Airline Pilots, Copilots, and Flight Engineers
53-2011.00Median wage $232,140/yr103,560 employed (US)Rank #674 of 923 scored · top 73% by substitution
Pilot and navigate the flight of fixed-wing aircraft, usually on scheduled air carrier routes, for the transport of passengers and cargo. Requires Federal Air Transport certificate and rating for specific aircraft type used. Includes regional, national, and international airline pilots and flight instructors of airline pilots.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution20
Exposure24
Augmentation62
Substitution — the headline: capability discounted by cost, barriers and adoption.
Exposure — technical capability alone, regardless of whether anyone deploys it.
Augmentation — how much AI assists without replacing. High here + moderate substitution = a changing job, not a disappearing one.
Tasks on the substitution scale
24 rated tasks, binned by substitution score.
Position among all scored occupations
Distribution of 923 occupation scores; the marker is this occupation.
Tasks with substitution ≥ 70
4%
Run 1.0.0-draft.1 · computed 2026-08-05 · rater panel: claude-sonnet-5, claude-haiku-4-5-20251001 · intervals span rater disagreement.
Why this score
The five weighted dimensions of the composite, averaged across this occupation's tasks (importance-weighted, panel mean). Exact weights and formulas: /api/v1/methodology.
Task automatabilityw 35%22
panel mean rating 1.9/5 → substitution pressure 22/100
Technical feasibility todayw 20%29
panel mean rating 2.2/5 → substitution pressure 29/100
Cost vs. human wagew 15%21
panel mean rating 1.9/5 → substitution pressure 21/100
Adoption barriersw 20%inverted — strong barriers lower the score6
panel mean rating 4.8/5 (barrier strength) → substitution pressure 6/100
Sector adoption velocityw 10%22
panel mean rating 1.9/5 → substitution pressure 22/100
Task breakdown (24 tasks)
Substitution pressure per task, weighted by O*NET importance in the composite. Expand a task for the full rater audit trail — every rating, every model, every rationale.
Record in log books information, such as flight times, distances flown, and fuel consumption.
92CI 92–92 · exposure 100 · augmentation 75 · importance 3.7/5 · click for rater detail
Record in log books information, such as flight times, distances flown, and fuel consumption.
92| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Airlines operate in a digitally mature, safety-critical sector with strong incentives to automate administrative burden; many carriers have already transitioned to electronic flight logs with automatic data population. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Commercial aviation has rapidly adopted electronic flight bags and automated data logging over the past decade, though some smaller operators and general aviation still use manual logs. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI could assist by automatically populating digital logbooks, verifying data accuracy, and flagging anomalies, allowing pilots to focus on reviewing and certifying rather than manual entry. |
| Augmentation potential | claude-sonnet-5 | 4/5 | Where manual entry persists, automated systems and apps significantly speed up and reduce errors in logging, letting pilots focus on flight operations rather than paperwork. |
| Task automatability | claude-haiku-4-5-20251001 | 5/5 | Modern aircraft avionics systems already capture flight times, distances, and fuel consumption automatically; AI could easily extract and format this data into logbooks with minimal human oversight, easily meeting the 50% time-saving threshold. |
| Task automatability | claude-sonnet-5 | 5/5 | Recording flight times, distances, and fuel data is a structured data-logging task that avionics systems and existing electronic flight bag/logbook software already automate by pulling data directly from flight management systems, meeting the time-saving threshold easily. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | While some regulatory oversight of flight records exists, there is no legal requirement that pilots manually write logbooks; digital systems are already permitted and widely used, creating minimal barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 2/5 | Some regulatory record-keeping requirements exist (FAA/EASA logbook standards), but electronic logging is already accepted and widely certified, so barriers are modest rather than a hard requirement for manual human entry. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Automated data extraction from aircraft systems costs negligibly compared to the pilot labor required to manually record this information, making the cost ratio highly favorable for automation. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Automated sensor-based logging costs a tiny fraction of pilot time spent manually recording data, representing an order-of-magnitude cost advantage. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 5/5 | Flight data logging and extraction from avionics systems is mature and widely deployed; many airlines already have automated systems that pull flight parameters directly into digital logbooks without manual pilot entry. |
| Technical feasibility today | claude-sonnet-5 | 5/5 | Deployed electronic flight logging systems (e.g., ForeFlight, airline EFB systems, ACARS data feeds) automatically capture and log this data in production across most commercial fleets today. |
Steer aircraft along planned routes, using autopilot and flight management computers.
46CI 34–57 · exposure 62 · augmentation 88 · importance 4.7/5 · click for rater detail
Steer aircraft along planned routes, using autopilot and flight management computers.
46| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Despite technical feasibility for route-following automation, adoption of full autonomous steering (beyond existing autopilot) is extremely slow due to regulatory, safety, and insurance barriers. The aviation sector remains heavily conservative on crewed flight operations, with no commercial movement toward removing pilots from the cockpit. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Autopilot technology is already near-universal in commercial aviation, but further automation toward reduced crew or single-pilot operations is progressing slowly due to regulatory and safety certification hurdles. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | Autopilot and flight management computers substantially augment pilot productivity by handling repetitive manual steering during cruise, reducing fatigue and workload. This frees pilots to focus on monitoring, contingency planning, and higher-level navigation decisions, making the task significantly easier while the human retains full authority. |
| Augmentation potential | claude-sonnet-5 | 5/5 | Autopilot and FMS dramatically reduce pilot workload during route navigation, letting pilots focus on monitoring, decision-making, and handling contingencies, representing a mature and transformative augmentation tool. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Modern autopilot and flight management systems already handle the core steering task for the majority of a flight (cruise segments). Current AI systems could fully automate route-following with minimal human intervention, meeting the 50% time-saving bar; the main limitations are regulatory certification and edge-case handling during departure/approach. |
| Task automatability | claude-sonnet-5 | 3/5 | Autopilot and flight management systems already perform the mechanical steering along planned routes for most of cruise flight, but takeoff, landing, and exception handling still require certified human pilots, so full end-to-end automation is not yet realized.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Airline pilots are federally licensed professionals, and aviation is heavily regulated by the FAA and ICAO. Regulations explicitly mandate pilot presence and require humans to legally command the aircraft; full replacement of the steering task would require major regulatory change and liability reform in a safety-critical domain. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Aviation regulations (FAA/EASA) mandate licensed pilots to be in command and monitor automated systems at all times; unsupervised automation is legally prohibited for commercial passenger flight. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | The computational cost of running autopilot and flight management systems is minimal compared to pilot labor (six-figure annual salary for crew on long-haul flights). Even accounting for system maintenance and certification overhead, automation is orders of magnitude cheaper per flight-hour for the steering function alone. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | The autopilot hardware/software is relatively cheap to operate, but regulations require full pilot crews to be present and paid regardless, so the marginal human cost is not eliminated, keeping the effective cost ratio close to comparable. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Autopilot systems demonstrably perform sustained route-following in production aircraft daily, but full autonomous navigation without pilot oversight is not yet certified or deployed. Existing systems require human initiation, monitoring, and override capability, so they perform the task under human supervision rather than independently. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Autopilot/FMS systems are mature, certified products flying billions of passenger-miles daily and reliably hold routes, altitudes, and speeds in production use across commercial aviation. |
File instrument flight plans with air traffic control to ensure that flights are coordinated with other air traffic.
44CI 18–70 · exposure 50 · augmentation 75 · importance 4.1/5 · click for rater detail
File instrument flight plans with air traffic control to ensure that flights are coordinated with other air traffic.
44| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Aviation is a heavily regulated, safety-critical sector with slow adoption cycles; airlines have not deployed autonomous flight-plan filing in production, and regulatory conservatism ensures that even mature AI solutions face multi-year certification timelines. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Aviation dispatch and flight planning is a digitized, software-heavy domain with widespread adoption of automated planning tools across major and regional airlines. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist pilots by auto-populating standard fields, suggesting optimized routes based on weather and traffic data, and checking basic compliance rules, meaningfully improving planning speed and reducing cognitive load during preflight preparation. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI-assisted flight planning tools significantly speed up route optimization, weather integration, and plan filing while the pilot/dispatcher remains responsible for final approval. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Filing instrument flight plans involves structured data entry and rule-based form completion, which AI can assist with significantly, but the task requires human validation of route legality, weather considerations, and coordination with ATC that introduces unpredictability and liability. Current systems cannot reliably achieve 50% time savings at equal quality end-to-end because regulatory compliance and real-time adjustments demand human oversight. |
| Task automatability | claude-sonnet-5 | 4/5 | Filing an IFR flight plan is a structured, data-driven process (route, altitude, fuel, alternates) that flight planning software already automates heavily, with AI able to optimize routes and auto-populate ATC-compliant plans. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Federal Aviation Regulations (FAR) mandate that a licensed pilot-in-command must file and be responsible for the flight plan; automation of this task faces hard legal barriers requiring human sign-off and accountability, with liability remaining with the certificate holder. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Regulations require a certificated pilot-in-command or dispatcher to ultimately be responsible for the accuracy and filing of the flight plan, creating a sign-off barrier even though the drafting is automated. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI systems for aviation planning require specialized integration, regulatory compliance infrastructure, and continuous human oversight, making per-task costs substantial; they remain comparable to or higher than the marginal cost of a pilot filing a plan as part of their broader duty cycle. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated flight planning software is cheap per-plan compared to dedicated human labor time, though licensing and integration with ATC systems add some cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While prototype systems can draft flight plans from standard parameters, no deployed product reliably files plans with ATC independently in production aviation environments; airlines continue to require licensed pilots to review and submit plans manually. Existing automation is partial and assistive rather than end-to-end. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Commercial dispatch and flight planning systems (e.g., Jeppesen, ForeFlight, airline dispatch software) routinely generate and file IFR plans in production today, though a human dispatcher/pilot typically reviews and submits. |
Check passenger and cargo distributions and fuel amounts to ensure that weight and balance specifications are met.
41CI 30–51 · exposure 47 · augmentation 88 · importance 4.6/5 · click for rater detail
Check passenger and cargo distributions and fuel amounts to ensure that weight and balance specifications are met.
41| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Airlines have rapidly deployed automated weight-and-balance software to assist pilots and reduce manual calculation, though pilots retain sign-off authority. Most major carriers now use such systems, reflecting quick adoption of the assistive tool layer. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Commercial aviation has widely adopted electronic flight bags and automated load/balance software across major carriers for years. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | Automated calculation and real-time alerts substantially increase pilot productivity by eliminating manual arithmetic and highlighting violations instantly. This allows pilots to focus on verification and decision-making rather than computation. |
| Augmentation potential | claude-sonnet-5 | 5/5 | Software dramatically speeds up and improves accuracy of weight and balance calculations, letting pilots verify quickly rather than compute manually. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Automated systems can calculate weight and balance from sensor inputs and manifest data, but human verification and override authority are required for safety certification. Current AI cannot fully replace the pilot's legal responsibility to sign off on these critical specifications. |
| Task automatability | claude-sonnet-5 | 3/5 | Weight and balance calculations are highly formulaic and already computed by software today, but the pilot's verification and sign-off role remains a required check even if the underlying calculation is automated. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Federal Aviation Regulations (FAR Part 91/121) explicitly require the pilot-in-command to verify weight and balance before flight and bear legal liability. This hard regulatory and liability barrier prevents substitution of AI for the human attestation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Aviation regulations (FAA/EASA) require a certificated pilot to verify and be legally accountable for weight and balance compliance before flight. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Software tools reduce computation time but cannot eliminate the pilot labor cost, since federal regulation mandates pilot sign-off. The AI cost is low, but human oversight remains mandatory and expensive relative to the small portion of pilot time saved. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated weight and balance systems cost very little per calculation compared to pilot time, though the pilot must still review it as part of pre-flight duties. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Airlines use automated weight-and-balance calculation software that provides accurate computations, but pilots must manually verify and approve outputs. No fully autonomous system legally performs this task without human oversight in production aviation. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Weight and balance software is a mature, widely deployed product in commercial aviation, though final human verification remains standard practice. |
Choose routes, altitudes, and speeds that will provide the fastest, safest, and smoothest flights.
37CI 28–47 · exposure 47 · augmentation 100 · importance 4.0/5 · click for rater detail
Choose routes, altitudes, and speeds that will provide the fastest, safest, and smoothest flights.
37| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Major airlines have already integrated route and altitude optimization tools into daily operations (FMS with real-time weather integration, automated flight planning). Adoption is widespread in commercial aviation, though the legal requirement for pilot sign-off has prevented displacement of the human role. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Airlines have adopted flight optimization software and AI-assisted dispatch tools at moderate pace, but cockpit decision authority remains human-centered with slow regulatory change. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | FMS and route optimization systems substantially augment pilot capability by automating fuel calculations, weather-optimized routing suggestions, and real-time constraint checking, allowing pilots to focus on judgment, safety monitoring, and contingency planning. This is one of the clearest examples of human–AI co-operation in safety-critical work. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI-driven flight planning, weather routing, and fuel optimization tools substantially enhance pilot decision-making today, significantly improving efficiency and safety while keeping pilots in control. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | Modern flight management systems (FMS) can compute and suggest optimal routes, altitudes, and speeds based on weather, fuel, and airspace constraints—automating a substantial portion of the analytical work. However, final decision-making and real-time judgment calls (e.g., dynamic weather avoidance, contingency replanning) still require pilot oversight, and the task is not fully automatable without human approval at current deployment levels. |
| Task automatability | claude-sonnet-5 | 2/5 | Flight planning software and autopilot systems already automate much route/altitude optimization using weather and traffic data, but final selection integrates real-time judgment, ATC negotiation, and safety accountability that current AI cannot fully replace end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Federal aviation regulations (FAA, EASA, ICAO) mandate that a licensed pilot-in-command must authorize and remain responsible for route, altitude, and speed decisions. Liability and safety-critical authority legally rest with the human pilot, creating an unbreakable regulatory barrier to full automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | FAA/international regulations require a certificated pilot to make final flight path decisions and be present in the cockpit, creating a hard legal barrier to full automation of this task. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The AI infrastructure (FMS, optimization software, integration) has high capital and maintenance costs, and pilots remain essential for oversight, decision-making, and legal responsibility. The cost per task iteration is comparable to or exceeds the marginal value of removing the human decision-maker. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | Flight planning software is cheap relative to pilot wages, but the task remains bundled with in-flight decision-making requiring a present, licensed pilot, so net cost savings are moderate rather than transformative. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed FMS and route optimization tools (e.g., Airline route planners, OOOI systems) do perform significant portions of this task in production across major carriers. Reliability is high for cruise routing and altitude selection under normal conditions, though human pilots must still review and adapt recommendations in complex or degraded scenarios. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Flight management systems and dispatch optimization tools reliably suggest optimal routes/altitudes in production today, but pilots still actively verify and adjust these recommendations rather than accepting them autonomously. |
Make announcements regarding flights, using public address systems.
36CI 26–46 · exposure 38 · augmentation 38 · importance 3.2/5 · click for rater detail
Make announcements regarding flights, using public address systems.
36| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Airlines operate in a heavily regulated, conservative sector with strong safety cultures and strong human-contact expectations from passengers. Adoption of AI-only announcements has been minimal; most systems remain human-piloted or human-supervised. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Aviation is a highly regulated, safety-critical, physically-grounded sector with slow AI adoption for crew-facing functions. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI text-to-speech and script generation could assist pilots by drafting routine announcements and reducing microphone time, but the core task of deciding what to communicate and when remains human-driven, limiting transformative potential. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI could help draft or standardize announcement scripts, but offers limited real-time assistance for the live, judgment-based delivery of flight status updates. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can generate and synthesize natural speech announcements, current systems cannot reliably handle the full task end-to-end including real-time operational changes, regulatory compliance, passenger communication nuance, and emergency protocols. The task requires human judgment for tone, timing, and discretionary content beyond scripted templates. |
| Task automatability | claude-sonnet-5 | 3/5 | Text-to-speech and script generation could produce announcements, but real-time contextual updates (delays, weather, turbulence) still require a human in the loop to decide content and timing. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Federal Aviation Regulations and airline safety protocols implicitly expect a qualified pilot to make announcements, especially in emergencies. Customer expectations, liability concerns, and regulatory oversight of pilot duties create meaningful barriers to full automation of this communicative function. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Only licensed flight crew are authorized to be in the cockpit and communicate safety-critical information to passengers, and aviation regulations require crew control over cabin communications. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | The marginal cost of AI text-to-speech synthesis is very low compared to pilot labor, but this ignores the fact that pilots perform this task as a minor part of their broader flight-critical duties; the cost comparison is distorted by bundling. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | A TTS system would be cheap per announcement, but the task is a minor sliver of pilot duties and integration into cockpit workflow adds overhead comparable to the marginal cost already absorbed by the pilot's presence. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Text-to-speech and voice synthesis systems are deployed in limited airline settings for routine announcements, but they lack the flexibility and error-handling of human pilots. Production use remains narrow and typically supplements rather than replaces human announcement duties due to regulatory and safety expectations. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Automated PA systems exist for generic boarding announcements in airports, but pilot-delivered in-flight announcements requiring real-time judgment are not handled by deployed AI products today. |
Plan and formulate flight activities and test schedules and prepare flight evaluation reports.
25CI 25–25 · exposure 25 · augmentation 63 · importance 3.1/5 · click for rater detail
Plan and formulate flight activities and test schedules and prepare flight evaluation reports.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Aviation is a heavily regulated, cautious sector with strong incumbent processes. Adoption of autonomous flight planning and evaluation tools remains minimal despite digital infrastructure; human pilot judgment remains mandated in practice. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Aviation is a highly regulated, safety-critical, and slow-moving sector with cautious technology adoption cycles, especially for flight test and evaluation functions. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist with scheduling optimization, report formatting, compliance checklist generation, and data organization, meaningfully reducing administrative burden while pilots retain control of critical decisions. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-based flight planning tools, data analysis, and report drafting assistants can meaningfully speed up schedule creation and report generation, with the pilot/engineer retaining final judgment and sign-off. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with scheduling algorithms and report template generation, the task requires safety-critical judgment about flight testing parameters, regulatory compliance, and risk assessment that demands human expertise. Current AI cannot autonomously formulate comprehensive flight evaluation plans that meet certification requirements. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can assist with drafting schedules and reports but planning flight activities and test schedules requires integration of regulatory, safety, and operational judgment that current systems cannot fully replicate end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Aviation regulations require licensed pilots to plan and evaluate flight activities, and liability for safety-critical flight testing falls on human operators who must legally sign off on all testing protocols and evaluations. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Aviation is heavily regulated (FAA/EASA), and flight test planning and evaluation reports typically require sign-off by licensed pilots/engineers, creating strong legal and safety barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Integration costs for aviation-grade AI oversight, validation infrastructure, and regulatory compliance tracking are substantial relative to the time savings from automation of planning and report drafting alone. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | While software reduces some manual computation costs, the specialized aviation expertise and certification review still needed keeps overall cost closer to human-comparable rather than order-of-magnitude cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Some scheduling tools exist and AI can generate report drafts, but no deployed system reliably handles the full scope of flight activity planning with the safety and regulatory rigor required in aviation. Human pilots must validate and take responsibility for all outputs. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Flight planning software and scheduling tools exist and are widely used, but they support human planners rather than autonomously generating final flight test schedules or evaluation reports without expert review. |
Monitor engine operation, fuel consumption, and functioning of aircraft systems during flights.
25CI 25–25 · exposure 34 · augmentation 88 · importance 4.7/5 · click for rater detail
Monitor engine operation, fuel consumption, and functioning of aircraft systems during flights.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | While modern aircraft use sophisticated automation, the regulatory and safety-critical nature of this task means adoption of autonomous monitoring (without human oversight) is extremely slow and limited to narrow, pre-approved scenarios. Most modernization still involves human pilots with better tools, not replacement. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Aviation is a highly regulated, safety-conservative sector where automation is incrementally certified over long cycles; full autonomy or removal of monitoring pilots is not occurring at scale. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI-powered engine health monitoring, predictive maintenance alerts, and fuel optimization systems demonstrably assist pilots today, surfacing patterns and recommendations that improve situational awareness and efficiency. These tools materially raise pilot productivity while keeping the human in decision-making authority. |
| Augmentation potential | claude-sonnet-5 | 5/5 | Automated cockpit systems (EICAS, ECAM, predictive maintenance alerts) significantly enhance a pilot's ability to monitor complex systems, reducing workload while keeping the pilot in the loop. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Current AI can monitor some aircraft systems and flag anomalies in engine/fuel data, but cannot reliably make judgment calls on safety implications or take corrective action in real-time without human pilots. The task requires continuous situational awareness, cross-system synthesis, and critical decisions that demand human accountability. |
| Task automatability | claude-sonnet-5 | 2/5 | Automated flight monitoring systems already handle much of this via avionics and autopilot alerts, but the task as defined includes human vigilance and judgment that remains legally required and not fully replaceable end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Regulatory law (FAA, EASA, ICAO) mandates that certified pilots must be in command and actively monitor flight-critical systems; automation cannot legally replace this oversight without explicit regulatory waiver. Liability, certification requirements, and safety-of-life stakes create hard barriers to unsupervised AI deployment. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Aviation regulations (FAA, EASA) mandate certified pilots to monitor and control aircraft systems; this is one of the most heavily regulated safety-critical tasks with legal requirements for licensed human oversight. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Implementing and maintaining robust AI monitoring systems with requisite redundancy, certification, and oversight is expensive relative to the labor it would displace. The safety-critical nature mandates high reliability, driving integration costs that approach or exceed a single pilot's loaded wage. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | The monitoring hardware/software is already embedded in aircraft costs and doesn't eliminate the need for a paid, licensed pilot in the cockpit, so no meaningful cost substitution occurs at the human-labor level. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Deployed aviation systems (FADEC, aircraft health management systems) do automated monitoring and alert pilots to anomalies, but these are assistance tools, not autonomous performers of the task. Pilots remain required to interpret, validate, and act on system alerts in production environments. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Modern aircraft have sophisticated automated monitoring (EICAS/ECAM systems, autopilot) that reliably flags anomalies in production, but these augment rather than replace the pilot's monitoring role. |
Use instrumentation to guide flights when visibility is poor.
23CI 19–28 · exposure 34 · augmentation 88 · importance 4.9/5 · click for rater detail
Use instrumentation to guide flights when visibility is poor.
23| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Aviation is the most heavily regulated safety-critical sector globally; pilots remain legally required, and AI adoption is confined to autopilot enhancements under human control. No meaningful adoption of autonomous instrument-guided flight exists in commercial operations. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Airlines have long adopted autoflight and instrument landing technology, but progress toward reducing required human pilots is slow and heavily gated by regulatory certification processes. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | Modern glass cockpit avionics with autopilot, flight management systems, and weather radar substantially augment pilot capability during poor visibility by automating routine tracking, computing guidance, and surfacing alerts—freeing the pilot to focus on decision-making and monitoring. |
| Augmentation potential | claude-sonnet-5 | 5/5 | Instrumentation, autopilot, and flight management systems massively enhance pilot capability and safety in poor visibility, representing one of the most mature and impactful human-AI augmentation use cases in any occupation. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While modern aircraft autopilot systems can navigate using instruments (ILS, GPS) in poor visibility, they cannot legally handle the full task end-to-end today: a human pilot must monitor, make discretionary decisions (go/no-go for landing, divert decisions), and maintain situational awareness. Current AI lacks the real-time contextual judgment and regulatory authority needed for full autonomy. |
| Task automatability | claude-sonnet-5 | 2/5 | Autopilot and autoland systems already handle much of instrument-guided flight, but the task as stated (pilot using instrumentation) still requires certified human oversight and manual intervention capability, so full end-to-end automation with equal-quality time savings is not realized in the human role itself. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Federal Aviation Regulations (14 CFR Part 61) mandate that a licensed pilot must be in command and responsible for navigation and safety decisions during flight. Legal liability for accidents, air-traffic-control coordination responsibilities, and crew-resource requirements create hard barriers to full AI substitution. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Aviation is one of the most heavily regulated domains; FAA/EASA rules mandate licensed pilots for command and require human sign-off and presence, especially for low-visibility operations, creating near-absolute legal barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The infrastructure, integration, certification, and per-flight liability costs of AI-driven instrument guidance automation significantly exceed the cost of a pilot's time on a per-flight basis. Safety redundancy requirements and regulatory oversight add further integration overhead. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Highly automated avionics are already installed and amortized into aircraft cost, but replacing the pilot entirely would require enormous certification, redundancy, and liability infrastructure, making full substitution far more costly than retaining a human pilot today. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Flight management systems and autopilots operationally perform instrument guidance in production, but they function as assistive tools requiring continuous pilot oversight and decision-making, not as end-to-end task performers. Autonomous flight in poor visibility remains in research/limited trials, not routine deployment. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Autoland (CAT IIIb), ILS, and flight management systems are mature and widely deployed in commercial aviation for low-visibility operations, but they augment rather than replace the certified pilot who must monitor and be ready to intervene. |
Monitor gauges, warning devices, and control panels to verify aircraft performance and to regulate engine speed.
22CI 19–25 · exposure 34 · augmentation 88 · importance 4.7/5 · click for rater detail
Monitor gauges, warning devices, and control panels to verify aircraft performance and to regulate engine speed.
22| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Aviation is highly regulated and conservative; even incremental automation in cockpits faces multi-year certification cycles. Current adoption remains limited to narrow autopilot and engine-management assistance, with no displacement of pilot monitoring roles in commercial operations. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Aviation is a highly regulated, safety-critical sector with slow, incremental adoption of new automation, constrained heavily by certification cycles and conservative regulatory oversight. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | Modern glass cockpits and engine-management displays substantially assist pilots by automatically logging, displaying, and alerting on gauge anomalies, enabling faster diagnosis and response while the pilot retains oversight and decision authority. |
| Augmentation potential | claude-sonnet-5 | 5/5 | Modern flight deck automation (autopilot, flight management systems, alerting systems) dramatically reduces pilot workload for monitoring and control tasks while pilots remain in the loop for supervision and decision-making. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While current AI can monitor and log sensor data in real-time, the task requires active regulation of engine speed based on dynamic flight conditions, which demands integrated judgment and intervention that exceeds current autonomous system capability. Existing autopilot systems handle basic monitoring but humans remain essential for decision-making during anomalies or off-nominal conditions. |
| Task automatability | claude-sonnet-5 | 2/5 | Existing autopilot and flight management systems already automate much continuous monitoring and engine regulation in cruise, but full end-to-end task coverage across all flight phases including anomaly detection and judgment calls is not achieved by general AI systems today.jav |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Federal aviation regulations (FAA) legally require licensed pilots to monitor instruments and control engines; no automation can legally replace this function without explicit certification, and the catastrophic liability of engine failure creates insurmountable legal and regulatory barriers to full autonomy. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Aviation regulations (FAA/EASA) mandate certified human pilots to monitor and control aircraft systems, with strict certification and liability requirements preventing full automation of this task without a licensed human in the loop. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The cost of AI infrastructure, certification, redundancy, and liability insurance to replace a pilot function remains far higher than the loaded wage of a trained copilot or flight engineer, given the safety-critical nature and regulatory requirements. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Avionics systems are already embedded sunk costs in aircraft, but replacing the human oversight function entirely would require certification-grade AI and redundant safety systems, which are expensive to develop and certify relative to marginal pilot cost per flight. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Modern aircraft have automated monitoring and engine management systems (FADEC) that track gauges and control engine parameters in production, but these are narrow systems with heavy human oversight. No deployed AI system independently performs continuous gauge verification and full engine regulation without human pilots in the loop. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Autopilot, autothrottle, and EICAS/ECAM alerting systems are mature deployed products that handle much routine monitoring and regulation, but they are narrow, rule-based avionics rather than generalized AI, and pilots remain required to supervise and intervene. |
Contact control towers for takeoff clearances, arrival instructions, and other information, using radio equipment.
15CI 13–18 · exposure 20 · augmentation 50 · importance 4.8/5 · click for rater detail
Contact control towers for takeoff clearances, arrival instructions, and other information, using radio equipment.
15| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Aviation is a highly regulated, safety-critical domain with strong human oversight mandates and no visible adoption of autonomous ATC radio communication in production. The sector prioritizes human authority over efficiency gains in safety-sensitive tasks. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Aviation is a highly regulated, safety-conservative sector with slow adoption of autonomous systems for flight-critical communications. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI-assisted speech recognition, transcription, and readback verification tools could support pilot efficiency and reduce workload during high-traffic phases, but the pilot must remain the primary communicator and decision-maker under current regulations. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-based speech recognition and read-back verification tools can assist pilots in parsing and confirming ATC instructions, but do not perform the communication task itself. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Radio communication involves real-time, context-dependent dialogue with air traffic control requiring understanding of ambiguous, dynamic instructions and immediate verbal response. While AI can generate text and recognize speech, current systems cannot reliably handle the safety-critical, two-way conversational nature of ATC communication or consistently parse variable phraseology and background noise without human supervision. |
| Task automatability | claude-sonnet-5 | 2/5 | Radio communication with ATC requires real-time, safety-critical exchanges that current AI cannot reliably perform end-to-end in operational cockpits; automation exists only in limited data-link/CPDLC contexts, not full voice ATC interaction.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Federal aviation regulations (FAA Part 121) legally require a licensed pilot to receive, interpret, and respond to ATC clearances. Liability for communication errors is borne by the pilot certificate holder. These hard licensing and liability barriers make substitution legally and operationally infeasible. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Aviation regulations require licensed pilots to communicate with ATC; this is a legally mandated human responsibility with major safety and liability implications. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Automating radio communication would require extensive custom integration, validation, and redundant safety systems far exceeding the cost of pilot labor for this narrow task. The regulatory and liability burden makes AI solution more expensive than the current human-centered approach. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Any AI system attempting this would require extensive certified hardware, redundancy, and regulatory approval, making it costlier than current pilot labor for this specific function. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed aviation system performs this task autonomously. Experimental speech recognition exists, but production systems require human pilots to receive, interpret, and execute ATC instructions; the FAA mandates pilot readback and confirmation. Current AI cannot replace the pilot's role in this interaction loop. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously handles live voice ATC communications for commercial pilots; digital data-link systems supplement but do not replace this task. |
Start engines, operate controls, and pilot airplanes to transport passengers, mail, or freight, adhering to flight plans, regulations, and procedures.
13CI 5–20 · exposure 17 · augmentation 75 · importance 4.9/5 · click for rater detail
Start engines, operate controls, and pilot airplanes to transport passengers, mail, or freight, adhering to flight plans, regulations, and procedures.
13| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Despite decades of aviation industry digitization, pilot automation has stalled at assistance (autopilot, FMS). No sectors are deploying fully autonomous passenger or cargo aircraft; regulatory conservatism and safety liability have frozen adoption at the human-in-command model. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Commercial aviation is highly safety-regulated and conservative in adopting autonomy for full pilot replacement, though incremental automation (autoland, fly-by-wire) has existed for decades with slow further progress. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | Autopilot, flight management systems, weather routing, and real-time data integration already substantially augment pilot productivity and safety on this task. Modern cockpits offload routine cruise and many descent tasks, allowing pilots to focus on oversight, decision-making, and emergency response. |
| Augmentation potential | claude-sonnet-5 | 4/5 | Autopilot, flight management systems, and decision-support tools significantly reduce pilot workload and improve safety and efficiency while the pilot remains in command. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While modern aircraft have autopilot and automated systems for cruise phases, the full end-to-end task—engine startup, pre-flight checks, takeoff, navigation in variable conditions, landing, and emergency response—requires real-time human judgment and situational awareness. Current AI cannot safely handle the full operation at ≥50% time savings with equal safety margins. |
| Task automatability | claude-sonnet-5 | 1/5 | While autopilot systems handle much of cruise flight, the full task of starting engines, operating controls, and piloting through all phases (taxi, takeoff, landing, emergencies) end-to-end without a human pilot is not achievable with off-the-shelf AI today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Regulatory barriers are extreme: FAA and international aviation authorities mandate human pilots with specific certifications legally responsible for flight operations. Liability, safety-critical oversight, passenger trust, and decades of regulatory precedent create near-absolute barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Aviation regulations (FAA, EASA, ICAO) mandate licensed, certified human pilots to operate commercial aircraft; liability, safety certification, and legal requirements make full automation a hard barrier. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Full autonomous flight remains extremely costly (certification, liability, redundancy, sensors) and undeployed at scale. The loaded cost of pilot labor ($150k–$300k annually) is far lower than the R&D, liability, insurance, and systems cost per flight-hour for any autonomous alternative today. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Autopilot systems are already embedded and relatively cheap to run, but full replacement of pilots would require certified redundant hardware, sensors, and regulatory compliance costs that are not yet lower than pilot wages at scale. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Autopilot and flight management systems exist in production, but they handle only portions of flight (cruise, some approach phases). No deployed AI system can autonomously manage engine start, taxi, takeoff, complex routing decisions, weather adaptation, or emergency procedures reliably enough for certification without a human pilot in command. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Advanced autopilot and autoland systems are deployed and mature for specific flight phases, but no product independently performs the full pilot task in commercial operations without a licensed human pilot present and in control. |
Perform minor maintenance work, or arrange for major maintenance.
13CI 0–25 · exposure 13 · augmentation 50 · importance 4.0/5 · click for rater detail
Perform minor maintenance work, or arrange for major maintenance.
13| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Airlines operate under strict regulatory regimes with slow-moving certification cycles; maintenance automation adoption remains minimal, with most work performed by certified personnel following established procedures. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Aviation maintenance is a highly regulated, physical, safety-critical domain with minimal AI-driven displacement; adoption is limited to predictive maintenance analytics rather than task execution. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI could assist in scheduling maintenance, flagging anomalies from sensor data, or automating documentation, but pilots and engineers would remain primary decision-makers on actual inspection and maintenance coordination tasks. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-based predictive maintenance and diagnostic tools can help flight crews and engineers flag issues or streamline scheduling of major maintenance, offering moderate assistance without performing the physical task. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Minor maintenance tasks (e.g., checking fluid levels, inspecting components) involve physical inspection and judgment in confined aircraft spaces that current AI cannot perform autonomously. Arranging maintenance requires coordination with maintainers but could be partially automated; overall, only discrete parts (like scheduling systems) are automatable today. |
| Task automatability | claude-sonnet-5 | 1/5 | Physical minor maintenance and arranging major maintenance require hands-on inspection, judgment about airworthiness, and coordination with maintenance crews—no current AI system can perform these physical and decision-making acts end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strong regulatory and liability barriers exist: FAA certification and aircraft maintenance logs require sign-off by certified technicians or pilots, and error costs (safety-critical systems) are asymmetrically high. Legal and safety requirements mandate human responsibility and accountability. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Aviation maintenance is heavily regulated (FAA/EASA), requiring certified personnel and strict sign-off procedures, making unauthorized automation legally impossible. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The cost of integrating AI vision systems, robotics, and oversight infrastructure to inspect aircraft components would exceed the loaded wage of a qualified pilot or engineer performing these checks today. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this physical/regulatory task, so any AI cost comparison is moot; human labor is the only currently functioning option. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed AI system performs end-to-end maintenance inspection or arrangement autonomously in production. Scheduling tools exist, but the diagnostic and physical inspection components that define this task remain human-dependent in real airline operations. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product performs aircraft maintenance tasks or maintenance arrangement autonomously; this remains firmly in the human/mechanical domain with only ancillary digital scheduling tools. |
Instruct other pilots and student pilots in aircraft operations and the principles of flight.
13CI 5–20 · exposure 17 · augmentation 63 · importance 3.7/5 · click for rater detail
Instruct other pilots and student pilots in aircraft operations and the principles of flight.
13| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Despite aviation's high digitization, regulatory mandates and safety criticality mean that human flight instruction remains a hard requirement; adoption of AI for actual instructional duty is near-zero because the legal and liability barriers are non-negotiable. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Aviation training is a conservative, highly regulated, physically-grounded sector where AI adoption for instructional delivery is still nascent and mostly limited to ground-school supplements and simulator analytics. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist instructors by generating drill materials, summarizing aircraft systems, or supporting pre-flight briefings, but augmentation is modest because the core task—live assessment, adaptive feedback, and error correction in safety-critical scenarios—remains firmly human-dependent. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-powered simulators, adaptive e-learning, and performance analytics meaningfully enhance how instructors prepare materials, track student progress, and provide feedback, improving instructional productivity substantially. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Instructing pilots requires real-time adaptive teaching, assessment of understanding, correction of dangerous habits, and dynamic response to student questions and performance—tasks that demand human judgment, situational awareness, and the ability to make high-stakes safety decisions that current AI cannot reliably perform end-to-end. |
| Task automatability | claude-sonnet-5 | 2/5 | Some instructional content (ground school lectures, procedures explanation) could be AI-delivered, but hands-on flight instruction, demonstration, and real-time coaching require human presence and judgment that current AI cannot replicate end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Aviation regulations (FAA, EASA, ICAO) explicitly require that flight instruction be conducted by licensed, qualified human instructors with direct accountability for student competency and safety; there is no legal pathway for AI to satisfy this requirement without a human pilot signing off. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Flight instruction is heavily regulated; only certificated flight instructors (CFIs) with specific ratings can legally sign off on training and endorse pilots, creating a hard licensing barrier against full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Pilot instruction is delivered by highly experienced, credentialed professionals commanding substantial salaries; AI systems capable of equivalent instruction do not exist at scale, and oversight costs for any partial automation would likely exceed the economic value given the small population of instructors relative to their expertise. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI-based ground training modules can be cheap, but certified flight instruction still requires a licensed human instructor in the cockpit or simulator, keeping overall costs comparable to or only marginally cheaper than human instruction. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can generate written materials on flight principles or simulate basic scenario explanations, no deployed system reliably conducts live pilot instruction with the safety accountability, real-time feedback, and dynamic problem-solving that aviation regulators and operators require; simulation aids exist but do not replace the instructor role. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI tutoring products and flight simulators with AI-assisted feedback exist, but no deployed product independently instructs student pilots in actual aircraft operations at scale; instruction remains human-led with AI as a supplementary tool. |
Confer with flight dispatchers and weather forecasters to keep abreast of flight conditions.
11CI 0–23 · exposure 13 · augmentation 50 · importance 4.4/5 · click for rater detail
Confer with flight dispatchers and weather forecasters to keep abreast of flight conditions.
11| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Aviation is among the most heavily regulated and conservatively managed sectors; adoption of automation in crew-communication roles is minimal and faces strong organizational and regulatory resistance. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Aviation has adopted digital flight planning and weather integration tools steadily, but adoption of autonomous decision-making agents in this safety-critical loop remains slow and cautious. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | While AI could potentially assist by summarizing or organizing weather data for pilot review, current systems offer limited meaningful assistance to the conferencing dialogue itself, and pilots already have mature tools (OOOI systems, weather briefings) for this purpose. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-driven weather modeling, turbulence prediction, and flight data dashboards significantly enhance the information available to pilots and dispatchers during these conferences, improving decision quality and speed. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task is fundamentally communicative and relational—conferring with human experts to assess dynamic conditions. Current AI cannot independently replace the human judgment, expertise synthesis, and two-way dialogue between pilots and dispatchers that this task requires. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can aggregate and summarize weather and flight condition data, but the collaborative, judgment-based conferring process with dispatchers involves real-time decision-making and accountability that current systems cannot fully replace end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Federal Aviation Regulations (FAA Part 121) explicitly require flight crew interaction with dispatch and weather information as part of mandatory pre-flight procedures. A licensed pilot and dispatcher are legally responsible for flight planning decisions; automation cannot substitute for this regulatory requirement. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Aviation safety regulations mandate certified pilots and dispatchers to jointly assess flight conditions, making this a hard legal and licensing barrier to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Even if automation were feasible, the cost of integrating and maintaining such a system would exceed the minimal time cost of a pilot receiving a weather briefing from existing human dispatch personnel. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | While AI weather/data tools are cheap to run, the task still requires licensed pilots and dispatchers to communicate and decide, so overall cost savings from AI are limited to information-gathering support rather than full task replacement. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product reliably performs real-time pilot-dispatcher-forecaster conferencing. This requires live integration with aviation communication systems, meteorological data, and human expert judgment in ways that are not yet productionized in the airline industry. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Flight planning software and weather aggregation tools are deployed, but the actual human-to-human conferring and negotiation over conditions and contingencies is not performed by AI products in production today. |
Order changes in fuel supplies, loads, routes, or schedules to ensure safety of flights.
9CI 7–11 · exposure 9 · augmentation 75 · importance 4.3/5 · click for rater detail
Order changes in fuel supplies, loads, routes, or schedules to ensure safety of flights.
9| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Aviation is highly regulated and conservative; adoption of AI for operational decision-making is limited to narrow advisory roles (flight planning tools), not autonomous ordering of safety-critical changes. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Commercial aviation adopts digital flight-planning and advisory tools steadily, but adoption of AI for actual command decisions is essentially nonexistent due to certification and liability regimes. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI tools effectively augment pilots by analyzing fuel burn data, optimizing routes, predicting weather impacts, and flagging schedule conflicts, materially improving decision quality and efficiency while pilots retain control. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-powered flight planning, weather prediction, and fuel optimization tools meaningfully assist pilots and dispatchers in making better-informed decisions, even though the final order remains human-issued. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires real-time decision-making under safety-critical constraints, balancing competing technical, regulatory, and operational variables. Current AI systems cannot legally or reliably make autonomous decisions about flight safety parameters that have direct liability consequences. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires real-time judgment integrating weather, mechanical status, regulatory constraints, and passenger safety with legal accountability; no off-the-shelf AI system performs this end-to-end today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Federal aviation regulations (FAA, EASA) legally require licensed pilots to make safety-critical decisions and maintain command authority; liability, certification, and human-contact requirements are absolute barriers. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Aviation regulations (FAA/ICAO) require a licensed, certified pilot-in-command to make and be accountable for these safety-critical decisions, making this one of the most legally protected task types. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI support tools exist but require significant human oversight and integration costs; the labor cost of a pilot remains high, and AI cannot eliminate the need for licensed personnel to make final decisions. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no standalone AI product substituting for the pilot's authority here; the cost comparison is not meaningful since AI cannot legally or practically replace this decision-making role. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with data analysis and recommendations (e.g., fuel calculations, route optimization), no deployed system independently orders operational changes for flights in production; human pilots retain mandatory authority and sign-off. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Decision-support tools (flight planning, weather routing software) exist and are widely used, but the actual ordering of changes remains a human command decision, not something a deployed product performs autonomously. |
Inspect aircraft for defects and malfunctions, according to pre-flight checklists.
9CI 0–18 · exposure 13 · augmentation 38 · importance 4.8/5 · click for rater detail
Inspect aircraft for defects and malfunctions, according to pre-flight checklists.
9| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Aviation is a heavily regulated, safety-first sector with extreme risk aversion and low appetite for replacing human preflight inspection. Adoption of autonomous or AI-primary inspection is negligible in commercial operations. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Aviation safety-critical physical inspection tasks show essentially no AI adoption; the sector is highly regulated and conservative regarding automation of safety sign-offs. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist pilots by cross-checking checklist compliance, flagging sensor anomalies, and organizing inspection data—raising efficiency—but the human pilot remains the decision-maker and legally responsible party. |
| Augmentation potential | claude-sonnet-5 | 2/5 | Digital checklists and some diagnostic sensor systems provide minor assistance in flagging anomalies, but the core physical inspection and judgment remain manual with limited AI augmentation. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist in analyzing sensor data and checklists, the task requires human judgment to visually inspect the aircraft, detect subtle defects, and make go/no-go decisions with safety-critical consequences. Current AI falls short of the 50% time-saving threshold for end-to-end performance. |
| Task automatability | claude-sonnet-5 | 1/5 | Physical pre-flight walkaround inspection requires visual and tactile assessment of a physical aircraft, which current AI cannot perform end-to-end without robotic embodiment and sensor deployment not in general use. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Federal aviation regulations (FAA, EASA) mandate that a licensed pilot must physically inspect and sign off on aircraft airworthiness before flight. Legal liability for missed defects is severe, and automation of this safety-critical task faces hard regulatory barriers. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Aviation regulations (FAA/EASA) mandate certified pilot or crew sign-off on pre-flight checks, making this a hard legal requirement that cannot be delegated to automation without regulatory overhaul. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Integrating AI systems for partial inspection assistance (sensor analysis, checklist optimization) requires substantial infrastructure, training, and oversight costs that approach or exceed the cost of a trained pilot performing the inspection. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this physical task, so cost comparison favors the human by default; any robotic solution would require expensive specialized hardware exceeding pilot labor cost for this narrow task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed AI system in production reliably performs autonomous pre-flight aircraft inspections independently. Some tools exist for analyzing avionics data and checklists, but visual defect detection and safety assessment remain human-dependent in all commercial operations. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously performs physical pre-flight inspections in commercial aviation; some sensor-based diagnostic tools assist maintenance crews but do not replace the pilot's checklist inspection. |
Brief crews about flight details, such as destinations, duties, and responsibilities.
7CI 0–15 · exposure 8 · augmentation 38 · importance 4.2/5 · click for rater detail
Brief crews about flight details, such as destinations, duties, and responsibilities.
7| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Aviation is a heavily regulated, safety-critical domain with strong resistance to removing human accountability from core operational tasks. There is no adoption of AI crew briefing systems in production, and regulatory and safety culture make such adoption unlikely in the foreseeable future. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Aviation is a highly regulated, safety-critical industry with slow AI adoption for operational/cockpit procedures, though AI is used more in back-office flight planning and scheduling support. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist by pre-generating briefing materials or checklists from flight data, but the briefing itself—the interpersonal communication and acknowledgment of crew understanding—remains fundamentally a pilot responsibility with limited augmentation benefit beyond information preparation. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI tools can help prepare briefing materials, weather summaries, and flight plan details that pilots then relay to crew, offering moderate productivity support. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires communicating critical safety information to human crew members in a clear, contextually appropriate manner and responding to questions. Current AI cannot reliably conduct this interpersonal, real-time communication with the judgment and accountability needed in aviation, nor would it reduce task time by 50% compared to a pilot doing it directly. |
| Task automatability | claude-sonnet-5 | 2/5 | While AI could generate briefing content, the task requires live human-to-human verbal communication, real-time crew coordination, and answering situational questions, limiting full automation despite content-generation assistance being feasible. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Federal Aviation Regulations (FAR 121.438 and similar) explicitly require a pilot-in-command to conduct a crew briefing before flight. Human authorization and sign-off for this safety-critical task is legally mandated, creating a hard regulatory barrier to full automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Aviation regulations require certificated pilots to conduct pre-flight briefings and assume responsibility for crew coordination and safety, making this a hard legal/licensing barrier. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The task involves a pilot speaking to crew members, which takes a small fraction of the pilot's salaried time. An AI system capable of reliably conducting safety-critical briefings would require substantial infrastructure and oversight costs that would exceed the minimal labor cost of the pilot simply performing it themselves. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI could cheaply draft briefing materials, but the actual briefing delivery still requires a paid, licensed pilot present, so overall cost savings are marginal relative to the full task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product in aviation currently performs this task autonomously. Pilots remain responsible for crew briefing as a regulatory requirement, and no AI system is certified or used in production to replace pilot-to-crew briefing in commercial aviation. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously conducts crew briefings in production; this remains a human-led interpersonal and procedural function with no commercial substitute in operation. |
Coordinate flight activities with ground crews and air traffic control and inform crew members of flight and test procedures.
