Commercial Pilots

53-2012.00
Median wage $123,220/yr47,630 employed (US)Rank #693 of 923 scored · top 75% by substitution

Pilot and navigate the flight of fixed-wing aircraft on nonscheduled air carrier routes, or helicopters. Requires Commercial Pilot certificate. Includes charter pilots with similar certification, and air ambulance and air tour pilots. Excludes regional, national, and international airline pilots.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution19
Exposure22
Augmentation57

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

8%

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%21

panel mean rating 1.9/5 → substitution pressure 21/100

Technical feasibility todayw 20%24

panel mean rating 2.0/5 → substitution pressure 24/100

Cost vs. human wagew 15%22

panel mean rating 1.9/5 → substitution pressure 22/100

Adoption barriersw 20%inverted — strong barriers lower the score7

panel mean rating 4.7/5 (barrier strength) → substitution pressure 7/100

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.

Consider airport altitudes, outside temperatures, plane weights, and wind speeds and directions to calculate the speed needed to become airborne.

83

CI 76–90 · exposure 92 · augmentation 88 · importance 4.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510015/5Commercial aviation is a digitized, safety-critical sector with rapid adoption of automation. Flight management systems that compute these parameters are standard on all modern commercial aircraft and have been in widespread use for decades.
Sector adoption velocityclaude-sonnet-55/5Commercial aviation has near-universal adoption of EFB-based performance calculation tools, replacing manual chart lookups across major airlines for years.
Augmentation potentialclaude-haiku-4-5-202510014/5AI systems augment pilots by providing accurate, real-time calculations and decision support that reduce cognitive load and error. Pilots retain final authority while benefiting substantially from automated computation and display of required airspeed.
Augmentation potentialclaude-sonnet-55/5These tools substantially speed up and improve accuracy of pilot decision-making, with the pilot remaining the accountable authority who reviews and confirms outputs.
Task automatabilityclaude-haiku-4-5-202510015/5This task is a pure mathematical calculation involving well-defined inputs (altitude, temperature, weight, wind) and established aerodynamic formulas. Modern flight management systems and AI can compute required airspeed end-to-end with 100% accuracy and near-zero time cost, easily meeting the 50% threshold.
Task automatabilityclaude-sonnet-54/5This is a well-defined computation using standard performance charts/formulas already automated in flight management systems and takeoff performance calculators, easily replicable by software with high reliability.
Adoption barriersclaude-haiku-4-5-202510013/5While the calculation itself can be fully automated, regulatory frameworks (FAA, EASA) and airline procedures typically require pilot involvement in flight planning and decision-making, creating organizational and procedural friction even though the technical barrier is minimal.
Adoption barriersclaude-sonnet-54/5Regulatory frameworks require certified performance data and typically mandate pilot verification/sign-off of takeoff speeds, so a licensed pilot must still confirm the calculation even though software performs it.
Cost vs. human wageclaude-haiku-4-5-202510015/5The inference cost of this calculation is negligible (microseconds of computation on existing avionics), while a human pilot performing manual calculation incurs significant time cost. AI cost is orders of magnitude lower.
Cost vs. human wageclaude-sonnet-55/5Once integrated, the marginal cost of running a takeoff performance calculation is negligible compared to pilot time, and the software is amortized across thousands of flights.
Technical feasibility todayclaude-haiku-4-5-202510015/5Flight management systems in commercial aircraft already perform this calculation automatically and reliably in production every day. Dedicated aviation software and AI-based tools exist and are deployed at scale across the industry.
Technical feasibility todayclaude-sonnet-55/5Electronic Flight Bags and integrated FMS/takeoff performance calculators are deployed industry-wide today and reliably compute V-speeds from these exact inputs in production airline operations.

Write specified information in flight records, such as flight times, altitudes flown, and fuel consumption.

73

CI 57–87 · exposure 83 · augmentation 88 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510014/5Commercial aviation is highly digitized and regulated; major carriers already use electronic flight bags and automated logbook systems, showing rapid adoption of digital record-keeping across the industry.
Sector adoption velocityclaude-sonnet-53/5Aviation is adopting electronic flight bags and digital logging steadily, but full automation of flight records is uneven across carriers and aircraft types, with legacy paper/manual processes still common in smaller operators.
Augmentation potentialclaude-haiku-4-5-202510015/5AI augmentation (auto-population of data from aircraft systems, anomaly detection) dramatically reduces pilot workload in post-flight administration while maintaining human oversight of accuracy and sign-off.
Augmentation potentialclaude-sonnet-54/5AI and automated systems significantly reduce the manual burden of recording flight data, letting pilots focus on verification rather than manual entry.
Task automatabilityclaude-haiku-4-5-202510015/5Flight data logging is a highly structured task with standardized inputs (altitude, fuel, time) from aircraft systems that are already digitally captured; automation can extract and populate records at >50% time savings with no quality loss compared to manual entry.
Task automatabilityclaude-sonnet-54/5Logging standardized flight data (times, altitudes, fuel) is largely a data-transcription task that avionics systems already capture digitally and could auto-populate records with minimal human input, though final sign-off is still expected.
Adoption barriersclaude-haiku-4-5-202510014/5Regulatory bodies (FAA, EASA) require pilots or authorized personnel to sign off on flight records for legal compliance; while data capture can be automated, human certification is mandated, creating a hard barrier to full substitution.
Adoption barriersclaude-sonnet-53/5Regulatory record-keeping requirements (FAA/EASA) mandate accurate, often pilot-certified logs, creating documentation and liability requirements even if the data capture itself is automated.
Cost vs. human wageclaude-haiku-4-5-202510015/5Automated data extraction from avionics costs pennies per flight; manual pilot entry costs loaded pilot wages (~$100+/hour); the cost ratio favors automation by orders of magnitude.
Cost vs. human wageclaude-sonnet-54/5Automated data logging via onboard sensors and software is far cheaper per record than manual pilot transcription, though some integration and system costs exist.
Technical feasibility todayclaude-haiku-4-5-202510015/5Aircraft already generate digital flight data automatically (FDR, QAR systems); software to parse and populate electronic logbooks is mature and widely deployed in commercial aviation operations today.
Technical feasibility todayclaude-sonnet-53/5Many aircraft already have digital flight data recorders and electronic logbook systems that auto-fill much of this information, but universal integration and fully automated record generation across all commercial fleets is not yet standard.

Monitor engine operation, fuel consumption, and functioning of aircraft systems during flights.

48

CI 25–71 · exposure 62 · augmentation 88 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510015/5Commercial aviation has been systematically adopting automated monitoring and decision-support systems for decades; glass cockpits and predictive maintenance platforms are now standard across large carriers and modern aircraft, with rapid deployment of AI-augmented diagnostics in major fleet operations.
Sector adoption velocityclaude-sonnet-52/5Aviation is a highly regulated, slow-moving sector where new automation must undergo lengthy certification, so despite advanced avionics, actual adoption of AI-driven autonomous monitoring replacing pilots is minimal and slow.
Augmentation potentialclaude-haiku-4-5-202510015/5AI systems dramatically augment pilot productivity by automating real-time alerts, trend analysis, and diagnostic interpretation; pilots retain decision authority while AI reduces cognitive load and flag-detection time, enabling safer and more efficient operations.
Augmentation potentialclaude-sonnet-54/5Flight management systems, engine indication and crew alerting systems (EICAS/ECAM), and predictive maintenance analytics substantially augment a pilot's ability to monitor systems and detect anomalies in real time.
Task automatabilityclaude-haiku-4-5-202510015/5Modern aircraft already feature extensive automated monitoring systems (e.g., Engine Indication and Crew Alerting Systems, EICAS) that continuously track engine parameters, fuel consumption, and system status, displaying alerts and diagnostics that reduce pilot workload by >50% compared to manual monitoring alone. AI-enhanced predictive maintenance and real-time anomaly detection could further automate the detection and interpretation of system malfunctions.
Task automatabilityclaude-sonnet-52/5Automated monitoring systems and autopilot already track engine and system parameters, but the human pilot's continuous oversight and judgment role in the cockpit remains largely intact, so this doesn't yet meet the 50% time-saving bar for the task as performed by the pilot.'
Adoption barriersclaude-haiku-4-5-202510015/5Regulatory authorities (FAA, EASA, ICAO) legally mandate that a licensed pilot must be present in the cockpit and responsible for monitoring critical systems; pilots must personally certify flight readiness and respond to anomalies, creating a hard licensing and liability barrier that prevents full substitution of human monitoring.
Adoption barriersclaude-sonnet-55/5Aviation regulations (FAA/EASA) mandate certified pilots to monitor and control aircraft systems, and liability, safety certification, and passenger trust create very strong barriers to removing human oversight.
Cost vs. human wageclaude-haiku-4-5-202510014/5AI-driven monitoring systems cost a fraction of pilot salaries, and airlines have already absorbed avionics costs into fleet operations; the marginal cost of advanced AI monitoring is negligible compared to paying human pilots, making the cost ratio strongly favorable to automation.
Cost vs. human wageclaude-sonnet-52/5Onboard monitoring systems are already integrated into aircraft cost, but replacing the human monitoring function entirely would require additional certified autonomous systems, which are costly to develop and certify relative to marginal pilot labor cost.
Technical feasibility todayclaude-haiku-4-5-202510014/5Deployed aviation systems (Boeing, Airbus glass cockpits, airline fleet management systems) already automate substantial engine and system monitoring in production; however, the requirement for human pilot situational awareness and sign-off means the task cannot be performed purely end-to-end by AI without human oversight, limiting it to near-full rather than complete feasibility.
Technical feasibility todayclaude-sonnet-53/5Modern avionics and flight management systems reliably monitor engine and fuel data and alert crews to anomalies, but these are decision-support tools rather than full replacements for pilot monitoring duties.

