Air Traffic Controllers

53-2021.00
Median wage $148,080/yr22,510 employed (US)Rank #867 of 923 scored · top 94% by substitution

Control air traffic on and within vicinity of airport, and movement of air traffic between altitude sectors and control centers, according to established procedures and policies. Authorize, regulate, and control commercial airline flights according to government or company regulations to expedite and ensure flight safety.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution11
Exposure15
Augmentation47

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

23 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

0%

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%14

panel mean rating 1.6/5 → substitution pressure 14/100

Technical feasibility todayw 20%15

panel mean rating 1.6/5 → substitution pressure 15/100

Cost vs. human wagew 15%12

panel mean rating 1.5/5 → substitution pressure 12/100

Adoption barriersw 20%inverted — strong barriers lower the score2

panel mean rating 4.9/5 (barrier strength) → substitution pressure 2/100

Sector adoption velocityw 10%5

panel mean rating 1.2/5 → substitution pressure 5/100

Task breakdown (23 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.

Compile information about flights from flight plans, pilot reports, radar, or observations.

49

CI 18–80 · exposure 62 · augmentation 75 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510015/5Automated flight data compilation and display systems have been standard infrastructure in ATC facilities globally for decades, with continuous upgrades (NextGen, SESAR) embedding deeper automation into operational workflows.
Sector adoption velocityclaude-sonnet-51/5Aviation safety systems are notoriously slow to adopt new automation due to certification cycles, regulatory approval, and conservative safety culture, with AI-driven changes to core ATC functions essentially absent in production.
Augmentation potentialclaude-haiku-4-5-202510015/5Modern ATC displays, conflict detection algorithms, and data fusion tools substantially enhance controller productivity by presenting synthesized, real-time flight information that humans use to make faster, more informed separation and sequencing decisions.
Augmentation potentialclaude-sonnet-53/5AI-based decision support tools (e.g., conflict alert systems, data fusion displays) already help controllers synthesize radar and flight plan data, improving situational awareness while humans retain full authority.
Task automatabilityclaude-haiku-4-5-202510015/5Modern ADS-B, radar feeds, and flight plan databases provide structured data that can be automatically parsed, correlated, and compiled into integrated flight information displays with minimal human intervention, easily meeting the 50% time-saving threshold.
Task automatabilityclaude-sonnet-52/5Aggregating flight plans, pilot reports, radar, and observations into a coherent flight picture requires real-time, safety-critical fusion under strict tolerances; current AI can assist but not autonomously replace this end-to-end today.itto
Adoption barriersclaude-haiku-4-5-202510015/5FAA and international aviation regulations (14 CFR Part 65, ICAO standards) mandate that licensed air traffic controllers must maintain situational awareness and approve critical air traffic decisions; automated compilation does not eliminate the human sign-off requirement for tactical decisions.
Adoption barriersclaude-sonnet-55/5Air traffic control is a highly regulated, licensed profession with strict FAA/ICAO certification requirements and severe liability for errors, mandating human control over flight information compilation.
Cost vs. human wageclaude-haiku-4-5-202510015/5The infrastructure for automated flight data compilation (radar, ADS-B receivers, data fusion servers) operates continuously at marginal cost per additional flight once established, orders of magnitude cheaper than human compilation labor.
Cost vs. human wageclaude-sonnet-52/5Building and certifying an AI system to safety standards for this task requires expensive redundant infrastructure, validation, and human backup, making it not clearly cheaper than existing controller labor plus current automation.
Technical feasibility todayclaude-haiku-4-5-202510015/5Automated flight information systems (radar data integration, flight strips, traffic management systems) are already deployed and reliably perform real-time compilation of flight data in operational environments at scale across the world.
Technical feasibility todayclaude-sonnet-52/5Some decision-support tools exist that fuse radar and flight data for controllers, but no deployed product independently compiles this information without human oversight in live ATC operations.

Complete daily activity reports and keep records of messages from aircraft.

40

CI 25–55 · exposure 50 · augmentation 63 · importance 3.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5ATC is a heavily regulated, safety-critical domain with slow institutional change and strong union resistance to automation. Adoption of even supplementary logging AI is cautious and piecemeal, with many facilities still relying on manual methods.
Sector adoption velocityclaude-sonnet-52/5Aviation is a highly regulated, safety-critical sector with slow technology adoption cycles, especially for tools touching official records rather than administrative back-office functions.
Augmentation potentialclaude-haiku-4-5-202510013/5AI-assisted transcription and message parsing can reduce clerical burden and help organize data for review, but the controller must still perform final judgment and certification. Moderate productivity gain is possible through better data organization and search, but the core verification task remains human-driven.
Augmentation potentialclaude-sonnet-54/5AI-assisted transcription and auto-population of reports from radio and radar logs can substantially speed up documentation while controllers retain oversight and final sign-off.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can transcribe and parse aircraft messages with high accuracy, the task requires domain judgment about what constitutes reportable activity, context-dependent filtering, and integration with live operational data. Current systems cannot reliably decide what merits inclusion in official records without substantial human oversight, falling well short of the 50% time-saving threshold.
Task automatabilityclaude-sonnet-54/5Logging messages and compiling structured daily activity reports is a routine documentation task that current NLP/transcription and record-keeping systems can largely automate, especially when integrated with radio/communications logs.
Adoption barriersclaude-haiku-4-5-202510015/5FAA regulations require licensed air traffic controllers to author and take responsibility for official records; liability for inaccurate reporting falls on the controller, creating a hard legal barrier to full automation. No automation can replace the controller's signature and accountability on these records.
Adoption barriersclaude-sonnet-54/5Air traffic systems are heavily regulated, and record-keeping tied to safety-critical operations typically requires certified systems and human verification, creating significant regulatory and liability barriers.
Cost vs. human wageclaude-haiku-4-5-202510012/5Transcription and logging tools are cheap, but the human oversight required to validate, interpret, and sign off on safety-critical records remains necessary and costly. Integration with legacy ATC systems and mandatory quality checks keep total cost per task comparable to or exceeding human labor.
Cost vs. human wageclaude-sonnet-54/5Automated logging/transcription software is inexpensive to run continuously compared to controller time spent on manual documentation, though integration with certified systems adds some cost.
Technical feasibility todayclaude-haiku-4-5-202510013/5Speech-to-text and message logging systems exist in some ATC facilities, but deployed products typically handle transcription only; human controllers still filter, interpret, and officially author the reports. No mature, end-to-end automated reporting system is known to operate reliably across diverse ATC facilities.
Technical feasibility todayclaude-sonnet-53/5Aviation communication logging and transcription tools exist and are used in some ATC facilities, but full automated report generation integrated into FAA/ANSP workflows is not yet universally deployed at scale.

Review records or reports for clarity and completeness and maintain records or reports, as required under federal law.