6CI 0–11 · exposure 8 · augmentation 38 · importance 4.3/5 · click for rater detail
Coordinate flight activities with ground crews and air traffic control and inform crew members of flight and test procedures.
6| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Aviation is heavily regulated and conservative by necessity. Adoption of AI for safety-critical piloting tasks is minimal because federal law and industry standards explicitly require human pilot authority and accountability over these coordination functions. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Aviation is a highly regulated, safety-critical sector with slow AI adoption for flight-critical communication tasks, though AI is being piloted in areas like scheduling and diagnostics. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could marginally assist with procedure documentation or communication templates, but human pilots remain fully responsible and in control. The task's real-time, judgment-heavy nature and legal accountability requirements limit meaningful augmentation. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-based tools (e.g., digital checklists, automated flight data systems, predictive weather/traffic tools) assist pilots in preparing for and executing coordination tasks, improving efficiency without replacing the human role. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task fundamentally requires real-time two-way communication, judgment, and decision-making with human stakeholders (ATC, ground crews, flight crew) in safety-critical contexts. Current AI cannot autonomously coordinate these interactions or make the contextual decisions needed without human pilots actively in command. |
| Task automatability | claude-sonnet-5 | 2/5 | Real-time coordination with ATC and ground crews requires split-second verbal communication, situational judgment, and legal responsibility that current AI cannot fully replicate end-to-end in live operations. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Extremely strong regulatory and legal barriers exist: FAA regulations mandate that a licensed captain or copilot must be in command of the aircraft and directly responsible for all crew coordination and ATC communication. No automation can circumvent this requirement. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Aviation regulations (FAA, ICAO) require licensed pilots to communicate with ATC and coordinate flight operations; this is a hard legal requirement with severe liability implications. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI systems cannot yet replace human pilots in this role, making cost comparison premature. The overhead of human oversight, integration, and liability would far exceed the value of any partial automation for such a safety-critical task. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Given the safety-critical nature and lack of viable autonomous substitutes, any AI system would require extensive human oversight and certification, making it more costly than simply having a trained pilot perform this task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product reliably performs this task end-to-end. While AI can assist with communication drafting or procedure documentation, the actual coordination, real-time problem-solving, and legal authority to communicate with ATC and crew remain exclusively human pilot functions. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously handles live ATC coordination and crew briefing in commercial flight operations; this remains research/prototype territory (e.g., experimental voice-recognition ATC tools). |
Work as part of a flight team with other crew members, especially during takeoffs and landings.
4CI 0–7 · exposure 5 · augmentation 50 · importance 4.9/5 · click for rater detail
Work as part of a flight team with other crew members, especially during takeoffs and landings.
4| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Despite decades of autopilot technology, cockpit crew requirements have not diminished and no commercial adoption of unmanned or single-pilot commercial aviation exists. Regulatory and safety culture actively resist removal of human crew from this safety-critical task. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Aviation is a highly regulated, safety-critical physical sector where automation is adopted cautiously and slowly despite advances in flight automation systems. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | Modern avionics and flight management systems provide useful assistance—alerting, automation of routine procedures, data integration—that augment crew performance during coordination tasks. However, the assistance is largely procedural support rather than transformative to the core crew-team dynamic. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-enhanced avionics, alerting systems, and decision-support tools assist crew coordination and situational awareness, but the core teamwork remains human-driven. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Current AI systems cannot fully automate crew coordination and decision-making during high-stakes takeoffs and landings. While autopilot exists, it does not replace the collaborative judgment, communication, and real-time contingency management that flight crews perform together, and full automation would require removing human pilots entirely—a fundamentally different scenario than task automation. |
| Task automatability | claude-sonnet-5 | 1/5 | Crew resource management and real-time physical/verbal coordination during critical flight phases requires embodied presence and human judgment that current AI cannot replicate end-to-end.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Extremely strong regulatory and legal barriers: aviation law mandates specific numbers of licensed pilots in the cockpit, certification requirements are rigid, and liability for accidents falls on the airline and crew. No automation of this task is legally permissible without explicit regulatory approval, which remains absent. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Aviation regulations (FAA/EASA) mandate certified human pilots and specific crew complements for takeoff and landing, an absolute legal and safety barrier. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI systems that could theoretically replace crew coordination would require extensive redundancy, certification, and integration costs far exceeding pilot wages. Current autopilot and flight management systems are expensive capital investments that supplement rather than replace crew cost. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this task, so cost comparison favors the human by default; AI cannot yet deliver the output at all. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Some aircraft have advanced autopilot capable of hands-off takeoff and landing in controlled conditions, but these are narrow implementations that still require human oversight and do not perform the collaborative crew-team aspects (cross-checking, role-sharing, mutual monitoring) that define this task. Deployed systems cannot reliably replace the team coordination function. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product replaces a human crew member's role in team-based cockpit coordination; autopilot assists flight control but not team dynamics or CRM. |
Respond to and report in-flight emergencies and malfunctions.
4CI 0–7 · exposure 5 · augmentation 50 · importance 4.9/5 · click for rater detail
Respond to and report in-flight emergencies and malfunctions.
4| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Aviation is among the most heavily regulated sectors; emergency response automation is subject to extreme scrutiny and certification cycles. Even incremental automation of emergency procedures faces multi-year validation requirements, and the industry remains conservatively human-centric in safety-critical roles. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Aviation is a highly regulated, safety-critical, slow-moving sector; while cockpit automation has advanced steadily, deep AI-driven adoption for emergency handling remains minimal and cautious. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | Modern avionics and flight management systems already augment pilot response by offering real-time diagnostics, procedure checklists, and automation of routine corrective actions during emergencies. However, the augmentation is bounded by the complexity and novelty of rare, truly novel emergency scenarios. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-based decision-support tools (e.g., diagnostic checklists, predictive maintenance alerts, flight management computers) assist pilots in identifying and managing malfunctions, but the human remains fully in control of response and reporting. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | While AI can assist with diagnostics and procedure retrieval, responding to emergencies requires real-time sensorimotor control, split-second judgment under uncertainty, and legal authority that remain firmly human responsibilities today. No current system can autonomously handle the full spectrum of in-flight emergencies end-to-end. |
| Task automatability | claude-sonnet-5 | 1/5 | Responding to in-flight emergencies requires real-time physical control of an aircraft, split-second judgment, and accountability that no current AI system can perform end-to-end without a human pilot; autopilot systems handle routine flight, not emergency decision-making and crew coordination. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Pilots hold FAA certifications and legal responsibility for aircraft safety; federal regulations (FARs) explicitly require a licensed pilot-in-command to make emergency decisions and direct the aircraft. Liability, certification, and regulatory barriers are absolute. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Aviation regulations (FAA, EASA) mandate licensed, type-rated pilots to be in command and legally responsible for emergency response, with strict liability and certification requirements preventing full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The cost of integrating, validating, and maintaining AI systems for emergency handling, plus required oversight and liability management, far exceeds the cost of trained pilots whose salary is already embedded in operations. The criticality of the task pushes integration costs very high. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI-only substitute for this task, so cost comparison favors the human pilot entirely; any AI assistance adds to rather than replaces certified crew costs. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Deployed systems can provide diagnostic support and checklist guidance (e.g., integrated avionics), but no product reliably performs the full task of emergency response and reporting independently. Production aviation systems are advisory only, with the pilot retaining all decision and control authority. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously handles emergency response and reporting for commercial aircraft; existing autoland and envelope-protection systems assist but do not replace pilot judgment during malfunctions. |
Conduct in-flight tests and evaluations at specified altitudes and in all types of weather to determine the receptivity and other characteristics of equipment and systems.
