Transportation Security Screeners

33-9093.00
Median wage $66,770/yr50,290 employed (US)Rank #774 of 923 scored · top 84% by substitution

Conduct screening of passengers, baggage, or cargo to ensure compliance with Transportation Security Administration (TSA) regulations. May operate basic security equipment such as x-ray machines and hand wands at screening checkpoints.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution16
Exposure17
Augmentation45

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

26 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%17

panel mean rating 1.7/5 → substitution pressure 17/100

Technical feasibility todayw 20%17

panel mean rating 1.7/5 → substitution pressure 17/100

Cost vs. human wagew 15%18

panel mean rating 1.7/5 → substitution pressure 18/100

Adoption barriersw 20%inverted — strong barriers lower the score10

panel mean rating 4.6/5 (barrier strength) → substitution pressure 10/100

Sector adoption velocityw 10%14

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

Task breakdown (26 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.

Provide directions and respond to passenger inquiries.

49

CI 30–67 · exposure 50 · augmentation 63 · importance 3.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5Major airports have piloted self-service kiosks and AI chatbots for passenger inquiries, but full displacement remains inconsistent; most facilities still maintain human staff for directions and inquiries, suggesting middling adoption with pilots common but not yet universal production rollout.
Sector adoption velocityclaude-sonnet-52/5Government security/aviation sector is a slow adopter of AI-driven passenger service tools compared to fast-moving information or finance sectors, with pilots for kiosks/apps but limited deep integration.
Augmentation potentialclaude-haiku-4-5-202510014/5AI assistants already augment human screeners by providing instant access to procedures, waits, and FAQ responses via mobile or voice interfaces, enabling staff to focus on higher-value security decisions while remaining informed and productive.
Augmentation potentialclaude-sonnet-53/5AI-powered apps, digital signage, and chatbots can offload routine directional questions, allowing screeners to focus on security tasks, providing moderate productivity assistance.
Task automatabilityclaude-haiku-4-5-202510014/5Current AI systems can handle a majority of standard passenger inquiries and provide directions via chatbots and voice assistants with high accuracy. However, edge cases involving complex security concerns, accessibility needs, or unusual situations would require human intervention, preventing a perfect 5.
Task automatabilityclaude-sonnet-52/5Basic wayfinding and FAQ-type inquiries could be handled by kiosks or chatbots, but real-time in-person direction-giving in a dynamic, secure airport environment with diverse passenger needs and safety implications requires human presence and judgment.
Adoption barriersclaude-haiku-4-5-202510013/5While there are no strict legal barriers preventing automation of informational tasks, security screening environments have customer preference for human presence, oversight requirements for sensitive security information, and organizational friction around replacing visible staff at checkpoints.
Adoption barriersclaude-sonnet-53/5No strict licensing requirement for answering questions, but TSA screeners are federally trained/certified personnel whose physical presence is mandated for security operations, limiting full substitution even for ancillary tasks like directions.
Cost vs. human wageclaude-haiku-4-5-202510014/5AI systems scale to handle thousands of inquiries at minimal marginal cost compared to the loaded wage of a human screener. Integration and maintenance are modest; the cost-per-inquiry is substantially lower than labor.
Cost vs. human wageclaude-sonnet-52/5Screeners are already staffed for security duties, so adding AI kiosks for inquiries incurs additional infrastructure cost without eliminating the human presence needed for security functions, making cost savings marginal.
Technical feasibility todayclaude-haiku-4-5-202510014/5Deployed products (airport kiosks, chatbots, mobile apps, voice assistants) demonstrate reliable performance on routine inquiries and directions in many airports today. Some systems have material limitations in handling ambiguity or context, but production-grade solutions are widely operational.
Technical feasibility todayclaude-sonnet-52/5Airport signage, apps, and some chatbot/kiosk systems exist for basic wayfinding, but no deployed AI product reliably handles the full range of live, ad-hoc passenger inquiries at security checkpoints.

Check passengers' tickets to ensure that they are valid, and to determine whether passengers have designations that require special handling, such as providing photo identification.

36

CI 23–49 · exposure 38 · augmentation 63 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Despite decades of TSA operations, ticket/designation checking remains primarily manual; adoption of AI screening tools in actual airports is minimal, reflecting regulatory constraints, security sensitivity, and organizational reluctance to deploy automation in this federally controlled domain.
Sector adoption velocityclaude-sonnet-53/5Airports and TSA have been steadily rolling out biometric and automated ID/boarding pass verification (e.g., CAT machines, facial recognition), indicating moderate but accelerating adoption in this specific security-sector niche.
Augmentation potentialclaude-haiku-4-5-202510013/5AI could assist by pre-scanning tickets and flagging missing IDs or designation codes before the human screener reviews them, speeding up the human's workflow without removing their authority or judgment.
Augmentation potentialclaude-sonnet-54/5AI-powered ID verification and biometric matching significantly speeds up and improves accuracy of the ticket/ID check step, meaningfully augmenting human screeners' throughput and accuracy.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can read and validate ticket barcodes/text and flag missing photo IDs, the task requires human judgment for edge cases (altered documents, suspicious behavior, special circumstances) and legal authority to deny boarding that cannot yet be fully automated end-to-end with consistent quality.
Task automatabilityclaude-sonnet-53/5Ticket/ID validation and flagging special-handling designations is largely a data-matching task that could be automated via document scanning and database lookups, but the current human role also involves physical verification and behavioral judgment that AI alone doesn't fully replace end-to-end.'
Adoption barriersclaude-haiku-4-5-202510014/5TSA regulations and federal law require trained, certified human security screeners; automation of ticket validation and passenger designation checking would likely require regulatory approval and cannot fully replace the human-authority requirement to screen passengers and deny boarding.
Adoption barriersclaude-sonnet-54/5TSA screening functions are federally regulated and screeners are government-authorized; identity/security verification carries high liability and legal requirements for certified personnel or systems, creating strong regulatory barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Current AI document-reading systems are relatively inexpensive per scan, but the integration, oversight infrastructure, and liability costs in a safety-critical setting approach human screener costs, especially given the need for human backup and training.
Cost vs. human wageclaude-sonnet-54/5Automated scanning/kiosk systems are cheap per-passenger compared to staffing costs once installed, though capital and integration costs are non-trivial versus a simple human check.
Technical feasibility todayclaude-haiku-4-5-202510012/5Document scanning and barcode reading are deployed, but production systems for this task remain limited; most TSA checkpoints still rely on human screeners, and AI systems lack the authority and real-world validation track record needed for consistent, liability-safe deployment.
Technical feasibility todayclaude-sonnet-53/5Automated e-gates, boarding pass scanners, and biometric/ID verification kiosks are deployed at many airports (e.g., CLEAR, TSA PreCheck biometric lanes), but full replacement of the human screener check with material error tolerance is not universal or complete.