39

CI 23–55 · exposure 45 · augmentation 63 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Air traffic is a heavily regulated, safety-critical sector where administrative automation adoption is cautious; controllers and facilities have shown minimal uptake of AI for compliance-sensitive record-keeping tasks, and regulatory culture discourages autonomous systems in this domain.
Sector adoption velocityclaude-sonnet-52/5Air traffic control is a highly regulated, safety-critical, and slow-to-digitize government-adjacent sector with cautious technology adoption cycles.
Augmentation potentialclaude-haiku-4-5-202510013/5AI tools can assist controllers by automating data formatting, highlighting missing fields, or flagging inconsistencies against regulatory checklists, meaningfully reducing review time while the controller retains final authority over compliance decisions.
Augmentation potentialclaude-sonnet-54/5AI tools can effectively flag inconsistencies, missing fields, or errors in records, significantly speeding up the human review process while keeping a person accountable for final sign-off.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can assist in parsing and flagging incomplete or unclear entries in structured records, the task requires judgment about regulatory compliance nuance and contextual interpretation that current systems cannot reliably automate end-to-end without substantial human review, especially given the safety-critical nature of aviation records.
Task automatabilityclaude-sonnet-54/5Reviewing records and reports for clarity/completeness and maintaining logs is largely a text-processing task well suited to current LLMs and document automation tools, though air traffic-specific formats and regulatory nuances require setup and validation.
Adoption barriersclaude-haiku-4-5-202510014/5Federal regulations (14 CFR) explicitly mandate that certain records be maintained and reviewed by licensed air traffic controllers; liability exposure for automation errors in safety-critical documentation is severe, and regulatory framework assumes human accountability for record integrity.
Adoption barriersclaude-sonnet-54/5Federal law mandates specific record-keeping and reporting standards in aviation safety, and controllers or certified staff are typically accountable for accuracy, creating regulatory and liability barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Integration, fine-tuning for FAA rule specifics, and mandatory human oversight would likely cost as much or more than the labor-loaded wage for a controller performing this administrative task, especially at the error-detection accuracy required.
Cost vs. human wageclaude-sonnet-54/5Automated document parsing and compliance-checking software is far cheaper per unit of review than controller or administrative staff time, once integrated with existing systems.
Technical feasibility todayclaude-haiku-4-5-202510012/5Document review products and data validation tools exist, but no deployed system reliably performs full regulatory compliance audits for FAA-mandated air traffic records without expert human oversight; the liability and error consequences are too high for fully autonomous execution in production.
Technical feasibility todayclaude-sonnet-53/5Document review and record-keeping automation products exist broadly in other regulated industries, but specialized deployment for FAA-mandated ATC records is not widely evidenced in production today.

Conduct pre-flight briefings on weather conditions, suggested routes, altitudes, indications of turbulence, or other flight safety information.

21

CI 18–24 · exposure 25 · augmentation 75 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Adoption of autonomous briefing systems is virtually non-existent because regulatory barriers and safety-critical requirements prevent deployment. Pilot programs and research exist, but real-world substitution of human controllers remains extremely limited.
Sector adoption velocityclaude-sonnet-52/5Aviation is a highly regulated, safety-critical sector with slow, cautious AI adoption; automation is limited to decision-support tools, not autonomous task replacement.
Augmentation potentialclaude-haiku-4-5-202510014/5AI systems demonstrate strong augmentation potential: real-time weather aggregation, traffic pattern analysis, turbulence prediction, and route optimization can substantially assist human controllers in preparing more comprehensive and timely briefings while the controller retains decision authority.
Augmentation potentialclaude-sonnet-54/5AI-driven weather aggregation, turbulence prediction, and route optimization tools meaningfully speed up and improve the quality of briefings prepared by controllers/dispatchers.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can aggregate and summarize weather data and flight information rapidly, air traffic control briefings require real-time judgment, legal sign-off, and coordination with human pilots that current systems cannot fully perform end-to-end. The task involves discretionary safety decisions and accountability that remain outside reliable automation scope today.
Task automatabilityclaude-sonnet-52/5AI can compile and summarize weather and route data, but delivering authoritative pre-flight briefings requires real-time judgment, liability, and interactive Q&A that current systems cannot fully replace end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5FAA regulations explicitly require licensed air traffic controllers to conduct pre-flight briefings and hold accountability for safety information provided to pilots. Legal authorization, liability frameworks, and mandatory human sign-off create hard barriers to full automation.
Adoption barriersclaude-sonnet-55/5Air traffic safety communications are heavily regulated (FAA/ICAO), requiring certified controllers, making autonomous AI briefing legally and operationally prohibited without human sign-off.
Cost vs. human wageclaude-haiku-4-5-202510012/5The integration, oversight, and liability management of an AI-assisted briefing system would likely approach or exceed the cost of a human controller's briefing time, especially given regulatory compliance and error-checking burdens.
Cost vs. human wageclaude-sonnet-53/5Automated data aggregation is cheap, but the human oversight, certification, and liability requirements keep the effective cost comparable rather than order-of-magnitude cheaper.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product reliably conducts full pre-flight briefings as a substitute for human air traffic controllers. Weather integration and flight planning tools exist, but they serve as assistive systems rather than autonomous briefing agents, and regulatory requirements mandate human controller involvement.
Technical feasibility todayclaude-sonnet-52/5Automated weather briefing tools (e.g., Flight Service aviation weather systems) exist and provide data, but they are decision-support tools rather than autonomous replacements for controller/briefer judgment in production today.

Analyze factors such as weather reports, fuel requirements, or maps to determine air routes.

21

CI 18–24 · exposure 25 · augmentation 63 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Despite digitization in aviation, the air traffic control sector has slow adoption of autonomous AI decision-making due to safety-critical regulation, high organizational friction around liability and certification, and the requirement for human controllers to remain responsible. Adoption of AI is primarily in supporting tools, not autonomous routing.
Sector adoption velocityclaude-sonnet-52/5Aviation and air traffic control are highly regulated, safety-critical sectors with slow, cautious technology adoption cycles despite growing use of decision-support tools.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist controllers by analyzing weather patterns, predicting congestion, and suggesting alternative routes, which improves situational awareness and decision speed. However, augmentation is limited to analysis and recommendation; controllers retain full authority over final route approval.
Augmentation potentialclaude-sonnet-54/5AI-powered weather analytics, fuel optimization algorithms, and route-planning software meaningfully speed up and improve the quality of a controller's route analysis while the human retains final decision authority.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can process weather data and fuel requirements, analyzing these factors holistically to determine safe, efficient air routes requires real-time decision-making, adherence to complex airspace regulations, and integration of dynamic variables that current systems cannot fully automate. This task demands continuous human oversight and judgment that exceeds the 50% time-saving bar for autonomous performance.
Task automatabilityclaude-sonnet-52/5While AI can process weather data and calculate fuel-optimal routes computationally, the task requires real-time integration with dynamic constraints and legal responsibility that current systems cannot fully replace end-to-end., so only partial time savings are realistic today.
Adoption barriersclaude-haiku-4-5-202510015/5This task has hard regulatory and legal barriers: licensed air traffic controllers must legally perform route determination and separation assurance. FAA certification and standards mandate human authorization, and liability for routing errors rests with certified personnel, preventing full substitution of AI for human judgment.
Adoption barriersclaude-sonnet-55/5Air traffic control is a heavily regulated, licensed profession where a certified human controller must be legally responsible for route determination and safety-critical decisions.
Cost vs. human wageclaude-haiku-4-5-202510012/5Integration of AI decision-support tools into ATM systems requires significant infrastructure, certification, and liability oversight, while the cost of errors is extremely high. The all-in cost of AI routing systems (development, validation, compliance, insurance) remains comparable to or higher than human controller labor for this critical safety function.
Cost vs. human wageclaude-sonnet-53/5Software-assisted route calculation is already cheaper than pure manual analysis, but the need for certified oversight and integration with ATC systems keeps all-in costs roughly comparable to current staffing costs.
Technical feasibility todayclaude-haiku-4-5-202510012/5Flight planning software assists with route optimization and can suggest alternatives based on structured inputs, but no deployed product independently determines safe air routes without human air traffic control validation and approval. Existing tools are decision-support systems rather than autonomous route-determination systems operating in production airspace.
Technical feasibility todayclaude-sonnet-52/5Flight planning software and dispatch tools already incorporate weather and fuel data, but these are advisory tools used by controllers/dispatchers rather than autonomous systems performing this analysis independently in production.

Relay air traffic information, such as courses, altitudes, or expected arrival times, to control centers.