File instrument flight plans with air traffic control so that flights can be coordinated with other air traffic.

37

CI 25–49 · exposure 38 · augmentation 75 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Commercial aviation is highly regulated and risk-averse; adoption of any autonomous flight planning changes is slow and constrained by certification processes (type certification, operational approval). Current industry practice retains human pilots in the planning loop; no measurable displacement of this task by autonomous systems is occurring in production fleets.
Sector adoption velocityclaude-sonnet-53/5Aviation has adopted digital flight planning tools widely, but full AI-driven autonomous filing without pilot review remains uncommon due to regulatory and safety constraints.
Augmentation potentialclaude-haiku-4-5-202510013/5AI systems that pre-populate flight plan forms, suggest optimal routes, or integrate real-time weather and airspace data assist pilots in faster, more accurate plan construction. Tools that draft or validate instrument flight plans against constraints do improve pilot productivity on this task while keeping the human in command and responsible for final approval and filing.
Augmentation potentialclaude-sonnet-55/5Software dramatically speeds up route generation, fuel calculations, and filing, transforming pilot productivity while keeping the pilot as final authority.
Task automatabilityclaude-haiku-4-5-202510012/5Filing an instrument flight plan involves structured data entry (route, altitude, aircraft type) that is highly automatable, but the task also requires real-time coordination, weather assessment integration, and compliance verification with current airspace restrictions—elements that demand human judgment and oversight. Current AI could assist with form population but cannot fully replace the human pilot's situational awareness and decision-making in coordination with ATC.
Task automatabilityclaude-sonnet-53/5Flight plan filing is largely form-based data entry that software already handles (route optimization, fuel/weight calcs), but a licensed pilot must review and file/accept it, limiting full end-to-end automation.dictating
Adoption barriersclaude-haiku-4-5-202510014/5Federal Aviation Regulations (14 CFR) and international civil aviation rules require the aircraft captain to be responsible for flight planning and adherence to filed plans. Filing with ATC is part of a legally mandated human duty; automation cannot substitute for the pilot's legal accountability. Liability and regulatory requirements create hard barriers to full autonomy.
Adoption barriersclaude-sonnet-54/5Regulatory frameworks (FAA/ICAO) require a certificated pilot-in-command to ultimately verify and be responsible for the filed flight plan, creating a hard licensing/liability barrier to full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5The task is a minor fraction of a pilot's duties and labor cost. Automating this one task does not substantially reduce pilot workload or wages, as pilots must remain present and legally responsible for all flight operations. AI implementation costs (integration with existing flight planning systems, validation, oversight) likely exceed the marginal labor savings from this narrow task.
Cost vs. human wageclaude-sonnet-54/5Automated flight planning tools cost a small subscription fee versus the pilot's time and cognitive load, making software-assisted filing far cheaper per instance.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product reliably performs end-to-end instrument flight plan filing autonomously in production. Airlines use electronic flight plan systems (EFPS) that are semi-automated but require human pilots to initiate, review, and approve filings before submission to ATC. AI systems exist to parse and draft plan components but lack the integrated, real-world deployment required for consistent reliability.
Technical feasibility todayclaude-sonnet-53/5Flight planning software (e.g., ForeFlight, Jeppesen) reliably generates and files IFR flight plans in production today, though pilots still review and confirm before submission.

Request changes in altitudes or routes as circumstances dictate.

36

CI 0–73 · exposure 45 · augmentation 75 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510014/5Major airlines and aviation operators have already adopted AI-driven flight optimization, fuel management, and dispatch systems; these are in active production use across the industry, demonstrating rapid adoption in this highly digitized sector.
Sector adoption velocityclaude-sonnet-51/5Commercial aviation is a highly regulated, safety-critical sector with extremely slow adoption of autonomous decision-making systems for flight control communications.
Augmentation potentialclaude-haiku-4-5-202510015/5AI systems provide transformative assistance to pilots by continuously analyzing real-time weather, air traffic, and fuel data to recommend optimal altitude and route changes, significantly enhancing pilot decision-making and operational efficiency while the pilot retains control.
Augmentation potentialclaude-sonnet-53/5AI-based flight planning, weather analysis, and decision-support tools can help pilots anticipate needed route/altitude changes, but the actual request and communication remain human-driven.
Task automatabilityclaude-haiku-4-5-202510015/5Monitoring flight conditions and making data-driven decisions to request altitude or route changes are tasks current AI systems can perform reliably. Modern flight management systems and AI-assisted tools can analyze weather, air traffic, and fuel optimization to suggest changes that would save substantial flight time and resources.
Task automatabilityclaude-sonnet-51/5This requires real-time judgment, radio communication with ATC, and split-second decision-making based on weather, traffic, and aircraft status that current AI cannot autonomously execute in the cockpit today.rating
Adoption barriersclaude-haiku-4-5-202510015/5Federal aviation regulations (FAA, EASA) mandate that licensed pilots must make final decisions on altitude and route changes; automation may assist but cannot legally replace human pilot authority and sign-off on safety-critical decisions.
Adoption barriersclaude-sonnet-55/5FAA/ICAO regulations require a licensed, certified pilot to communicate with ATC and make real-time flight decisions, making this a hard legal and safety barrier.
Cost vs. human wageclaude-haiku-4-5-202510015/5AI-based flight optimization systems cost far less than pilot labor per task execution; fuel savings and route optimization alone typically justify deployment costs many times over, making the cost per optimization decision orders of magnitude cheaper than human pilot time.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this specific task, so no meaningful cost comparison exists; the human pilot is mandatory.
Technical feasibility todayclaude-haiku-4-5-202510014/5Flight management and optimization software exists in production use today, though human pilots retain final authority on most operations. Many airlines deploy AI-assisted dispatch and route optimization systems that recommend altitude/route changes, though full end-to-end automation without pilot sign-off remains limited by regulation.
Technical feasibility todayclaude-sonnet-51/5No deployed product allows an AI system to independently request altitude/route changes from ATC in commercial aviation; this remains a pilot-only function under strict regulatory oversight.

Obtain and review data such as load weights, fuel supplies, weather conditions, and flight schedules to determine flight plans and identify needed changes.

