm Beach, FL | 4,810 | $216,070 -7% |
| Atlanta-Sandy Springs-Roswell, GA | 8,700 | $618,090 +166% |
| San Francisco-Oakland-Fremont, CA | 2,900 | $374,680 +61% |
| Los Angeles-Long Beach-Anaheim, CA | 6,710 | $364,730 +57% |
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.