13

CI 5–20 · exposure 17 · augmentation 38 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Adoption of autonomous AI in air traffic control remains virtually zero; the sector is extremely conservative due to safety criticality, regulatory constraints, and pilot/controller reliance on human judgment. Pilots and controllers actively resist removing the human element.
Sector adoption velocityclaude-sonnet-52/5Aviation is a highly regulated, safety-critical sector with slow, cautious technology adoption cycles despite some automation in flight data systems.
Augmentation potentialclaude-haiku-4-5-202510012/5AI can assist by auto-formatting or highlighting data (e.g., flagging conflicts), but the core task of dynamic relay and decision-making under time pressure requires human expertise and liability that AI tools have not materially enhanced to date in operational ATC.
Augmentation potentialclaude-sonnet-53/5Automated data feeds and decision-support tools help controllers track and relay information faster and more accurately, improving efficiency while humans remain in control.
Task automatabilityclaude-haiku-4-5-202510011/5Relaying air traffic information requires real-time situational awareness, judgment about safety-critical details, and coordination with human pilots and other controllers in a dynamic safety system where any error can be catastrophic. Current AI cannot reliably handle the full complexity and liability of this task end-to-end.
Task automatabilityclaude-sonnet-52/5Relaying structured flight data could partly be automated via data links and automated handoff systems, but real-time verbal coordination with judgment about traffic conflicts still requires human controllers today.
Adoption barriersclaude-haiku-4-5-202510015/5Air traffic control is heavily regulated under FAA and ICAO rules; a licensed air traffic controller must legally perform or directly supervise these tasks. Liability for accidents is extreme, and there is no regulatory pathway for autonomous AI to make safety-critical decisions in this domain.
Adoption barriersclaude-sonnet-55/5Air traffic control is heavily regulated, requires licensed controllers, and involves extreme liability and safety requirements that legally mandate human authority over these communications.
Cost vs. human wageclaude-haiku-4-5-202510011/5The liability, oversight, and verification costs of AI-assisted or autonomous ATC relay are enormous relative to a trained human controller's salary, and any meaningful deployment would require extensive certification and redundancy, making AI far more expensive than human labor.
Cost vs. human wageclaude-sonnet-52/5Existing automated data systems are cheap to run, but the human-in-the-loop verification and communication layer required for safety keeps overall cost comparable to or higher than pure automation savings would suggest.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can extract and format basic flight data from radar/systems, no deployed product reliably performs the full task of controlling and relaying complex, real-time air traffic information in production. Regulatory approval for AI in safety-critical ATC decisions remains absent.
Technical feasibility todayclaude-sonnet-52/5Automated data-sharing systems (e.g., flight data processing, ADS-B feeds) exist and are used operationally, but voice relay of course/altitude/ETA between controllers remains largely human-performed and safety-critical.

Direct ground traffic, including taxiing aircraft, maintenance or baggage vehicles, or airport workers.

11

CI 0–23 · exposure 13 · augmentation 50 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Aviation is a heavily regulated, safety-critical sector with high organizational friction and conservative adoption patterns. Despite decades of opportunity, full automation of ground traffic direction remains limited to experimental or narrow deployments, with no broad sector-wide adoption.
Sector adoption velocityclaude-sonnet-51/5Aviation ground control is an extremely conservative, safety-regulated sector with minimal AI deployment for actual control functions; adoption is limited to advisory tools.
Augmentation potentialclaude-haiku-4-5-202510013/5AI tools can assist controllers by providing enhanced radar/surface tracking visualization, automated conflict alerts, and routing suggestions, improving situational awareness and decision support. However, the human must remain actively engaged in the control loop for safety reasons.
Augmentation potentialclaude-sonnet-53/5AI-enabled surface surveillance, conflict alerts, and traffic flow tools already assist controllers in monitoring ground movements, improving situational awareness while humans retain full control.
Task automatabilityclaude-haiku-4-5-202510012/5While some elements like vehicle detection and basic routing could be partially automated, the task requires real-time judgment calls involving safety coordination, human worker safety, and dynamic decision-making under high-stakes conditions. Current AI cannot reliably manage the full complexity of ground traffic coordination to meet a ≥50% time-saving threshold with equivalent safety.
Task automatabilityclaude-sonnet-51/5Real-time, safety-critical ground traffic direction requires continuous perception, split-second judgment, and accountability that current AI cannot perform end-to-end without human control.
Adoption barriersclaude-haiku-4-5-202510014/5This task is heavily regulated under FAA and international aviation authorities, which mandate human responsibility for ground safety. Legal liability and safety certification requirements create substantial barriers to removing human oversight, and the task involves direct responsibility for worker and aircraft safety.
Adoption barriersclaude-sonnet-55/5Air traffic control is heavily regulated, requires certified licensed controllers, and carries extreme liability and safety consequences, creating hard legal and regulatory barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Integrating AI systems for ground traffic direction requires significant infrastructure investment, continuous oversight, and liability management. The all-in cost of robust automation with safety assurance likely exceeds the cost of trained air traffic controllers performing this task.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this task alone, so cost comparison favors the human controller who is legally required and currently irreplaceable.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed system reliably performs end-to-end ground traffic direction autonomously. Some airports have tested automated taxiing systems and vehicle detection, but these operate in narrow, controlled scenarios and require human oversight. Production automation of full ground traffic direction is not demonstrated at scale.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously directs airport ground traffic; existing surface-movement radar and A-SMGCS systems are decision-support tools for human controllers, not autonomous directors.

Inspect, adjust, or control radio equipment or airport lights.

11

CI 0–23 · exposure 13 · augmentation 38 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Adoption of automation in airport infrastructure maintenance is slow; these tasks remain heavily regulated and require certified technicians. No evidence of significant AI-driven displacement in this domain.
Sector adoption velocityclaude-sonnet-51/5Air traffic control and airport infrastructure maintenance are highly regulated, safety-critical physical domains with minimal AI adoption for hands-on equipment tasks.
Augmentation potentialclaude-haiku-4-5-202510013/5AI could meaningfully assist technicians by providing real-time diagnostics, predictive maintenance alerts, or equipment status monitoring to guide their physical inspection and adjustment work, improving efficiency without removing the human from the loop.
Augmentation potentialclaude-sonnet-52/5AI-based diagnostic and monitoring tools can flag anomalies or predict maintenance needs, offering some assistance, but the core physical inspection and adjustment work sees little AI augmentation.
Task automatabilityclaude-haiku-4-5-202510012/5Physical inspection, adjustment, and control of radio equipment and airport lights requires hands-on intervention in real-world infrastructure that current AI systems cannot perform remotely or autonomously. While AI could assist in diagnostics or monitoring, the core task of physically manipulating equipment remains outside current automation capability.
Task automatabilityclaude-sonnet-51/5This involves physical inspection and manual adjustment of radio equipment and airport lighting systems, requiring hands-on manipulation and safety-critical judgment that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5FAA and ICAO regulations require licensed, trained personnel to maintain and adjust critical air navigation infrastructure. Safety-critical nature of airport lighting and radio equipment creates legal and regulatory requirements that a human must sign off on or directly perform the work.
Adoption barriersclaude-sonnet-55/5Airport lighting and radio navigation equipment are safety-critical, FAA-regulated systems requiring certified personnel to inspect and maintain them, creating hard regulatory and liability barriers.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI systems capable of assisting with diagnostics or monitoring would still require human technicians to perform the physical work, resulting in minimal cost savings compared to the loaded wage of skilled airport maintenance personnel who specialize in this equipment.
Cost vs. human wageclaude-sonnet-51/5AI has no capability to perform the physical inspection/adjustment, so there is no viable cost comparison—human technicians remain the only option.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed AI systems reliably perform physical equipment inspection and adjustment at airport installations. Remote monitoring tools exist, but the actual hands-on control and adjustment of radio and lighting systems requires human technicians and is not automated in production.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product autonomously inspects or physically adjusts radio equipment or airport lights; this remains a physical maintenance/technician task performed by humans.

Monitor or direct the movement of aircraft within an assigned air space or on the ground at airports to minimize delays and maximize safety.