35

CI 25–45 · exposure 47 · augmentation 88 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5While aviation is digitized, adoption of autonomous flight-plan modification is slow because existing cockpit-crew workflows are deeply integrated with regulatory requirements and safety culture that favor human decision authority over algorithmic substitution.
Sector adoption velocityclaude-sonnet-53/5Aviation has adopted digital flight planning tools broadly, but the industry is cautious and highly regulated, so deeper AI-driven automation progresses slowly.
Augmentation potentialclaude-haiku-4-5-202510014/5AI systems demonstrably assist pilots by rapidly aggregating and highlighting relevant weather, weight, and fuel data, enabling faster scenario analysis and decision-making while the pilot retains final authority and judgment.
Augmentation potentialclaude-sonnet-55/5AI-assisted flight planning tools significantly speed up data aggregation and route optimization, letting pilots focus judgment on exceptions while staying fully in the loop.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can retrieve and summarize weather, weight, and fuel data, determining safe flight plans requires integration of real-time operational constraints, regulatory compliance, and human judgment about risk trade-offs that current AI systems cannot reliably execute end-to-end without substantial human oversight.
Task automatabilityclaude-sonnet-53/5Flight planning software already aggregates weather, fuel, and load data and can compute optimal routes, but final review and judgment on anomalies still requires pilot interpretation and sign-off.
Adoption barriersclaude-haiku-4-5-202510015/5Federal Aviation Regulations (FAR) mandate that a licensed commercial pilot retain authority over flight planning and decisions; automation cannot substitute for the legally required human sign-off, creating a hard regulatory barrier to full automation.
Adoption barriersclaude-sonnet-55/5Aviation regulations require a certificated pilot-in-command to review and approve the flight plan before flight, creating a hard legal barrier to full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI-assisted data retrieval is inexpensive, but the task's safety-critical nature means integration, validation, and regulatory sign-off requirements make total cost comparable to or potentially exceeding a pilot's time for independent review.
Cost vs. human wageclaude-sonnet-53/5Software reduces manual calculation time substantially and is cheap per use, but licensed pilot review/oversight time still adds significant cost, keeping it roughly comparable rather than order-of-magnitude cheaper.
Technical feasibility todayclaude-haiku-4-5-202510013/5Data aggregation and formatting tools exist and are deployed in flight planning systems, but reliable autonomous determination of modified flight plans with safety-critical implications remains narrow in scope and typically requires human pilot validation before implementation.
Technical feasibility todayclaude-sonnet-54/5Dispatch and flight planning systems (e.g., Jeppesen, ForeFlight) are mature and widely deployed in production, though pilots still manually review and confirm outputs.

Plan flights according to government and company regulations, using aeronautical charts and navigation instruments.

31

CI 25–36 · exposure 30 · augmentation 75 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Adoption remains slow and limited to narrow decision-support tools rather than autonomous planning. Aviation is a highly regulated, safety-critical sector where automation adoption is conservative and incremental. No evidence of industry-wide shift toward AI-led flight planning in production.
Sector adoption velocityclaude-sonnet-53/5Aviation has adopted computerized flight planning and dispatch tools extensively, but full AI-driven autonomous planning without human review remains rare in production due to regulatory constraints.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist with chart interpretation, weather briefing synthesis, regulatory requirement lookup, and route optimization suggestions, raising dispatcher and pilot productivity. However, the human remains the authoritative planner and decision-maker, making this solidly assistive rather than transformative.
Augmentation potentialclaude-sonnet-55/5AI-powered flight planning and navigation software dramatically speeds up route optimization, weather integration, and regulatory compliance checks, greatly enhancing pilot productivity while keeping the pilot in the loop.
Task automatabilityclaude-haiku-4-5-202510012/5Flight planning involves complex decision-making around weather, airspace regulations, fuel optimization, and safety margins that currently requires human judgment. While AI can assist with route calculations and regulatory lookups, the task requires integration of real-time data, regulatory interpretation, and safety accountability that today's systems cannot reliably handle end-to-end without substantial human oversight.
Task automatabilityclaude-sonnet-52/5Flight planning software already automates much of the computation, but the task as stated includes regulatory compliance judgment and final responsibility that current AI cannot fully assume end-to-end without human sign-off.
Adoption barriersclaude-haiku-4-5-202510014/5Regulatory barriers are substantial: FAA regulations require a licensed pilot or authorized dispatcher to file flight plans and make final decisions on route legality and safety. Liability for navigation errors and accident causation rests with accountable humans, creating strong legal and contractual barriers to full automation.
Adoption barriersclaude-sonnet-54/5Aviation regulations (FAA/EASA) require a certificated pilot in command to review and accept the flight plan, creating a hard legal barrier to full automation of this task.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI tools for flight planning (weather parsing, route optimization) require significant integration and expert human oversight to validate. The cost of reliable AI assistance plus required human review likely approaches or exceeds the cost of a pilot or dispatcher performing the task directly.
Cost vs. human wageclaude-sonnet-53/5Flight planning software is already cheap and widely used, but the pilot's oversight and legal review time still factor into cost, keeping it roughly comparable rather than order-of-magnitude cheaper for the full task.
Technical feasibility todayclaude-haiku-4-5-202510012/5No production systems perform full flight planning autonomously. Some airline dispatch systems use AI-assisted tools for weather analysis and route optimization, but human pilots and dispatchers remain decision-makers. Regulatory and liability frameworks require human sign-off, and no deployed product reliably replaces the human planner.
Technical feasibility todayclaude-sonnet-53/5Commercial flight planning software (e.g., dispatch systems) reliably generates routes, fuel calculations, and weather-optimized plans in production, but a licensed pilot or dispatcher must review and approve them.

Plan and formulate flight activities and test schedules and prepare flight evaluation reports.

25

CI 25–25 · exposure 25 · augmentation 63 · importance 3.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Aviation remains a conservative, safety-critical sector with slow digital transformation in administrative workflows; while some airlines use scheduling software, AI-driven evaluation reporting and test planning have not yet seen meaningful production adoption.
Sector adoption velocityclaude-sonnet-52/5Aviation is a highly regulated, safety-critical, capital-intensive sector where AI adoption for core flight planning and evaluation functions remains slow, with pilots still in firm control of these functions despite some digitization of tools.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can usefully assist pilots in drafting initial schedules, organizing flight data, and suggesting report structures, thereby reducing clerical burden; however, the augmentation is limited to scaffolding tasks that still require significant pilot expertise and validation.
Augmentation potentialclaude-sonnet-54/5AI-powered flight planning software, weather analytics, and automated report drafting tools meaningfully speed up route planning and documentation, letting pilots and dispatchers focus on judgment-intensive decisions.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can assist in drafting flight schedules and generating routine reports from structured flight data, the task requires expert judgment about regulatory compliance, aircraft-specific considerations, and safety-critical decisions that current systems cannot reliably perform end-to-end without substantial human oversight.
Task automatabilityclaude-sonnet-52/5Drafting portions of flight plans and reports can be assisted by software, but formulating test schedules and evaluation reports requires domain judgment, regulatory awareness, and integration with real-time conditions that current AI cannot fully replicate end-to-end.dz
Adoption barriersclaude-haiku-4-5-202510014/5Commercial aviation is heavily regulated; flight evaluation reports and test schedules must be authorized and signed by appropriately rated pilots or certified flight instructors, creating a hard legal requirement that a qualified human must review and validate the output.
Adoption barriersclaude-sonnet-54/5Aviation is heavily regulated (FAA/EASA), requiring licensed pilots and certified personnel to formulate and sign off on flight plans and evaluation reports, creating strong legal and safety-driven barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI assistance for scheduling and report drafting may reduce typing effort, but the cost of integration, compliance verification, and mandatory human review likely approaches or exceeds the loaded cost of a pilot's time spent on these administrative tasks.
Cost vs. human wageclaude-sonnet-52/5Software-assisted planning tools reduce some labor cost, but the need for licensed pilot review, liability, and specialized aviation knowledge keeps the all-in cost of AI-driven automation comparable to or only modestly cheaper than skilled human labor.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product reliably performs the full scope of flight evaluation report writing and activity planning for commercial aviation; existing tools handle narrow subtasks (data logging, template generation) but lack the domain expertise and accountability required for aviation certification and regulatory sign-off.
Technical feasibility todayclaude-sonnet-52/5Flight planning software and dispatch tools exist and are widely used, but they are decision-support tools requiring pilot/dispatcher review, not autonomous systems that reliably produce final flight plans and evaluation reports without human oversight.

Choose routes, altitudes, and speeds that will provide the fastest, safest, and smoothest flights.