0CI 0–0 · exposure 0 · augmentation 38 · importance 4.2/5 · click for rater detail
Conduct in-flight tests and evaluations at specified altitudes and in all types of weather to determine the receptivity and other characteristics of equipment and systems.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Aviation remains one of the most heavily regulated and safety-conservative sectors; in-flight testing and evaluation are core pilot responsibilities embedded in certification and operational law. No adoption of AI substitution is occurring or anticipated. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Aviation flight testing is a highly regulated, safety-critical, physical domain with minimal AI adoption for actual in-flight test conduct. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI systems can provide flight data logging, parameter visualization, and preliminary diagnostics to assist pilots during testing, but augmentation is minimal compared to the human pilot's irreplaceable role in judgment, equipment troubleshooting, and safety response. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with data analysis, simulation pre-testing, and processing telemetry from in-flight tests, but the core in-flight evaluation is human-conducted. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Conducting in-flight tests requires real-time judgment under variable weather conditions, equipment troubleshooting, safety decisions, and dynamic response to unexpected system behavior—tasks that demand human pilots' expertise, situational awareness, and legal responsibility. Current AI cannot reliably substitute for the integrated cognition and safety-critical decision-making this task demands. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires physically piloting aircraft under varied conditions and making real-time judgment calls about equipment performance; no AI system can perform this end-to-end today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Federal Aviation Regulations mandate that licensed pilots must conduct test flights and systems evaluations, and liability for aircraft safety and certification rests on human crew. Regulatory and legal barriers are absolute: automation of this task is legally prohibited without human pilot command and responsibility. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Flight testing requires licensed test pilots/flight engineers, strict FAA/aviation authority certification, and legal liability for safety-critical evaluations, making human authorization mandatory. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Automating this task would require a fully autonomous aircraft with AI capable of conducting systems diagnostics, weather analysis, and safety-critical troubleshooting—infrastructure far more expensive than deploying qualified pilots trained and certified for the role. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this task, so cost comparison favors the human by default; any automation would require extensive certified hardware/software far exceeding pilot costs. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI system independently conducts in-flight tests and equipment evaluations. While autopilot and flight management systems can maintain altitude and heading, the diagnostic evaluation, weather adaptation, and systems assessment require human expertise and certification that no autonomous AI product performs in production. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product conducts in-flight equipment test evaluations autonomously; this remains a specialized human piloting and engineering task. |
Direct activities of aircraft crews during flights.
0CI 0–0 · exposure 0 · augmentation 38 · importance 4.0/5 · click for rater detail
Direct activities of aircraft crews during flights.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Aviation is the slowest-adopting sector for crew autonomy due to stringent regulatory requirements, high safety stakes, and risk aversion. No measurable displacement of crew direction roles is occurring or expected in the near term. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Commercial aviation is highly regulated and conservative in adopting autonomous decision-making systems for crew command roles, with change occurring very slowly. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | Current flight management systems provide some assistance with navigation and monitoring, but they augment rather than transform crew coordination. AI has minimal direct impact on the captain's ability to direct activities—most augmentation remains in the technical flight domain, not crew management. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-based decision support, checklists, and monitoring systems assist pilots in coordinating crew tasks and situational awareness, but do not replace the directive function itself. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | While modern aircraft have automated flight systems, the task of directing crew activities requires real-time human judgment, communication, and decision-making under unpredictable conditions. No current AI system can reliably assume the captain's role of coordinating crew actions, handling emergencies, and making critical in-flight decisions. |
| Task automatability | claude-sonnet-5 | 1/5 | Directing crew during flight requires real-time human leadership, judgment under uncertainty, and legal command authority that AI cannot perform end-to-end today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Federal aviation regulations (FAA, EASA, ICAO) mandate that a licensed pilot-in-command must be physically present and in direct control of the aircraft. Legal and safety liability requirements create an absolute hard barrier to automation of crew direction. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Aviation regulations (FAA/EASA) mandate a licensed, certified human pilot-in-command to direct crew activities; this is a hard legal and safety-critical barrier. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The cost of deploying AI to supervise crew activities (including integration, oversight, and liability) would far exceed the salary of a single pilot or flight engineer, especially given the safety-critical nature and current lack of proven systems. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this function, so cost comparison favors the human pilot who is legally required and functionally necessary. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product today can autonomously direct airline crews during actual flights. Autopilot systems handle navigation and altitude but do not manage crew coordination, resource allocation, or emergency response—functions that remain entirely human-dependent in production aviation. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product directs flight crew activities; autopilot systems handle narrow flight control functions, not crew management and command decisions. |
Evaluate other pilots or pilot-license applicants for proficiency.
0CI 0–0 · exposure 0 · augmentation 38 · importance 3.8/5 · click for rater detail
Evaluate other pilots or pilot-license applicants for proficiency.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Aviation is a highly regulated, safety-critical sector where pilot licensing is non-negotiable and centralized under federal authority. No adoption of AI for proficiency evaluation has occurred or is permissible under current regulations. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Aviation training and certification is a highly regulated, safety-critical sector with minimal AI adoption for examiner roles, relying on structured, human-led evaluation. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist by analyzing simulator data, flagging performance anomalies, or organizing evaluation metrics, but these are minor aids to a fundamentally human gatekeeping function. The core task—making the judgment call on proficiency—remains purely human. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-based flight simulators and data analytics can support objective performance metrics and debriefing, aiding examiners without replacing their judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Evaluating pilot proficiency requires nuanced judgment about safety-critical decision-making, situational awareness, and handling of edge cases—competencies that current AI cannot assess end-to-end. While AI might score simulator performance metrics, it cannot replicate the human evaluator's ability to probe decision-making under stress, interpret non-verbal cues, or make final licensing determinations. |
| Task automatability | claude-sonnet-5 | 1/5 | Evaluating pilot proficiency requires nuanced judgment of airmanship, decision-making, and human factors in real or simulated flight scenarios that current AI cannot perform end-to-end.direct. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Federal Aviation Administration (FAA) regulations explicitly require evaluation by certified flight instructors and designated pilot examiners (DPEs). The legal authority to issue or revoke pilot licenses is vested exclusively in human examiners; no automation can substitute for this regulatory requirement. |
| Adoption barriers | claude-sonnet-5 | 5/5 | FAA/EASA regulations require licensed, designated pilot examiners to certify proficiency and sign off on checkrides, making this a hard legal barrier to automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI evaluation systems do not exist in production, making cost comparison speculative. Moreover, the liability and regulatory requirement for a human examiner means any AI system would be overhead, not replacement, keeping all-in costs above the human baseline. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this evaluative role, so cost comparison favors the human examiner entirely. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product performs pilot proficiency evaluation independently. Evaluation is conducted by human certified flight instructors and examiners; AI plays no operational role in actual checkrides or license assessments in commercial aviation. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product independently certifies or evaluates pilot proficiency; this remains a research-stage concept with no operational examiner-replacement systems. |
Related occupations — Transportation & Material Moving
How to read this
A high substitution score does not mean this job disappears — it means a large share of its current tasks face replacement pressure, so the mix of tasks is likely to change. High augmentation alongside substitution typically means the occupation reorganizes around the protected tasks. Wide confidence intervals mean the rater panel disagreed: treat those scores as open questions, not verdicts.
What would change this score
New model capabilities (automatability, feasibility), falling inference costs (cost ratio), regulation and licensing shifts (barriers), and measured sector adoption (velocity) all re-enter at every index release. Each release is recomputed, versioned and kept queryable — scores are claims with a date on them, not permanent labels.