9

CI 0–18 · exposure 13 · augmentation 63 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Adoption of AI for autonomous or near-autonomous air traffic control is minimal; the sector is highly regulated, conservative, and human-controller employment is politically protected. Pilots in advanced automation (e.g., conflict detection aids) progress slowly due to certification and safety validation requirements.
Sector adoption velocityclaude-sonnet-51/5Aviation safety systems are notoriously slow to adopt automation for core control functions due to certification cycles, safety culture, and regulatory conservatism.
Augmentation potentialclaude-haiku-4-5-202510013/5AI tools that augment controllers—enhanced radar displays, automated conflict alerts, weather prediction integration, and workload-reducing data fusion—are increasingly deployed and do improve situational awareness and efficiency. However, the core decision and communication task remains human-driven.
Augmentation potentialclaude-sonnet-54/5AI and automation (e.g., conflict detection, trajectory prediction, decision-support tools) meaningfully assist controllers in managing traffic flow and reducing workload, even though humans retain full control authority.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can assist with radar tracking and conflict detection, the task requires real-time decision-making in dynamic, safety-critical conditions where human judgment, communication, and responsibility are irreplaceable. Current systems cannot reliably handle the full scope—coordinating multiple aircraft, weather changes, pilot communications, and emergency scenarios—without human oversight, falling short of 50% time savings at equal safety.
Task automatabilityclaude-sonnet-51/5Real-time, safety-critical separation of aircraft requires continuous perception, split-second judgment, and accountability that no current off-the-shelf AI system can perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Air traffic control is governed by strict FAA (and international) regulations that explicitly require licensed human controllers to exercise operational authority. Legal liability for accidents falls on humans and organizations; automation of safety-critical movement decisions faces hard regulatory and legal barriers to substitution.
Adoption barriersclaude-sonnet-55/5Air traffic control is one of the most heavily regulated, licensed professions in the world, with strict certification, legal liability, and international safety regulation (FAA, ICAO) requiring human controllers by law.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI infrastructure for air traffic support is expensive (specialized hardware, integration with legacy systems, continuous operation and maintenance), and controllers' salaries, while substantial, are justified by the irreducible human responsibility. AI is not yet cheaper all-in when reliability and liability are factored in.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute providing equivalent output, so cost comparison favors the human by default; any AI system would need extensive redundant safety infrastructure making it far more expensive than current ATC labor per unit of coverage.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product performs end-to-end air traffic control autonomously today. Research prototypes exist for conflict detection and automation within narrow bounds, but production systems still rely on human controllers making final decisions. The regulatory environment and liability structure preclude full automation in practice.
Technical feasibility todayclaude-sonnet-51/5No deployed product independently controls live air traffic; decision-support and automation tools exist but a certified human controller remains in full control at all times.

Determine the timing or procedures for flight vector changes.

9

CI 0–18 · exposure 13 · augmentation 50 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5ATC adoption of AI for core operational decisions remains minimal despite decades of research; the sector is highly risk-averse, safety-critical, and regulatory-constrained, resulting in slow and cautious pilot deployments only.
Sector adoption velocityclaude-sonnet-51/5Aviation safety systems are highly conservative and slow-moving; despite decades of automation research (e.g., NextGen), controllers still manually issue vector instructions with only decision-support augmentation.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist controllers by providing traffic conflict alerts, suggested maneuvers, and workload analysis, meaningfully improving decision support; however, the human controller remains central and legally responsible for all vector determinations.
Augmentation potentialclaude-sonnet-53/5Conflict-alert and trajectory-prediction tools already assist controllers by flagging potential conflicts and suggesting resolutions, improving situational awareness even though humans retain full decision authority.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can process flight data and suggest vector changes, determining timing and procedures requires real-time judgment of complex, dynamic airspace interactions and safety-critical decision-making that current systems cannot reliably handle end-to-end without significant human oversight.
Task automatabilityclaude-sonnet-51/5Real-time separation of aircraft under dynamic, safety-critical conditions requires split-second judgment, weather adaptation, and accountability that current AI cannot reliably replicate end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Air traffic control is heavily regulated by FAA and international bodies; controllers must hold licenses and maintain legal accountability for flight safety. Automation of vector decisions faces hard regulatory barriers requiring certified human approval and sign-off.
Adoption barriersclaude-sonnet-55/5Air traffic control is one of the most tightly regulated safety-critical professions, requiring FAA/ICAO-certified controllers with legal authority; automation of core separation decisions is essentially prohibited without new regulatory frameworks.
Cost vs. human wageclaude-haiku-4-5-202510012/5Implementing reliable AI-assisted or autonomous vector-change systems requires substantial infrastructure, validation, and ongoing human oversight costs that approach or exceed the loaded cost of experienced air traffic controllers.
Cost vs. human wageclaude-sonnet-51/5Given the extreme cost of any error (catastrophic collision risk), any AI system would require redundant human oversight, making all-in cost comparable to or higher than current staffing.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed ATC system makes autonomous decisions on vector changes in operational airspace; existing tools provide decision support only. Prototype systems exist in research and limited trials, but production systems still require human controllers to make final determinations.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously determines live vector changes for air traffic control; decision-support tools exist but humans make and execute all separation decisions.

Organize flight plans or traffic management plans to prepare for planes about to enter assigned airspace.

8

CI 0–16 · exposure 13 · augmentation 50 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Air traffic control is a government-regulated, safety-critical sector with strict licensing and strict resistance to replacing human authority. Adoption of AI in this domain has been slow, pilot-focused, and primarily assistive rather than replacive due to regulatory, liability, and institutional friction.
Sector adoption velocityclaude-sonnet-52/5Aviation is a highly regulated, safety-critical sector where AI adoption for core control functions is minimal and proceeds cautiously via decision-support tools rather than substitution.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist ATCs by automating routine flight-plan optimizations, flagging conflicts, and organizing data presentations, improving controller efficiency. However, augmentation remains limited by the need for human judgment on safety, weather, emergencies, and strategic decisions.
Augmentation potentialclaude-sonnet-53/5AI-based traffic flow prediction and conflict-detection tools can help controllers organize and anticipate incoming traffic, improving planning efficiency while humans remain fully in control.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can assist with organizing and optimizing flight plans algorithmically, the task requires real-time decision-making, conflict resolution, and integration with live traffic data that demands human oversight. Current AI systems cannot reliably handle the full end-to-end task independently without supervised control, falling short of the 50% time-saving threshold for unsupervised performance.
Task automatabilityclaude-sonnet-51/5Real-time organization of flight plans requires split-second judgment integrating weather, aircraft performance, and safety-critical coordination that current AI cannot perform end-to-end without human control.
Adoption barriersclaude-haiku-4-5-202510015/5Air traffic control is one of the most heavily regulated safety-critical domains; Federal Air Regulations mandate that licensed ATCs legally perform and sign off on traffic management decisions. Liability, certification, and human-contact requirements create hard barriers to full automation.
Adoption barriersclaude-sonnet-55/5Air traffic control is heavily regulated (FAA/ICAO) and requires licensed, certified controllers; legal and safety frameworks mandate human control of live traffic.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI systems for traffic management require significant infrastructure investment, domain expertise for integration, continuous oversight, and liability insurance. The all-in cost exceeds the loaded wage of a single controller, particularly given regulatory compliance and safety validation overhead.
Cost vs. human wageclaude-sonnet-51/5Given safety-critical certification requirements and the catastrophic cost of errors, any AI system would require extensive redundant human oversight, making it more expensive than current staffing models today.
Technical feasibility todayclaude-haiku-4-5-202510012/5Some decision-support tools and flight optimization software exist in production, but no deployed AI system independently organizes and prepares traffic management plans without human ATCs making final decisions and approvals. Real-world air traffic control retains mandatory human authority due to safety criticality.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously organizes live traffic management plans in operational airspace; decision-support tools exist but humans retain full responsibility.