21

CI 14–29 · exposure 30 · augmentation 75 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Commercial aviation is a highly regulated, safety-critical sector with strong incumbent practices and slow regulatory change. AI adoption in routing remains confined to non-autonomous flight planning tools; real-time autonomous decision-making is not in production use.
Sector adoption velocityclaude-sonnet-52/5Commercial aviation adopts decision-support software steadily but is a highly regulated, safety-critical sector where autonomous replacement of pilot judgment moves slowly.
Augmentation potentialclaude-haiku-4-5-202510013/5Modern flight management systems and weather-integrated planning tools meaningfully assist pilots by pre-computing route options and alerting to wind patterns and turbulence. These augment pilot decision-making but do not replace the human judgment required for the final choice.
Augmentation potentialclaude-sonnet-55/5Modern flight management systems, weather routing tools, and fuel-optimization software substantially enhance pilots' ability to choose optimal routes while keeping the pilot in control.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can compute optimal routes and altitudes using weather data and flight physics, the task requires real-time judgment about safety tradeoffs, passenger comfort, and dynamic airspace constraints that currently demand human pilots. No AI system today performs this end-to-end with the 50% time-saving threshold at equal safety and quality standards.
Task automatabilityclaude-sonnet-52/5Flight planning software already computes optimal routes/altitudes using weather and fuel data, but final decision-making integrates real-time judgment, ATC negotiation, and safety accountability that current AI cannot fully replace end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Federal Aviation Regulations legally require a licensed, certified pilot-in-command to make in-flight routing and altitude decisions; liability for deviations and safety outcomes rests on the human pilot. This is a hard regulatory and legal barrier to autonomous automation.
Adoption barriersclaude-sonnet-55/5Aviation regulations require a licensed pilot-in-command to make and be accountable for these decisions; full automation of this authority is legally prohibited outside experimental contexts.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI route-optimization services remain niche, integration costs are non-trivial, and the oversight and liability burden mean total automation cost is not yet cheaper than the human pilot labor required for this task in commercial aviation.
Cost vs. human wageclaude-sonnet-53/5Flight planning software is already cheap relative to pilot salaries, but the task also includes real-time judgment during flight, which still requires the human pilot's presence regardless of software cost.
Technical feasibility todayclaude-haiku-4-5-202510012/5Current flight management systems assist with route optimization and autopilot execution, but no deployed product autonomously selects routes, altitudes, and speeds with the integrated safety judgment required of commercial pilots. Existing tools are narrow assistants, not independent performers of this task.
Technical feasibility todayclaude-sonnet-53/5Flight planning and dispatch systems reliably generate route/altitude recommendations in production today, but pilots still actively verify and adjust these in real time, especially for dynamic weather and turbulence.

Instruct other pilots and student pilots in aircraft operations.

11

CI 11–11 · exposure 9 · augmentation 50 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Aviation training is heavily regulated and risk-averse; adoption of AI for supplementary content or simulator design occurs slowly, and no production replacements for human instructors exist in real flight training programs.
Sector adoption velocityclaude-sonnet-52/5Aviation training is a highly regulated, safety-critical physical-world sector with slow AI adoption limited mostly to ground-school and simulator analytics tools.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist instructors by generating scenario briefs, analyzing flight data to identify learning gaps, or providing adaptive practice problems—useful support that improves instructor efficiency without removing human judgment.
Augmentation potentialclaude-sonnet-53/5AI can assist with lesson planning, knowledge quizzes, and simulator data analysis, usefully supporting instructors without replacing the hands-on teaching itself.
Task automatabilityclaude-haiku-4-5-202510011/5Instructing pilots requires real-time judgment, safety-critical decision-making in variable scenarios, and direct assessment of human performance under stress—tasks that demand live interaction and legal responsibility that current AI cannot assume end-to-end.
Task automatabilityclaude-sonnet-51/5Flight instruction requires live physical demonstration, real-time judgment of a student's control inputs, and hands-on correction in a cockpit or simulator, which current AI cannot perform end-to-end.'
Adoption barriersclaude-haiku-4-5-202510015/5Federal aviation regulations (FAA Part 61/141) mandate that flight instruction be delivered by a certificated flight instructor; legal liability for student performance and safety requires a licensed human to conduct and attest to instruction.
Adoption barriersclaude-sonnet-55/5Flight instructors must hold FAA/regulatory certification (CFI or equivalent) and legally sign off on training and competency, making this a hard licensing barrier.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI tools for simulator content or pre-flight briefings are relatively cheap, but cannot replace the instructor's presence; integration costs for scenario generation and oversight by human instructors offset modest savings.
Cost vs. human wageclaude-sonnet-52/5AI can cheaply supplement ground instruction materials, but the core in-aircraft/simulator instruction still requires a certified instructor, so overall cost savings versus a human instructor are limited.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can generate instructional content, create training materials, or simulate scenarios, no deployed system reliably performs the full instructional task of evaluating student competency and making safety-critical sign-off decisions in actual flight operations.
Technical feasibility todayclaude-sonnet-52/5AI-based ground school tutoring and simulator debrief tools exist, but no deployed product independently conducts flight instruction or check rides in production at scale.

Teach company regulations and procedures to other pilots.

11

CI 6–16 · exposure 5 · augmentation 50 · importance 3.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Aviation is a conservative, heavily regulated sector; pilot training remains dominated by human instructors with face-to-face accountability; while some airlines experiment with AI-assisted training materials, autonomous or primary AI teaching of company procedures has not achieved measurable adoption in the industry.
Sector adoption velocityclaude-sonnet-52/5Aviation training is a highly regulated, safety-critical sector with slow AI adoption for instructional authority, though e-learning tools are used for supplementary content.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist pilots preparing or delivering training by drafting materials, generating quizzes, or providing reference documents, improving preparation efficiency; however, the core teaching interaction remains human-led and judgment-driven.
Augmentation potentialclaude-sonnet-53/5AI can help create training materials, quizzes, and simulations to support instructors, improving efficiency without replacing the human teaching role.
Task automatabilityclaude-haiku-4-5-202510011/5Teaching company regulations requires explaining nuanced procedural knowledge, answering pilot questions, assessing understanding, and adapting to learner needs—complex pedagogical tasks that current AI cannot perform end-to-end with documented 50% time savings and equal quality in production settings.
Task automatabilityclaude-sonnet-51/5Teaching company-specific regulations and procedures requires live instruction, contextual judgment, and interaction grounded in operational authority that current AI cannot replicate end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Regulatory requirements and liability concerns likely mandate that a qualified, accountable human (often certificated) deliver or directly oversee safety-critical training; companies have strong incentive to retain human instructors who can be held responsible for training quality and regulatory compliance.
Adoption barriersclaude-sonnet-55/5Aviation regulations (e.g., FAA/EASA) require certified instructors or check airmen to conduct and sign off on this training, making it a hard licensing barrier.
Cost vs. human wageclaude-haiku-4-5-202510012/5Developing, customizing, and maintaining AI-driven instruction systems for company-specific regulations, plus required human oversight and verification, remains comparably costly to or more expensive than assigning an experienced pilot instructor.
Cost vs. human wageclaude-sonnet-52/5AI could supply reference materials cheaply, but the actual instructional and evaluative labor still requires a paid, qualified human instructor, keeping costs comparable or higher when factoring oversight.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI can generate training materials and answer basic procedural questions, but live teaching demands real-time interaction, credibility, correcting misconceptions, and legal accountability for safety-critical knowledge that no production system currently handles reliably without human oversight.
Technical feasibility todayclaude-sonnet-51/5No deployed product independently trains and certifies pilots on company procedures; this remains a human instructor/check-airman function.

Use instrumentation to pilot aircraft when visibility is poor.

