SAFE
The job is physical safety work in a moving metal tube: securing cabins, operating exit doors, restraining unruly passengers, administering oxygen and first aid, and evacuating 180 people in 90 seconds. FAA regulation requires a minimum number of certificated attendants per aircraft based on seat count, which is a hard staffing floor no software can lower. AI erodes only the paperwork edges — service scripting, catering counts, irregular-ops rebooking messages — not the core.
Dipped in 2020, then grew past where it started.
Median pay $56,640 → $63,580 -10.2% 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
+9.2% 130,800 → 142,900 on the projections basis
Hard to automate, and growing
The work resists current AI and the BLS projects +9.2% 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.
~19,800 openings a year on average, including replacing people who leave.
PurserStewardFlight CrewFlight HostessFlight StewardMeal AttendantAirline HostessCabin AttendantFlight AttendantAirline AttendantAirline StewardessFlight Crew MemberAircraft Cabin AgentIn-Flight Crew MemberRamp Flight AttendantAirline Flight AttendantAirplane Flight AttendantDoor and Arrival AttendantInternational Flight AttendantInflight Services Flight Attendant
Holding it up: embodiment . Weakest point: judgment & accountability .
Tasks largely resist digitisation At 15, the tasks that survive are the ones that require a body in the aisle — arming door slides, checking overhead bin latches row by row, hands-on CPR and AED use, physically wrestling an intoxicated passenger into a seat — while the automatable slice (PA announcements, meal preference tracking, connecting-gate lookups on the crew tablet) is real enough to keep it out of the 17-plus range.
Hands-on in uncontrolled environments 19 reflects an office that is 35,000 feet up and vibrating: pushing a 200-pound cart uphill on a pitching floor, working in 6-8% cabin humidity and reduced oxygen partial pressure, and being expected to open a 60-pound overwing exit and slide-evacuate strangers in smoke, which is about as uncontrolled as environments get short of firefighting.
Licensed human required and personally liable The FAA Certificate of Demonstrated Proficiency is a real, personally-held certificate with annual recurrent training and checkrides under 14 CFR 121.417, and you are the crewmember whose lawful instruction passengers must obey under 49 USC 46504 — but you don't sign the airworthiness release or hold the operating certificate, which is why this sits at 13 and not with pilots at 18.
Meaningful discretion At 13 you make calls the manual only frames: whether an ill passenger warrants a diversion recommendation to the captain, whether to deny further alcohol service, whether the door goes or stays shut when there's fire outside your exit — high-stakes discretion, but bounded by an FOM, checklists, and a captain who holds final authority under 14 CFR 91.3.
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 (13/20) 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 (13/20) 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 39 of this occupation's 73 points (53%).
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.
FAA rulemaking or statute tightening 14 CFR 121.391 ratios — e.g. codifying the 1-per-50-seats floor against carrier petitions for single-attendant relief on small jets, or the AFA-CWA-backed push to keep minimum crew counts explicitly in the FAA Reauthorization Act. Certificate-holding attendants remain the legally required, individually liable safety crew under FAA enforcement (they hold personal certificates since 2004 and can be fined/decertified).
Task-mix shift: the occupation genuinely has two tiers. If AI absorbs the service tier (pre-order fulfillment, catering, IROPS rebooking on crew tablets, passenger comms), the residual job is emergency response, medical judgment, and conflict de-escalation — the tier no current system does. This raises the fraction of the day AI cannot do without any rule change.
Expanded in-flight medical and self-defense authority: if carriers formalize attendant authority to deviate from telemedicine ground advice (MedAire/STAT-MD), to call for diversion, or if DOT rules on unruly-passenger restraint place the decision to restrain/deplane explicitly with the purser rather than the captain, the role owns more consequential ambiguous calls.
Mandatory recurrent training scope expansion — e.g. FAA-required naloxone/AED/tourniquet and self-defense training (already in FAA Reauthorization language), each addition deepening the certificated-human requirement per aircraft.
The limit. Trust premium has no realistic route: passengers buy fares on price and schedule, not on crew identity, and no carrier has ever successfully charged for human cabin service as such. Embodiment is already near maximum at 19. Total realistic headroom is a few points on the institutional side; the score is high because the FAA staffing floor already does the work.
| New York-Newark-Jersey City, NY-NJ | 18,030 | $127,740 +101% |
| Los Angeles-Long Beach-Anaheim, CA | 10,080 | $79,390 +25% |
| Chicago-Naperville-Elgin, IL-IN | 9,240 | $62,120 -2% |
| Miami-Fort Lauderdale-West Palm Beach, FL | 8,100 | $47,320 -26% |
| Dallas-Fort Worth-Arlington, TX | 7,680 | $61,100 -4% |
| Denver-Aurora-Centennial, CO | 6,930 | $59,500 -6% |
| Phoenix-Mesa-Chandler, AZ | 4,510 | $61,780 -3% |
| San Francisco-Oakland-Fremont, CA | 4,420 | $89,430 +41% |
| New York-Newark-Jersey City, NY-NJ | 18,030 | $127,740 +101% |
| San Francisco-Oakland-Fremont, CA | 4,420 | $89,430 +41% |
| Columbus, OH | 290 | $83,450 +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 73. 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.