Inform pilots about nearby planes or potentially hazardous conditions, such as weather, speed and direction of wind, or visibility problems.

7

CI 0–14 · exposure 13 · augmentation 50 · importance 4.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Air traffic control is heavily regulated and operates in a slow-moving, safety-obsessed sector with deep organizational and legal resistance to full automation of pilot communication; adoption of AI in this domain has been minimal and only in narrow, human-supervised support roles.
Sector adoption velocityclaude-sonnet-51/5Aviation safety systems are extremely slow to adopt automation for controller-facing communication tasks due to regulatory and safety-of-life constraints.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist controllers by automatically detecting hazards, alerting to traffic conflicts, and synthesizing weather and radar data into clearer displays, moderately raising controller productivity while humans retain decision-making and communication authority.
Augmentation potentialclaude-sonnet-53/5AI-based weather prediction, conflict detection, and decision-support tools already help controllers anticipate hazards and manage workload, improving situational awareness without replacing the human role.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can process radar data and weather information to identify hazards, the task requires real-time human judgment, split-second decision-making, and direct two-way communication with pilots in complex, safety-critical situations. Current AI systems cannot reliably replace the full end-to-end task with 50% time savings at equal safety quality.
Task automatabilityclaude-sonnet-51/5Real-time separation and hazard communication requires split-second judgment, liability, and integration with live radar/voice systems that current AI cannot autonomously replace end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5This task is protected by strict FAA regulations requiring a licensed human air traffic controller to make safety decisions and communicate with pilots; legal liability for safety-critical failures rests on the human controller, and there is no path to substitute automation without fundamental regulatory change.
Adoption barriersclaude-sonnet-55/5Air traffic control is a licensed, safety-critical function governed by strict aviation regulations (FAA/ICAO) requiring certified human controllers to communicate hazards.
Cost vs. human wageclaude-haiku-4-5-202510011/5An air traffic controller's loaded cost is moderate, but the infrastructure and oversight required to implement AI-based pilot communication systems would be substantial, and any remaining human oversight would negate cost savings.
Cost vs. human wageclaude-sonnet-51/5Given certification, redundancy, and failure-cost requirements, any AI system replacing this function would be far more costly than the marginal cost of an already-employed controller.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed air traffic control system operates without human controllers making the final safety decisions and communications. Research systems can assist with hazard detection and alert generation, but production air traffic control relies entirely on licensed human controllers for this communication task.
Technical feasibility todayclaude-sonnet-51/5No deployed product independently performs live pilot-hazard communication in operational airspace; ATC automation exists only as decision-support tools, not autonomous communicators.

Direct pilots to runways when space is available or direct them to maintain a traffic pattern until there is space for them to land.

7

CI 0–14 · exposure 13 · augmentation 50 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Adoption of AI in core ATC decision-making is extremely limited. Safety requirements, regulatory caution, and the critical nature of the task mean only very slow, incremental pilot projects are underway. The aviation sector is inherently conservative on automation of life-safety functions.
Sector adoption velocityclaude-sonnet-51/5Aviation ATC is an extremely conservative, safety-regulated sector with negligible AI-driven displacement of controllers; adoption is confined to research and limited decision-support pilots.
Augmentation potentialclaude-haiku-4-5-202510013/5AI tools can assist by predicting traffic flow, recommending sequences, and flagging conflicts, allowing controllers to make faster and more informed decisions. However, the human controller remains essential to the loop, providing moderate productivity gains rather than transformative enhancement.
Augmentation potentialclaude-sonnet-53/5AI-based conflict prediction, workload management, and sequencing tools can assist controllers in planning traffic flow, though the core directive communication with pilots remains human-led.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can assist in sequencing and pattern analysis, this task requires real-time safety-critical decisions balancing multiple dynamic variables (weather, aircraft systems, pilot inputs, separation standards). Current AI cannot reliably handle the full responsibility of directing aircraft to runways or traffic patterns without continuous human oversight, falling short of the 50% time-saving threshold.
Task automatabilityclaude-sonnet-51/5Real-time separation of aircraft and dynamic runway sequencing under safety-critical constraints requires split-second judgment, communication, and accountability that current AI cannot perform end-to-end with equal quality and safety assurance.
Adoption barriersclaude-haiku-4-5-202510015/5Air traffic control is heavily regulated (FAA, ICAO) and legally requires licensed controllers to bear responsibility for flight safety. Controllers must hold certification and are personally accountable for decisions, creating a hard regulatory and legal barrier to full automation.
Adoption barriersclaude-sonnet-55/5Air traffic control is heavily regulated, requires licensed certified controllers, and carries extreme liability for error, making full automation a hard legal and safety barrier.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of AI infrastructure, integration with legacy ATC systems, continuous monitoring, and required human oversight significantly exceeds the salary cost of a human controller. Liability and safety certification requirements add substantial overhead.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this function, so cost comparison favors humans by default; any AI system would require redundant certified human oversight, adding cost rather than saving it.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed AI system currently makes autonomous landing-sequencing decisions in production air traffic control. Research and simulation systems exist, but active air traffic facilities still rely on human controllers, with AI only in advisory/decision-support roles—not performing the task end-to-end reliably.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously directs live aircraft traffic in operational airspace; decision-support tools exist but a human controller issues all instructions.

Monitor aircraft within a specific airspace, using radar, computer equipment, or visual references.

7

CI 0–14 · exposure 13 · augmentation 50 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Adoption of full monitoring automation in air traffic control is negligible; the sector remains highly conservative due to safety criticality and regulatory lock-in. Modernization efforts focus on incremental tool improvements, not controller displacement.
Sector adoption velocityclaude-sonnet-51/5Aviation safety systems are extremely conservative and slow-moving; AI adoption in live ATC operations is essentially nonexistent beyond limited advisory tools in pilot programs.
Augmentation potentialclaude-haiku-4-5-202510013/5AI-assisted tools (automated conflict detection, enhanced radar visualization, weather integration) do improve controller productivity and situational awareness in production systems, providing useful assistance on specific subtasks like data filtering and alerting.
Augmentation potentialclaude-sonnet-53/5AI-enhanced radar processing, conflict-alert systems, and decision-support tools already assist controllers in tracking and predicting traffic, improving situational awareness without replacing their monitoring role.
Task automatabilityclaude-haiku-4-5-202510012/5While radar and computer monitoring can be partially automated for data collection and anomaly detection, the dynamic integration of multiple information streams, rapid decision-making under uncertainty, and continuous human judgment required for safe separation and conflict resolution cannot be reliably automated end-to-end today. Current systems assist but do not meet the ≥50% time-saving threshold for full task replacement at equal safety quality.
Task automatabilityclaude-sonnet-51/5Real-time separation of aircraft with life-safety consequences requires continuous human judgment and legal accountability; no current AI system operates end-to-end without a licensed controller in the loop.
Adoption barriersclaude-haiku-4-5-202510015/5Air traffic control is heavily regulated; FAA and international aviation authorities require licensed controllers to perform this task legally. Liability asymmetry is severe—automation failures risk lives and face massive penalties, creating structural barriers to unsupervised substitution.
Adoption barriersclaude-sonnet-55/5Air traffic control is one of the most tightly regulated, licensed professions in the world with strict certification, liability, and safety-of-life requirements mandating human control.
Cost vs. human wageclaude-haiku-4-5-202510011/5The infrastructure cost of aviation-grade monitoring systems, integration with legacy ATC equipment, regulatory validation, and ongoing human oversight far exceeds the cost savings from any partial automation. Full system replacement is not economically viable given safety requirements.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this task alone, so cost comparison favors humans by default; any AI overlay adds cost on top of required human staffing rather than replacing it.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product performs full air traffic monitoring autonomously in production; aviation regulations require licensed human controllers. Partial automation exists (tools like conflict alerting, weather radar processing), but these are assistive, not replacements. The task remains legally and operationally dependent on human controllers.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously monitors and separates live air traffic; decision-support and conflict-alert tools exist but the core monitoring/control function remains fully human-performed in production ATC systems.