11

CI 5–16 · exposure 17 · augmentation 88 · importance 4.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Despite decades of autopilot technology, the regulatory and safety barriers have prevented any meaningful automation of the pilot's legal responsibility. Adoption of autonomous flight remains negligible in commercial aviation.
Sector adoption velocityclaude-sonnet-52/5Aviation is a highly regulated, safety-critical sector with slow adoption cycles for new automation; existing autopilot tech evolves incrementally and full autonomy pilots are not in production despite years of R&D.
Augmentation potentialclaude-haiku-4-5-202510014/5Modern avionics, flight-management systems, and autopilot provide substantial assistance to pilots conducting instrument flight—reducing workload, improving precision, and supporting decision-making. These systems significantly enhance pilot productivity within the regulatory framework.
Augmentation potentialclaude-sonnet-55/5Modern flight management systems, autopilot, and instrument displays massively augment pilot capability in low-visibility conditions, letting pilots safely execute precision approaches with reduced workload.
Task automatabilityclaude-haiku-4-5-202510012/5While modern aircraft autopilot systems can maintain altitude, heading, and speed in instrument conditions, they cannot autonomously handle takeoff, landing, emergency management, or the full judgment required for poor-visibility flight. Current AI falls far short of the 50% time-saving bar for the complete task.
Task automatabilityclaude-sonnet-51/5Instrument flight in commercial aircraft requires certified human pilots to actively fly or supervise autopilot systems; no off-the-shelf AI system independently performs this end-to-end in commercial operations today.
Adoption barriersclaude-haiku-4-5-202510015/5Federal aviation regulations (FAA, EASA, ICAO) mandate that a licensed pilot with type-rating and current instrument certification must be physically present and in command of the aircraft. This is a hard legal requirement, not a preference.
Adoption barriersclaude-sonnet-55/5Aviation regulators (FAA, EASA) require type-rated, licensed pilots to operate and be legally responsible for instrument flight; certification, liability, and safety-of-life requirements create hard legal barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5The infrastructure cost of AI-enabled autonomous flight (certification, redundancy, liability coverage) vastly exceeds the salary of a commercial pilot, and regulatory barriers prevent cost comparison at parity.
Cost vs. human wageclaude-sonnet-51/5Regulatory mandates require licensed pilots aboard, so there is no scenario where AI alone substitutes at lower cost; redundant human staffing costs remain regardless of automation level.
Technical feasibility todayclaude-haiku-4-5-202510012/5Autopilot and flight-management systems exist in production aircraft, but they are explicitly designed as assistive tools requiring continuous human supervision and intervention, not autonomous task completion. No deployed product performs instrument flight without a licensed pilot in control.
Technical feasibility todayclaude-sonnet-52/5Autopilot and autoland systems are mature and widely deployed as aids, but fully autonomous IFR flight without a human pilot in command is not deployed in commercial aviation; it remains research/military-stage for full autonomy.

Check baggage or cargo to ensure that it has been loaded correctly.

9

CI 0–18 · exposure 13 · augmentation 25 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Aviation is a highly regulated, safety-critical sector with strong resistance to unproven automation. Adoption of AI for loading verification in commercial operations is negligible; regulatory barriers and liability concerns prevent pilots from delegating this sign-off to machines.
Sector adoption velocityclaude-sonnet-51/5Aviation is a highly regulated, physical, safety-critical sector where automation of manual inspection tasks has seen minimal real-world deployment.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist by flagging potential loading anomalies via computer vision analysis, but the pilot must still visually inspect and certify the load. The augmentation value is limited because the core task—human verification and sign-off—cannot be eliminated or substantially accelerated by current tools.
Augmentation potentialclaude-sonnet-52/5Weight-and-balance software and digital manifests assist in calculating load distribution, but the physical verification checklist itself sees limited direct AI augmentation.
Task automatabilityclaude-haiku-4-5-202510012/5Visual inspection of baggage/cargo placement requires spatial reasoning and knowledge of aircraft-specific loading rules, but current computer vision can only partially verify correct placement. Full automation would need reliable detection of load distribution, weight balance, and secure positioning—tasks where AI today shows material gaps and cannot replace human judgment on safety compliance.
Task automatabilityclaude-sonnet-51/5Physical inspection and verification of loaded baggage/cargo requires on-site visual and physical checks that current AI systems cannot perform end-to-end without robotic embodiment.
Adoption barriersclaude-haiku-4-5-202510015/5Commercial aviation is heavily regulated; FAA and international standards mandate that a licensed crew member must verify cargo loading and weight distribution. Legal liability for loading errors is substantial, and regulations explicitly require human sign-off before flight.
Adoption barriersclaude-sonnet-55/5Aviation safety regulations require certified pilots/crew to verify load and weight distribution for legal and safety compliance, creating a hard regulatory and liability barrier.
Cost vs. human wageclaude-haiku-4-5-202510012/5Deploying computer vision hardware, integration with aircraft systems, and continuous oversight by qualified personnel would cost significantly more than a pilot or ground crew member performing a brief visual check. The ROI is poor for a task taking minutes.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this physical task, so cost comparison favors the human who must be present regardless.
Technical feasibility todayclaude-haiku-4-5-202510012/5While computer vision systems exist for object detection, no deployed products reliably verify correct baggage/cargo loading in commercial aviation with the safety assurance required. Existing systems lack integration into cockpit workflows and cannot substitute for mandated human verification in regulatory contexts.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product autonomously performs physical cargo/baggage loading verification for commercial aircraft; this remains a manual pilot/ground crew responsibility.

Start engines, operate controls, and pilot airplanes to transport passengers, mail, or freight according to flight plans, regulations, and procedures.

8

CI 0–16 · exposure 13 · augmentation 75 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Commercial aviation has adopted autopilot assistance, but actual displacement of pilots toward autonomous operation is negligible and moving very slowly due to regulatory, safety, and liability constraints. Adoption of autonomous commercial flight remains in the research phase with no production deployment.
Sector adoption velocityclaude-sonnet-52/5Aviation is a highly regulated, safety-critical sector where autonomy adoption is slow and cautious, despite incremental automation (autopilot, autoland) being long-standing; full pilot replacement is not on near-term regulatory horizons.
Augmentation potentialclaude-haiku-4-5-202510014/5Modern aircraft autopilot, flight management systems, and AI-assisted monitoring tools substantially augment pilot productivity and reduce workload during cruise and some approach phases. AI can assist with weather analysis, route optimization, and anomaly detection while pilots retain final authority and decision-making.
Augmentation potentialclaude-sonnet-54/5Flight management systems, autopilot, and decision-support tools significantly reduce pilot workload and improve safety margins, meaningfully augmenting the human pilot throughout the flight.
Task automatabilityclaude-haiku-4-5-202510012/5While modern aircraft have autopilot systems that can handle cruise flight and some approach phases, full end-to-end autonomous flight remains undeployed in commercial settings. Current AI cannot reliably handle emergency decision-making, variable weather conditions, air traffic control coordination, or the full spectrum of preflight checks and takeoff/landing procedures required by regulation.
Task automatabilityclaude-sonnet-51/5While autopilot systems handle much of cruise flight, actual piloting including takeoff, landing, emergency response, and legal command authority requires a certified human pilot; no system today performs this end-to-end without a human at controls.
Adoption barriersclaude-haiku-4-5-202510015/5Extremely strong regulatory barriers: FAA, EASA, and international aviation authorities mandate a licensed pilot in command for all commercial flights, with specific medical and training certifications. Liability and safety-critical requirements are among the highest in any profession, and full autonomous operation faces near-insurmountable legal and insurance obstacles.
Adoption barriersclaude-sonnet-55/5Aviation regulations (FAA, ICAO) mandate certified, licensed pilots physically operating commercial aircraft, with strict liability, safety certification, and passenger-safety requirements making full automation legally prohibited today.
Cost vs. human wageclaude-haiku-4-5-202510011/5The infrastructure cost of developing, certifying, and maintaining autonomous flight systems vastly exceeds the cost of pilot labor. Current autopilot systems are expensive add-ons that require pilot oversight, not replacements for the pilot's salary.
Cost vs. human wageclaude-sonnet-51/5Fully autonomous commercial passenger flight does not exist; the cost comparison is moot since human pilots remain mandatory, and certifying autonomous systems would require massive additional investment.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed commercial product autonomously pilots passenger or freight aircraft end-to-end today. Experimental autonomous flight systems exist in research and limited testing, but are not in production use by commercial operators, and regulatory certification for fully autonomous commercial flight does not exist.
Technical feasibility todayclaude-sonnet-52/5Autopilot and autoland systems are mature and widely deployed as aids, but no product independently starts engines, taxis, and pilots a full commercial flight without licensed pilots present and in command.

Check aircraft prior to flights to ensure that the engines, controls, instruments, and other systems are functioning properly.