Contact pilots by radio to provide meteorological, navigational, or other information.

7

CI 0–14 · exposure 13 · augmentation 50 · 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 laggard in automation of safety-critical human-contact tasks; regulatory conservatism, union agreements, and safety-first culture mean adoption of autonomous pilot communication is negligible to zero in practice.
Sector adoption velocityclaude-sonnet-51/5Aviation ATC is a highly regulated, safety-critical, low-digitization-for-automation sector where deployment of autonomous communication systems is proceeding extremely slowly, if at all.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist controllers by pre-generating weather briefings, formatting navigational alerts, and suggesting routine information to communicate, reducing drafting time; however, the controller must review, authorize, and deliver all communications, so augmentation is significant but bounded.
Augmentation potentialclaude-sonnet-53/5AI-assisted tools (e.g., automated weather briefing generation, data-link relay, transcription aids) can help controllers prepare or deliver information faster, though the core radio interaction remains human-performed.
Task automatabilityclaude-haiku-4-5-202510012/5While AI could generate meteorological and navigational information with high accuracy, the task requires real-time two-way radio communication with pilots making dynamic safety decisions. Current AI cannot reliably conduct emergent, safety-critical radio dialogue with appropriate tone, exception handling, and pilot-specific context at the speed and reliability required; manual intervention would still be necessary, preventing the ≥50% time saving threshold.
Task automatabilityclaude-sonnet-51/5This is real-time, safety-critical, two-way voice communication requiring situational awareness across multiple aircraft; no off-the-shelf system can perform this end-to-end today with equal reliability and quality.
Adoption barriersclaude-haiku-4-5-202510015/5Air traffic control is one of the most heavily regulated and safety-critical domains; FAA certification, legal liability for pilot safety, and explicit human authorization/responsibility requirements make autonomous AI radio contact legally prohibited without a licensed controller in the loop and legally accountable.
Adoption barriersclaude-sonnet-55/5Air traffic control is a licensed, government-regulated function with strict legal requirements that a certified controller communicate with pilots; safety regulation and liability make substitution essentially prohibited.
Cost vs. human wageclaude-haiku-4-5-202510011/5Current AI infrastructure (speech synthesis, NLU, integration, and required oversight) combined with the liability and re-verification costs would exceed the hourly wage of a human controller, particularly given the zero-tolerance error environment.
Cost vs. human wageclaude-sonnet-51/5Given the extreme reliability, certification, and redundancy requirements of ATC systems, any AI solution would require costly validation, hardware, and human oversight, making it more expensive than current staffing in practice.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed air traffic control system uses AI to autonomously conduct pilot-radio communication. Text-to-speech and information retrieval exist, but FAA-certified autonomous radio contact with pilots is not in production anywhere; this remains research-stage and regulatory-blocked.
Technical feasibility todayclaude-sonnet-51/5There are research prototypes for automated data-link and voice weather/NOTAM relay, but no deployed product independently contacts pilots and provides operational information without a licensed controller.

Check conditions and traffic at different altitudes in response to pilots' requests for altitude changes.

7

CI 0–14 · exposure 13 · augmentation 50 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Adoption of autonomous ATC decision-making is negligible in practice; regulatory constraints and safety culture create structural resistance, and even research testbeds remain isolated from operational deployment, making this a laggard sector for AI substitution despite high digitization.
Sector adoption velocityclaude-sonnet-51/5Aviation safety systems are notoriously slow to adopt automation due to certification, redundancy, and regulatory approval processes, with automation efforts focused on decision support rather than replacement.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist controllers by highlighting traffic conflicts, recommending altitude options, or predicting separation violations, improving situation awareness and reducing cognitive load; however, the human controller must evaluate and execute the final decision, making this a true support rather than a transformative augmentation of controller productivity.
Augmentation potentialclaude-sonnet-53/5AI-based decision-support tools (e.g., conflict detection, traffic flow prediction) can help controllers process complex airspace data faster, improving situational awareness while the human retains full authority.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can process radar data and flight information, this task critically requires real-time judgment about separation standards, conflict resolution, and dynamic decision-making under uncertainty that current automation cannot reliably perform end-to-end. The task involves responding to pilots' requests with nuanced safety determinations that remain beyond the >50% time-saving threshold for fully autonomous execution.
Task automatabilityclaude-sonnet-51/5This is real-time, safety-critical decision-making requiring integration of live radar, weather, and communication data with legal responsibility resting on the human controller; no current AI system can perform this end-to-end at equal quality with time savings.
Adoption barriersclaude-haiku-4-5-202510015/5Extremely high barriers exist: federal law (14 CFR) mandates that certified air traffic controllers must personally provide air traffic services; liability for separation violations rests with human controllers; and the FAA strictly regulates automation of safety-critical ATC functions, making algorithmic replacement legally infeasible.
Adoption barriersclaude-sonnet-55/5Air traffic control is a highly regulated, licensed profession where safety regulations (FAA, ICAO, etc.) mandate certified human controllers for such decisions, making legal substitution essentially prohibited today.
Cost vs. human wageclaude-haiku-4-5-202510011/5Current AI decision-support tools for ATC require significant human operator, infrastructure, and integration costs; the loaded cost of a trained air traffic controller remains lower than deploying and maintaining reliable AI systems with necessary fallback and oversight mechanisms for this safety-critical task.
Cost vs. human wageclaude-sonnet-51/5Given the total absence of deployable autonomous systems and the extreme liability cost of errors, any AI substitute would require redundant human oversight, making it costlier than the human baseline.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed AI system currently performs full altitude-change clearance decisions in production air traffic control without human controller oversight. Research prototypes exist for decision support, but no mature product reliably handles the dynamic, safety-critical judgment required to independently approve or deny altitude changes across diverse traffic scenarios.
Technical feasibility todayclaude-sonnet-51/5There are no deployed products that autonomously handle altitude clearance decisions in live ATC operations; all real-world systems remain human-operated with decision support tools only.

Provide on-the-job training to new air traffic controllers.

3

CI 0–5 · exposure 5 · augmentation 38 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Air traffic control is a heavily regulated, safety-critical domain with strict certification requirements. Adoption of AI for core training functions is negligible; the sector remains dependent on human instructors by regulatory mandate.
Sector adoption velocityclaude-sonnet-51/5Aviation safety training is a highly conservative, regulation-bound sector with minimal AI adoption for actual instructional authority or certification tasks.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist with scheduling, simulation setup, performance data analysis, or training material preparation, but augmentation in live operational training is limited. The inherent need for human judgment and real-time safety oversight restricts meaningful productivity gains.
Augmentation potentialclaude-sonnet-53/5AI-driven simulators, scenario generators, and performance analytics can meaningfully support trainers in preparing exercises and tracking trainee progress, even though the human instructor remains central.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires real-time mentoring, adaptive feedback, judgment of trainee competence, and high-stakes decision-making oversight in a safety-critical environment. Current AI systems cannot reliably replace human instructors who must continuously evaluate trainee performance and make live operational decisions.
Task automatabilityclaude-sonnet-51/5Training new controllers requires live mentorship, real-time judgment transfer, and adaptive coaching in high-stakes dynamic environments that current AI cannot replicate end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Federal Aviation Administration (FAA) regulations require certified, qualified air traffic controllers to provide on-the-job training. Legal and safety liability for trainee errors rests on the human instructor, creating a hard regulatory barrier to AI replacement.
Adoption barriersclaude-sonnet-55/5Air traffic control training is heavily regulated (FAA/ICAO), requires certified human instructors and evaluators, and carries extreme liability and safety stakes that legally mandate human sign-off.
Cost vs. human wageclaude-haiku-4-5-202510011/5Fully trained air traffic controllers are required by regulation to conduct this training, and their labor cost far exceeds any AI inference cost. Integration and oversight would still require certified instructors, making any AI-only approach infeasible and therefore not cost-competitive.
Cost vs. human wageclaude-sonnet-51/5Any AI-based training tools require expensive certified human oversight and validation, so total cost is not meaningfully lower than experienced human trainers who are already embedded in operations.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI systems can provide training materials, simulations, and some tutoring assistance, no deployed product reliably performs full on-the-job instruction of air traffic controllers. Training simulators exist but require human instructors for evaluation and live operational training, which remains dependent on human expertise.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product independently trains or certifies air traffic controllers; simulators exist but human trainers remain essential and are not replaced by any production system.