6

CI 0–11 · exposure 8 · augmentation 38 · importance 5.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Regulatory requirements and safety criticality mean this task is performed by certified humans across the entire aviation sector with no meaningful displacement by AI systems in production.
Sector adoption velocityclaude-sonnet-52/5Aviation is a highly regulated, safety-critical sector with slow adoption of autonomous systems for core flight safety tasks, though sensor diagnostics are increasingly used as aids.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist marginally through automated checklist management or anomaly flagging from sensor data, but current systems provide limited augmentation since human judgment and physical inspection remain central to the task's safety role.
Augmentation potentialclaude-sonnet-53/5AI-enabled diagnostic systems and digital checklists can flag anomalies and streamline data review, meaningfully assisting pilots without replacing the physical inspection process.
Task automatabilityclaude-haiku-4-5-202510012/5Pre-flight checks require physical inspection of dispersed aircraft components and tactile assessment (e.g., control surface movement, fuel cap security) that current AI systems cannot perform without robotics. While AI could theoretically assist with documentation or checklist verification, the majority of the safety-critical manual verification work remains beyond current automation.
Task automatabilityclaude-sonnet-51/5Preflight inspection requires physical presence, visual and tactile checks of the aircraft exterior, and manual verification of systems that current AI cannot perform end-to-end without robotic embodiment.
Adoption barriersclaude-haiku-4-5-202510015/5Federal Aviation Regulations (FAR) require a licensed pilot to personally conduct and sign off on pre-flight inspections before each flight; this legal mandate creates an absolute barrier to automation regardless of technical capability.
Adoption barriersclaude-sonnet-55/5Aviation regulations (e.g., FAA) mandate that a certificated pilot personally conduct and certify preflight checks, making this a hard legal requirement for human performance.
Cost vs. human wageclaude-haiku-4-5-202510011/5The specialized equipment and robotics required to perform physical aircraft inspection would be far more expensive than the cost of a licensed pilot conducting the check, making full automation economically infeasible today.
Cost vs. human wageclaude-sonnet-51/5Since no AI system can substitute for the physical inspection, there is no viable AI-only cost comparison—human labor remains required.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI system reliably performs the physical inspection and verification components of aircraft pre-flight checks in production environments. Computer vision for some visual inspections exists in research, but comprehensive pre-flight validation remains the domain of human pilots.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs full physical preflight inspections autonomously; some digital checklist and sensor-monitoring tools assist but do not replace the physical walk-around and manual checks.

Contact control towers for takeoff clearances, arrival instructions, and other information, using radio equipment.

4

CI 0–9 · exposure 8 · augmentation 25 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Zero adoption is occurring because the task is fundamentally prohibited by aviation law and regulation. There are no pilots, operators, or institutions attempting to automate ATC communication in commercial aviation.
Sector adoption velocityclaude-sonnet-51/5Aviation is a highly regulated, safety-critical sector with extremely slow AI adoption for flight-critical operations like ATC communication.
Augmentation potentialclaude-haiku-4-5-202510011/5Augmentation potential is minimal; the task involves direct regulatory-mandated human-ATC interaction. While AI could theoretically assist with transcription or data processing, the core task is already lean and requires immediate pilot presence and response.
Augmentation potentialclaude-sonnet-53/5AI-based speech-to-text, readback verification, and workload-reduction tools can assist pilots in parsing and confirming ATC instructions, but human pilots remain fully responsible for communication and decisions.
Task automatabilityclaude-haiku-4-5-202510012/5While AI could theoretically generate clearance requests, commercial aviation requires real-time human judgment, context-dependent decisions, and regulatory compliance. Current AI systems cannot autonomously handle the dynamic, safety-critical two-way communication with ATC required for legal flight operations.
Task automatabilityclaude-sonnet-51/5Real-time radio communication with ATC requires split-second situational judgment, safety-critical decision-making, and physical control of the aircraft; no off-the-shelf system performs this end-to-end today.
Adoption barriersclaude-haiku-4-5-202510015/5Extremely high barriers: commercial pilots are required by law (FAA Part 121 regulations) to maintain two-way radio communication with ATC. Liability for communication failures is severe, and no automation of this safety-critical function is permitted in current regulations.
Adoption barriersclaude-sonnet-55/5Aviation regulations (FAA/ICAO) require certified, licensed pilots to communicate with ATC and make real-time flight decisions; this is a hard legal and safety barrier.
Cost vs. human wageclaude-haiku-4-5-202510011/5Automating this task is not economically relevant because it is legally and operationally prohibited. The cost comparison is moot when the task cannot be performed by AI systems under any regulatory framework.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this function alone, so cost comparison favors the human by default; certification and safety costs for an AI alternative would be immense.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably performs this task end-to-end. Radio communication with control towers in commercial aviation remains entirely human-piloted; automation is not permitted under current FAA and international aviation regulations.
Technical feasibility todayclaude-sonnet-51/5No deployed commercial product has pilots' radio ATC communication duties automated; voice recognition for ATC is only research-stage/assistive (e.g., readback checking tools), not operational replacement.

Coordinate flight activities with ground crews and air traffic control, and inform crew members of flight and test procedures.

3

CI 0–5 · exposure 5 · augmentation 38 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Commercial aviation is heavily regulated and conservative in adoption of unproven automation for safety-critical crew coordination. The industry shows no trajectory toward replacing pilot-led coordination with autonomous AI systems; instead, research remains limited to narrow decision-support tools.
Sector adoption velocityclaude-sonnet-51/5Commercial aviation is a highly regulated, safety-critical physical-world sector with very slow AI adoption for actual flight-control and coordination functions.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could provide limited assistance (e.g., pre-populated checklists, ATC message summarization), but current systems offer minimal productivity gains for an experienced pilot already proficient in these standardized procedures. The task is communication-heavy and cannot be meaningfully augmented without introducing unacceptable latency or information loss.
Augmentation potentialclaude-sonnet-53/5AI-based tools like automated ATC transcription, digital checklists, and flight-planning software assist pilots with situational awareness and communication logging, but core coordination remains human-driven.
Task automatabilityclaude-haiku-4-5-202510011/5This task fundamentally requires real-time human judgment, interpersonal communication, and decision-making in dynamic conditions with safety-critical consequences. No current AI system can autonomously coordinate with ground crews and air traffic control while managing crew communications and procedures—human pilots must legally and operationally remain in command.
Task automatabilityclaude-sonnet-51/5Real-time flight coordination requires live sensory awareness, split-second judgment, and physical presence in the cockpit; no current AI system can perform this end-to-end task autonomously.
Adoption barriersclaude-haiku-4-5-202510015/5Extremely high barriers: Federal Aviation Regulations (FAA, EASA, etc.) legally mandate that licensed pilots conduct coordination with ATC and brief crews. Aviation liability, crew safety accountability, and regulatory certification create insurmountable legal and regulatory barriers to autonomous AI substitution.
Adoption barriersclaude-sonnet-55/5Aviation regulations (FAA/EASA) mandate licensed pilots for flight operations and communications with ATC, plus extreme liability and safety-critical certification requirements create hard legal barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5The loaded cost of a commercial pilot far exceeds current AI integration costs, but automation is not the relevant comparison here; the task requires active human presence and accountability. Attempting to replace this role with AI would actually increase costs due to liability, oversight, and certification requirements.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this task standalone, so no meaningful cost comparison favors AI; the human pilot remains mandatory infrastructure.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can draft communications or log procedures, no deployed system reliably performs the full coordination task independently. Existing products assist with procedural documentation or communication templates, but the core function—real-time negotiation with ATC, crew briefings, and on-the-fly procedure adjustments—remains entirely human-operated in production aviation.
Technical feasibility todayclaude-sonnet-51/5No deployed product independently coordinates with ATC and ground crews and briefs human crew members in real commercial operations; this remains firmly in the human pilot's role.

Order changes in fuel supplies, loads, routes, or schedules to ensure safety of flights.