Issue landing and take-off authorizations or instructions.

0

CI 0–0 · exposure 0 · augmentation 38 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Despite decades of research, adoption of AI for independent authorization remains zero in production worldwide. Regulatory capture, safety culture, and the catastrophic-error asymmetry mean this sector is a laggard in automation, with no signs of change.
Sector adoption velocityclaude-sonnet-51/5Aviation safety systems are extremely conservative and slow-moving in regulatory approval; no jurisdiction is deploying AI-issued clearances at scale or even in serious pilot programs for live traffic.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist controllers by filtering radar data or predicting conflicts, but current augmentation tools are limited; most assistance today comes from older decision-support systems, not modern AI. The safety-critical nature and cognitive load of the task limit practical augmentation.
Augmentation potentialclaude-sonnet-53/5AI-based decision-support tools (trajectory prediction, conflict detection, workload balancing) already assist controllers in preparing and timing instructions, improving efficiency without replacing the authorization role.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires real-time judgment under high-stakes conditions, continuous situational awareness of dynamic airspace, and live communication coordination with pilots—none of which current AI can do end-to-end safely or reliably. The task involves split-second decision-making that balances competing demands and emergency responses that AI cannot match today.
Task automatabilityclaude-sonnet-51/5Real-time, safety-critical separation decisions require continuous perception, split-second judgment, and legal accountability that current AI cannot replicate end-to-end; no system approaches the 50% time-savings threshold since a licensed controller must still perform the task.
Adoption barriersclaude-haiku-4-5-202510015/5This task has hard legal and regulatory barriers: the FAA and international aviation authorities mandate that licensed human air traffic controllers must personally authorize all take-offs and landings. Liability for accidents rests on the human controller, creating legal requirement for human sign-off.
Adoption barriersclaude-sonnet-55/5Air traffic control is one of the most heavily regulated human-in-the-loop professions; FAA/ICAO rules mandate licensed human controllers to issue clearances, with severe liability for errors.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of integrating and validating an AI system for safety-critical authorization would exceed the salary of air traffic controllers, and the liability and oversight costs would be enormous given zero tolerance for failure.
Cost vs. human wageclaude-sonnet-51/5Given the extreme liability, certification, and safety infrastructure needed for any autonomous ATC system, the all-in cost of an AI replacement (if it existed) would vastly exceed a controller's wage today.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI system performs live air traffic control authorization independently; this remains research-stage and is explicitly prohibited by aviation regulation. All operational systems today require human controllers to issue binding authorizations.
Technical feasibility todayclaude-sonnet-51/5No deployed product issues live landing/takeoff clearances autonomously; existing decision-support tools (conflict alerts, sequencing tools) only assist, they do not replace the controller issuing instructions.

Transfer control of departing flights to traffic control centers and accept control of arriving flights.

0

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

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Aviation remains one of the most conservative and heavily regulated sectors; automation of core control functions is explicitly prohibited by law, and adoption of AI for these tasks is not occurring.
Sector adoption velocityclaude-sonnet-51/5Aviation safety systems are among the slowest sectors to adopt autonomous AI due to regulatory certification requirements and catastrophic failure costs.
Augmentation potentialclaude-haiku-4-5-202510012/5While AI systems can provide limited decision support (e.g., conflict detection, weather alerts), they offer minimal augmentation to the actual transfer-of-control task, which is fundamentally a legal and procedural handoff requiring human authority.
Augmentation potentialclaude-sonnet-52/5Some automation tools assist with flight tracking and conflict alerts, but the actual verbal handoff and control transfer process sees minimal AI-driven productivity enhancement today.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires real-time judgment, safety-critical coordination with multiple stakeholders, and legal authority transfer—functions that demand human responsibility and cannot be delegated end-to-end to AI systems today.
Task automatabilityclaude-sonnet-51/5This involves real-time, safety-critical handoffs of aircraft control requiring split-second judgment, verbal coordination, and legal accountability; no current AI system performs this end-to-end.'
Adoption barriersclaude-haiku-4-5-202510015/5Air traffic control is among the most heavily regulated occupations globally; FAA, ICAO, and equivalent authorities mandate that a licensed, certified human controller must maintain direct operational authority and sign responsibility for all control transfers.
Adoption barriersclaude-sonnet-55/5Air traffic control is heavily regulated, requires FAA/ICAO-equivalent licensing, and involves extreme liability for safety failures, making human sign-off legally mandatory.
Cost vs. human wageclaude-haiku-4-5-202510011/5Air traffic control requires certified personnel with legal accountability; the human cost includes salary, training, and licensing but cannot be substituted by AI inference, making AI comparatively more expensive when safety and liability are factored in.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this function, so cost comparison favors humans entirely; any AI system would require redundant certified human oversight, adding cost rather than saving it.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product performs flight control handoffs in production; this remains a strictly human responsibility in all aviation jurisdictions due to regulatory and safety requirements.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously executes control handoffs between ATC facilities; this remains firmly in human hands with only decision-support tools in research or limited trial stages.

Provide flight path changes or directions to emergency landing fields for pilots traveling in bad weather or in emergency situations.

0

CI 0–0 · exposure 0 · augmentation 38 · 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 one of the most conservative and heavily regulated sectors; adoption of AI for core ATC functions, especially emergency response, is minimal and unlikely to accelerate given safety criticality and regulatory lock-in.
Sector adoption velocityclaude-sonnet-51/5Aviation safety systems are notoriously slow to adopt automation for control functions, especially for emergency scenarios, due to regulatory conservatism and safety-critical certification requirements.
Augmentation potentialclaude-haiku-4-5-202510012/5While AI could potentially assist with data visualization or weather prediction as supporting tools, the real-time, safety-critical, judgment-intensive nature of emergency landing decisions limits meaningful augmentation; controllers cannot afford latency or uncertainty in such scenarios.
Augmentation potentialclaude-sonnet-53/5AI can assist by providing weather data, suggesting alternate airports, or flagging conflicts, but the controller retains full responsibility for real-time emergency guidance.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires real-time decision-making under uncertainty, direct communication with pilots, dynamic coordination of multiple aircraft, and judgment calls about safety in novel emergency contexts—none of which current AI systems can reliably perform end-to-end without human oversight and intervention.
Task automatabilityclaude-sonnet-51/5This is a real-time, life-critical decision task requiring split-second judgment under uncertainty; no current AI system can autonomously direct emergency aircraft rerouting end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Air traffic control is among the most heavily regulated domains; federal law (FAA) mandates licensed human controllers with direct legal responsibility for safety decisions, and automation of emergency handling faces hard regulatory and liability barriers that prevent substitution.
Adoption barriersclaude-sonnet-55/5Air traffic control is heavily regulated and requires FAA/ICAO licensed controllers; emergency directives legally and operationally require certified human judgment and accountability.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of integrating, training, validating, and maintaining AI systems for this safety-critical domain far exceeds the salary of a human controller, especially given required redundancy, certification, and liability coverage.
Cost vs. human wageclaude-sonnet-51/5Given the near-total need for certified human oversight and catastrophic failure costs, any AI system would need extensive redundant human backup, making it more expensive than a human controller alone.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product today performs live air traffic control decision-making for emergency landings; the liability and safety-critical nature mean even high-accuracy research systems remain confined to simulation and are not deployed in production airspace.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs live emergency ATC direction; existing AI decision-support tools are advisory-only research or limited trials, not operational replacements.