1

CI 0–3 · exposure 0 · augmentation 50 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Aviation remains a highly regulated, safety-critical domain with minimal autonomous decision-making adoption. Pilots retain direct legal authority and responsibility for all operational changes.
Sector adoption velocityclaude-sonnet-52/5Aviation is a highly regulated, safety-critical physical sector with slow, cautious AI adoption limited mostly to advisory and planning tools, not decision authority transfer.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist by suggesting fuel calculations or alternative routes for pilot review, but the task's nature—making binding safety-critical orders—limits meaningful augmentation since human authority cannot be delegated.
Augmentation potentialclaude-sonnet-54/5AI-based flight planning, weather, and fuel optimization tools already meaningfully assist pilots and dispatchers in making these decisions faster and with better data, though humans retain final authority.
Task automatabilityclaude-haiku-4-5-202510011/5Ordering fuel, load, route, or schedule changes requires real-time human judgment about safety tradeoffs, regulatory compliance, and operational context that AI cannot currently perform end-to-end. This task demands authority to make binding operational decisions that remain strictly human.
Task automatabilityclaude-sonnet-51/5This requires real-time judgment, legal authority, and accountability for passenger safety in dynamic physical conditions; no current AI system can independently order such operational changes end-to-end.'
Adoption barriersclaude-haiku-4-5-202510015/5FAA regulations explicitly require licensed pilots to exercise command authority and make safety-critical operational decisions; no automation can legally substitute for this legal requirement.
Adoption barriersclaude-sonnet-55/5Aviation regulations (FAA/ICAO) require a licensed, certified pilot-in-command to make safety-critical operational decisions; this is a hard legal and liability barrier.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of an AI error in ordering fuel or schedule changes on a commercial flight is catastrophic liability and safety failure, making any AI-only solution prohibitively expensive relative to a human pilot's loaded wage.
Cost vs. human wageclaude-sonnet-51/5Since AI cannot legally or reliably perform this task independently, there is no valid substitute cost comparison—human pilots remain mandatory, making AI cost irrelevant or infinite for full substitution.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably makes autonomous decisions to order operational changes for commercial flights. This authority must remain with licensed pilots and dispatch personnel in production.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously issues binding flight safety decisions like fuel/load/route changes; decision-support tools exist but a human pilot always makes and owns the call.

Co-pilot aircraft or perform captain's duties, as required.

0

CI 0–0 · exposure 0 · augmentation 50 · importance 4.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Aviation is a laggard sector for autonomous systems adoption due to safety criticality, regulatory conservatism, and high liability costs. No airlines operate commercial routes with fully autonomous cockpits.
Sector adoption velocityclaude-sonnet-51/5Commercial aviation is a highly regulated, safety-critical physical sector with extremely slow adoption of autonomy in pilot roles; current AI use is limited to decision-support tools, not role replacement.
Augmentation potentialclaude-haiku-4-5-202510013/5Modern flight management systems, autopilot, and AI-assisted navigation and weather analysis do assist pilots with workload and decision-making, but augmentation is limited to specific phases and routine conditions rather than transformative across all captain or co-pilot duties.
Augmentation potentialclaude-sonnet-53/5AI-enhanced flight management systems, predictive maintenance alerts, and automated checklists assist pilots in workload reduction and situational awareness, but do not fundamentally transform the core piloting task.
Task automatabilityclaude-haiku-4-5-202510011/5Commercial aviation remains heavily dependent on human pilots for real-time decision-making, emergency response, and legal responsibility. Current AI cannot reliably handle the full spectrum of flight scenarios, weather changes, and contingencies that define co-pilot or captain duties.
Task automatabilityclaude-sonnet-51/5Flying as pilot-in-command or co-pilot involves real-time perception, judgment under uncertainty, emergency response, and legal accountability that no current AI system can perform end-to-end without a human physically present and in control.
Adoption barriersclaude-haiku-4-5-202510015/5Commercial aviation is among the most heavily regulated industries globally. Federal Aviation Administration (FAA) and international aviation authority regulations legally require licensed human pilots to operate aircraft; no automation can bypass this legal requirement.
Adoption barriersclaude-sonnet-55/5Aviation regulations (FAA, ICAO) legally require licensed, certified human pilots to occupy cockpit seats and make command decisions, with strict liability and safety certification regimes forming a hard barrier.
Cost vs. human wageclaude-haiku-4-5-202510011/5The infrastructure, certification, and liability costs of autonomous flight systems, combined with ongoing regulatory uncertainty, far exceed the cost of trained human pilots. Integration and oversight would be prohibitively expensive.
Cost vs. human wageclaude-sonnet-51/5There is no AI system that can substitute for the certified human role, so no meaningful cost comparison exists; any automation would require costly redundant hardware, certification, and oversight far exceeding pilot wages at this stage.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed commercial aviation system today operates aircraft end-to-end without human pilots in the cockpit. Autopilot and flight management systems assist with specific phases but are not autonomous replacements for pilot judgment and control.
Technical feasibility todayclaude-sonnet-51/5While autopilot and flight management systems handle segments of flight, no deployed product replaces a human co-pilot or captain's full role; fully autonomous commercial passenger flight is research-stage, not in production.

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

CI 0–0 · exposure 0 · augmentation 38 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Aviation is a highly regulated, safety-first sector with entrenched human-pilot requirements and slow adoption of autonomous systems; in-flight test and evaluation remains firmly in the domain of certified human pilots.
Sector adoption velocityclaude-sonnet-51/5Aviation flight-test operations are a highly regulated, physical, low-digitization niche with minimal AI agent deployment in production for this specific task.
Augmentation potentialclaude-haiku-4-5-202510012/5While AI could assist with data logging, real-time sensor analysis, or trend detection during flight, the core task of conducting tests at specified altitudes and evaluating equipment characteristics in variable weather remains pilot-centric with limited opportunity for meaningful augmentation.
Augmentation potentialclaude-sonnet-53/5AI-assisted data analysis, sensor monitoring, and post-flight diagnostics can help pilots and engineers interpret test results, though the in-flight testing itself remains human-performed.
Task automatabilityclaude-haiku-4-5-202510011/5In-flight testing requires real-time aircraft control, safety-critical decision-making under variable weather, and physical equipment interaction—tasks that demand human judgment and cannot be safely delegated to current AI systems end-to-end. No commercial product automates the full scope of in-flight evaluation at specified conditions.
Task automatabilityclaude-sonnet-51/5This requires physically piloting an aircraft, operating test equipment, and making real-time judgments under varying flight conditions; no current AI system can perform this end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Regulation (FAA, EASA, ICAO) mandates that in-flight testing and aircraft operation must be conducted by licensed pilots; liability and safety-critical error costs create hard legal barriers to automation or unsupervised AI operation.
Adoption barriersclaude-sonnet-55/5Flight testing requires licensed pilots, regulatory certification (FAA/EASA), and legal accountability for airworthiness determinations, making human authorization mandatory.
Cost vs. human wageclaude-haiku-4-5-202510011/5A qualified commercial pilot conducting in-flight evaluations commands a fully-loaded cost (salary, certification, insurance, liability) far lower than the infrastructure, AI training, redundancy, and safety certification required to perform these tests autonomously.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this task, so cost comparison favors the human pilot entirely; any hypothetical automation would require extremely expensive autonomous flight and sensor infrastructure.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI system reliably conducts independent in-flight tests and evaluations; this remains strictly within human pilot domain in production aviation. Research into autonomous flight exists, but no product performs the full evaluation task safely or at scale.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product conducts in-flight test evaluations of aircraft systems; this remains a specialized human flight-test function requiring physical presence and certified expertise.

Rescue and evacuate injured persons.

0

CI 0–0 · exposure 0 · augmentation 25 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Rescue and evacuation remains entirely human-dependent in commercial aviation. No sector data shows any automation or AI adoption in this emergency response task; regulatory and safety requirements maintain strict human control.
Sector adoption velocityclaude-sonnet-51/5Aviation emergency response and physical rescue functions show essentially no AI adoption; this is a physical, safety-critical task with minimal digitization potential.
Augmentation potentialclaude-haiku-4-5-202510012/5AI might provide limited assistance through real-time medical diagnostics displays or evacuation route optimization, but the core task of physically rescuing and evacuating injured persons must remain human-performed, limiting meaningful augmentation.
Augmentation potentialclaude-sonnet-52/5AI could assist with emergency protocols, communication, or triage guidance, but offers minimal direct augmentation to the physical act of rescue and evacuation.
Task automatabilityclaude-haiku-4-5-202510011/5Rescuing and evacuating injured persons requires physical manipulation of human bodies, aircraft control during dynamic emergencies, real-time medical assessment, and split-second decision-making in hazardous conditions. Current AI systems cannot physically intervene or safely operate aircraft during emergency scenarios requiring human judgment and motor skills.
Task automatabilityclaude-sonnet-51/5Physically rescuing and evacuating injured persons requires embodied physical action, judgment under chaotic conditions, and manual handling that no current AI system can perform.
Adoption barriersclaude-haiku-4-5-202510015/5Federal aviation regulations (FAA/ICAO) legally require a licensed, qualified pilot to operate aircraft and manage emergency procedures including evacuations. Additionally, patient safety during medical emergencies and evacuation creates strict liability requirements that mandate human professional judgment and accountability.
Adoption barriersclaude-sonnet-55/5Physical rescue involves life-safety responsibility, legal duty of care, and requires human physical presence and authority, creating hard barriers to any non-human substitution.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI systems cannot perform this task at all, making cost comparison moot; the human commercial pilot remains irreplaceable. Any attempt to substitute AI for rescue operations would be catastrophically expensive in lives and liability.
Cost vs. human wageclaude-sonnet-51/5There is no AI-based alternative performing this task, so any comparison favors the human entirely; AI cost is effectively infinite/non-applicable.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product can autonomously perform rescue and evacuation of injured persons. This task inherently requires physical presence, human-to-human interaction, medical judgment, and direct pilot control—capabilities that do not exist in production systems today.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product performs physical rescue/evacuation of injured people; this remains a purely human, physical-world task with no robotic substitute in production.

Perform minor aircraft maintenance and repair work, or arrange for major maintenance.

0

CI 0–0 · exposure 0 · augmentation 25 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Aviation maintenance is a heavily regulated, safety-critical domain with minimal AI adoption; no meaningful trend toward AI-driven maintenance or repair exists in commercial aviation today. Regulatory oversight and the high cost of failure maintain human dominance.
Sector adoption velocityclaude-sonnet-51/5Aviation maintenance is a highly regulated, physical, safety-critical domain with minimal AI adoption for hands-on repair work; digitization here lags far behind information-sector norms.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could potentially assist with maintenance scheduling or diagnostic checklist generation, but augmentation is minimal because pilots and mechanics rely on established protocols, physical inspection, and hands-on judgment that AI cannot meaningfully enhance in current systems.
Augmentation potentialclaude-sonnet-52/5AI can assist with maintenance scheduling, diagnostics from sensor data, or documentation, but offers little help with the hands-on repair and arrangement judgment calls themselves.
Task automatabilityclaude-haiku-4-5-202510011/5Aircraft maintenance requires hands-on mechanical work, safety-critical decision-making, and FAA compliance that cannot be performed remotely or autonomously by current AI systems. This task involves physical inspection, diagnostics, and repair that demand human expertise and cannot meet the 50% time-saving threshold end-to-end.
Task automatabilityclaude-sonnet-51/5Physical inspection, hands-on minor repairs, and coordinating maintenance logistics require physical manipulation and judgment that current AI cannot perform; no end-to-end automation is feasible today.
Adoption barriersclaude-haiku-4-5-202510015/5Federal Aviation Regulations (FAR Part 43) legally require that aircraft maintenance be performed or supervised by certified mechanics or pilots; liability for safety-critical failures is borne by the human responsible. This creates an absolute legal barrier to AI substitution.
Adoption barriersclaude-sonnet-55/5Aviation maintenance is heavily regulated (FAA/EASA), requiring licensed A&P mechanics or certified pilots to sign off on airworthiness, creating hard legal barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI systems have no role in reducing the cost of aircraft maintenance, which is already labor-intensive and highly specialized. The human expert (pilot or mechanic) remains the sole economic actor, making AI cost comparison irrelevant.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing the physical labor, so any comparison favors the human worker entirely; robotics for this niche task doesn't exist commercially.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product reliably performs or arranges aircraft maintenance independently; this remains entirely dependent on human pilots and certified mechanics. The regulatory and safety-critical nature of aviation maintenance means AI has not reached production deployment for this task.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product performs physical aircraft maintenance tasks; this remains firmly in the domain of certified human mechanics and pilots performing preflight checks.

Supervise other crew members.

0

CI 0–0 · exposure 0 · augmentation 25 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5There is zero adoption of AI for crew supervision in aviation, and regulatory structure makes adoption essentially impossible. The aviation sector is highly regulated, and crew oversight remains exclusively human-assigned by law.
Sector adoption velocityclaude-sonnet-51/5Commercial aviation is heavily regulated and safety-conservative, with crew supervision authority showing no movement toward AI adoption.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could potentially assist with monitoring crew fatigue data or performance metrics to support a human supervisor's judgment, but direct supervision remains a human function. The task is fundamentally about human authority and accountability, limiting meaningful augmentation.
Augmentation potentialclaude-sonnet-52/5AI can support crew coordination via checklists, scheduling, or communication tools, but offers minimal direct enhancement to the interpersonal and command aspects of supervision.
Task automatabilityclaude-haiku-4-5-202510011/5Supervising crew members requires real-time decision-making, situational awareness, interpersonal judgment, and authority recognition that current AI systems cannot replicate. This task involves personnel management, behavioral assessment, and safety-critical oversight that fundamentally depends on human judgment and accountability.
Task automatabilityclaude-sonnet-51/5Supervising crew requires real-time human judgment, authority, and accountability in a physical, safety-critical environment that current AI cannot execute end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Aviation regulations explicitly require human captain or first officer authority over crew supervision; this is a licensed, legally mandated function. FAA and international aviation authorities assign supervision and authority exclusively to qualified flight crew, creating hard regulatory barriers to any AI substitution.
Adoption barriersclaude-sonnet-55/5Aviation regulations legally require a certificated pilot-in-command to supervise crew and bear responsibility for safety, making this a hard licensing and liability barrier.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI systems cannot perform this task, making cost comparison moot. The pilot performing supervision is already on the flight deck; there is no cost advantage to any automation or AI system.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing this supervisory role, so no cost comparison favors AI; human captain authority is irreplaceable at any price point today.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product can independently supervise airline crew; this requires legal authority vested in a licensed human pilot. Current AI lacks the contextual judgment, accountability, and decision-making autonomy needed for crew supervision in safety-critical aviation environments.
Technical feasibility todayclaude-sonnet-51/5No deployed product supervises human flight crew members; this remains firmly a human command function with no automation analog in production.

Fly with other pilots or pilot-license applicants to evaluate their proficiency.

0

CI 0–0 · exposure 0 · augmentation 38 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Aviation is a heavily regulated, laggard sector for autonomous decision-making in safety-critical roles. Adoption of AI for pilot evaluation remains non-existent; regulatory and safety culture strongly favor human operators.
Sector adoption velocityclaude-sonnet-51/5Aviation training and certification is a highly regulated, safety-critical, low-digitization-of-judgment sector with essentially no movement toward AI-led proficiency checks.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist with data logging and post-flight analysis of performance metrics, but the core task of in-flight evaluation of a pilot's judgment, handling, and emergency response remains fundamentally human-centric, limiting meaningful augmentation.
Augmentation potentialclaude-sonnet-53/5AI-based flight simulators, data analytics on flight performance, and automated debriefing tools can assist evaluators in reviewing metrics, but the core evaluative judgment remains human-led.
Task automatabilityclaude-haiku-4-5-202510011/5Flying aircraft requires real-time physical control in a safety-critical environment where human judgment, emergency response, and split-second decisions are irreplaceable. Current AI cannot perform this end-to-end task or achieve meaningful time savings at equal quality.
Task automatabilityclaude-sonnet-51/5Evaluating another pilot's proficiency requires real-time judgment, in-cockpit presence, physical control handoff, and legal certification authority that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5This task has hard legal and regulatory barriers: federal aviation regulations require that an appropriately licensed and qualified human pilot or examiner be physically present and in control or ready to intervene. Insurance, liability, and certification standards mandate human responsibility.
Adoption barriersclaude-sonnet-55/5Check airman/examiner roles require FAA (or equivalent) certification and legal authority to sign off on pilot competency, an unambiguous licensed-human-required barrier.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI systems capable of flight are prohibitively expensive and still require human pilots on board; the all-in cost per flight hour far exceeds the loaded wage of a commercial pilot performing this task.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this task, so cost comparison favors the human by default; any AI system would add cost without replacing the core function.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product can reliably fly an aircraft with a trainee or evaluator on board in production settings. While autopilot systems exist, they do not perform the full task of evaluating another pilot's proficiency, which requires subjective assessment and real-time intervention.
Technical feasibility todayclaude-sonnet-51/5No deployed product conducts in-flight check rides or proficiency evaluations of human pilots; this remains firmly human and research-stage at best for any AI co-evaluator role.

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.