Alert airport emergency services in cases of emergency or when aircraft are experiencing difficulties.

0

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

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5The aviation sector has strict regulatory oversight and safety-first culture; actual adoption of autonomous emergency alerting is effectively zero, with no path toward displacement of this core human responsibility.
Sector adoption velocityclaude-sonnet-51/5Aviation safety systems are among the most conservative and slow-moving sectors for AI adoption due to certification requirements and extreme risk aversion.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist by automatically detecting anomalies in telemetry and flagging potential emergencies for controller review, but the human controller must retain decision authority and the task fundamentally remains human-centered.
Augmentation potentialclaude-sonnet-52/5AI-based decision-support tools can flag anomalies or aid situational awareness, but the actual judgment and alerting action remains human-driven with limited AI assistance in current systems.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires real-time human judgment to assess dynamic, safety-critical situations and make urgent decisions about resource deployment. No current AI system can reliably make the deterministic call to alert emergency services without human validation in production environments.
Task automatabilityclaude-sonnet-51/5This requires real-time judgment under high-stakes, safety-critical conditions with legal accountability; no current AI system performs this end-to-end.wed
Adoption barriersclaude-haiku-4-5-202510015/5Air traffic control is heavily regulated by aviation authorities (FAA, EASA, etc.); a human controller is legally required to assess and respond to emergencies. Liability for false or missed alerts creates strong legal barriers to full automation of emergency response decisions.
Adoption barriersclaude-sonnet-55/5Air traffic control is a licensed, heavily regulated profession with strict certification requirements and legal liability, making unauthorized automation of emergency alerting essentially prohibited.
Cost vs. human wageclaude-haiku-4-5-202510011/5Automating this task would require AI infrastructure with redundancy, regulatory compliance, and continuous monitoring that would exceed the cost of a human controller performing the alerting function, especially given liability concerns.
Cost vs. human wageclaude-sonnet-51/5Given the near-total absence of deployed automation for this task, there is no meaningful AI cost basis to compare against the human controller's wage.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product independently alerts emergency services for aircraft emergencies; this remains entirely within the domain of human air traffic controllers. Any such system would require human sign-off and verification before triggering real emergency responses.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously alerts emergency services during aircraft emergencies; this remains fully a human controller function in production ATC systems.

Maintain radio or telephone contact with adjacent control towers, terminal control units, or other area control centers to coordinate aircraft movement.

0

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

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Air traffic control remains heavily regulated and human-centric; automation in this domain is incremental (e.g., tools assisting coordination) rather than replacement. The FAA's cautious, safety-first approach means adoption of autonomous coordination systems is nearly zero.
Sector adoption velocityclaude-sonnet-51/5Aviation safety systems are notoriously slow to adopt automation for control functions due to regulatory certification cycles and extreme risk aversion; no meaningful AI deployment exists in this specific task domain.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist by providing real-time traffic visualization, conflict predictions, or communication logging, but the core negotiation and decision-making remain human. Assistive tools exist but offer limited productivity gains on the core coordination task itself.
Augmentation potentialclaude-sonnet-52/5AI could potentially assist with monitoring, alerting, or transcription support, but does not yet meaningfully transform the core coordination task, which remains a manual voice-based judgment process.
Task automatabilityclaude-haiku-4-5-202510011/5This task fundamentally requires real-time human judgment, negotiation, and accountability in safety-critical coordination. Current AI cannot reliably conduct two-way radio conversations with other human controllers, understand complex verbal instructions, or make dynamic aircraft coordination decisions that have legal liability.
Task automatabilityclaude-sonnet-51/5This involves live, safety-critical coordination requiring split-second judgment, negotiation, and accountability between human controllers; no current AI system performs this end-to-end reliably in operational airspace.
Adoption barriersclaude-haiku-4-5-202510015/5Federal Aviation Regulations (14 CFR Part 65) legally require licensed, human air traffic controllers to coordinate aircraft movement and maintain radio contact. Liability and safety-critical requirements mean only a licensed human can legally perform or sign off on these decisions.
Adoption barriersclaude-sonnet-55/5Air traffic control is one of the most heavily regulated, licensed professions in existence, with strict certification, legal liability, and safety-of-life requirements mandating human controllers.
Cost vs. human wageclaude-haiku-4-5-202510011/5The infrastructure, training, and oversight required to automate this task would exceed the cost of human controllers. Additionally, any error carries massive liability, making human-equivalent performance mandatory, which AI cannot yet achieve at lower cost.
Cost vs. human wageclaude-sonnet-51/5Given the near-total absence of viable automation and the extreme liability/safety requirements, any AI solution would require extensive human oversight, making it more costly than current staffing when accounting for certification and risk mitigation.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably performs inter-facility air traffic coordination today. This requires seamless bidirectional voice communication, situational understanding across jurisdictions, and regulatory approval—far beyond current AI capabilities. Research systems cannot yet handle the real-time, high-consequence nature of control tower communications.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously conducts inter-facility ATC coordination in production; research on AI-assisted ATC exists but nothing is operational at this level of responsibility.

Initiate or coordinate searches for missing aircraft.

0

CI 0–0 · exposure 0 · augmentation 25 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Air traffic control remains a tightly regulated, human-supervised domain with strong legal and safety mandates; automation adoption in this sector is extremely cautious and procedurally constrained, not following tech-sector rapid-adoption patterns.
Sector adoption velocityclaude-sonnet-51/5Air traffic control is a highly regulated, safety-critical sector with extremely slow AI adoption for core operational and emergency-response duties.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could potentially assist by aggregating radar/flight data or suggesting search patterns, but the core task of coordinating a search operation—communicating with authorities, making strategic decisions, and taking responsibility—must remain controller-driven, limiting genuine augmentation.
Augmentation potentialclaude-sonnet-52/5AI can assist with radar data analysis, flight tracking, or alerting systems to flag anomalies, but the coordination and decision-making for search initiation remains a human-led process with limited AI tool support.
Task automatabilityclaude-haiku-4-5-202510011/5Initiating or coordinating searches for missing aircraft requires dynamic decision-making under high uncertainty, real-time communication with multiple stakeholders (pilots, ground crews, authorities), and judgment about search scope and strategy that current AI systems cannot perform end-to-end autonomously.
Task automatabilityclaude-sonnet-51/5Initiating and coordinating search-and-rescue for missing aircraft requires real-time crisis judgment, multi-agency coordination, and legal authority that current AI cannot replicate end-to-end.'
Adoption barriersclaude-haiku-4-5-202510015/5Air Traffic Controllers are federally licensed professionals, and search coordination is legally mandated to be conducted by authorized personnel; regulatory requirements and liability asymmetry create hard barriers to automation or autonomous AI substitution.
Adoption barriersclaude-sonnet-55/5This is a safety-critical, legally mandated function requiring certified air traffic controllers with authority to declare emergencies and coordinate with SAR/military assets; strict FAA/ICAO regulation applies.
Cost vs. human wageclaude-haiku-4-5-202510011/5Any AI assistance would require expert human oversight, integration with multiple legacy systems, and validation by licensed professionals; the total cost would exceed that of direct human operation for such a critical, low-volume task.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this function, so cost comparison favors the human controller who is already embedded in required infrastructure and authority chains.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI system today reliably coordinates missing aircraft searches; this task involves complex human coordination, legal authority, and real-time situational judgment that remains firmly in the human domain and has not been demonstrated in production systems.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product initiates or coordinates missing-aircraft searches; this remains a human-led emergency response function involving multiple agencies (SAR, military, ATC facilities).

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.