Record information about any baggage that sets off alarms in monitoring equipment.

34

CI 25–43 · exposure 38 · augmentation 50 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Airport security screening is heavily regulated and government-operated; adoption of AI is constrained by federal procurement, TSA policy, and security protocols rather than market forces, resulting in slower, cautious deployment compared to commercial sectors.
Sector adoption velocityclaude-sonnet-52/5Airport security screening is a slow-moving, highly regulated, physically-anchored sector with limited AI agent deployment beyond narrow detection-assist tools.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist screeners by automatically flagging suspicious baggage and pre-populating alert records, reducing manual search time and documentation overhead, but the human must remain in the decision loop for regulatory and liability reasons.
Augmentation potentialclaude-sonnet-53/5AI-assisted data capture and auto-population of alarm records can meaningfully speed up screener documentation, though humans remain central to verification and physical inspection.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can detect baggage triggering alarms through computer vision of X-ray or scanning data, recording that information requires minimal human input and no meaningful time saving because the detection itself is instantaneous and the recording is trivial data entry. The bottleneck is security investigation, not documentation.
Task automatabilityclaude-sonnet-53/5Recording structured alarm data (bag ID, alarm type, timestamp) is a simple logging task that automation could largely handle, but it's tied to physical screening events requiring integration with detection hardware and human confirmation.5This limits full end-to-end automation today.
Adoption barriersclaude-haiku-4-5-202510014/5TSA operations are federally regulated; transportation security screening is a legally mandated human responsibility under 49 U.S.C. § 44901, and liability for missed threats creates strong regulatory and organizational barriers to full automation without human sign-off.
Adoption barriersclaude-sonnet-54/5This task sits within a highly regulated security context (TSA), where documentation accuracy and chain-of-custody matter for security and legal reasons, creating real institutional and regulatory barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Implementing AI-assisted baggage monitoring requires substantial infrastructure investment and integration costs per airport location, while the task itself (recording information) is low-wage clerical work, making the all-in cost comparable or higher than direct human performance.
Cost vs. human wageclaude-sonnet-53/5Automated logging software is cheap to run, but integration with legacy screening equipment, security certification, and required human oversight raise effective costs closer to parity with existing labor already performing this as part of their job.
Technical feasibility todayclaude-haiku-4-5-202510013/5AI systems can classify baggage images and flag anomalies in deployed airport screening systems, but current production systems still require human screeners to confirm and manually document findings, with meaningful error rates on edge cases that necessitate human oversight.
Technical feasibility todayclaude-sonnet-52/5Some screening systems auto-log alarm metadata, but comprehensive automated recording integrated with human decision workflows in TSA-type settings is not widely deployed as a mature standalone product.

Send checked baggage through automated screening machines, and set bags aside for searching or rescreening as indicated by equipment.

26

CI 20–32 · exposure 30 · augmentation 75 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5While airports deploy automated screening machines, adoption of AI-driven end-to-end automation remains limited due to regulatory constraints, security requirements, and the need for human judgment in high-stakes security decisions.
Sector adoption velocityclaude-sonnet-53/5Airports have adopted AI-enhanced CT scanners and automated screening lanes at a moderate pace, but full physical automation of bag handling and search triage remains limited to pilot programs at major hubs.
Augmentation potentialclaude-haiku-4-5-202510014/5AI-powered screening machines significantly augment human screeners by automatically detecting potential threats and flagging bags for secondary inspection, substantially improving both speed and consistency while the human retains final authority over screening decisions.
Augmentation potentialclaude-sonnet-54/5AI-based image recognition significantly assists screeners by flagging suspicious items and reducing manual scanning burden, improving speed and accuracy while humans still make final decisions.
Task automatabilityclaude-haiku-4-5-202510012/5While AI-powered baggage screening machines exist, the task also requires human judgment to decide when to set bags aside for manual search—decisions that depend on context, risk assessment, and regulatory compliance that current systems cannot fully automate reliably at the required quality threshold.
Task automatabilityclaude-sonnet-52/5Physical handling of luggage and responding to machine alerts requires manipulation and judgment in a dynamic environment that current general-purpose AI cannot fully replace, though the screening algorithms themselves are AI-assisted.atabase.
Adoption barriersclaude-haiku-4-5-202510015/5TSA regulations and security protocols legally mandate human screeners for baggage inspection, and liability for missed threats or security failures creates strong regulatory and legal barriers to full automation of this safety-critical function.
Adoption barriersclaude-sonnet-54/5TSA regulations require certified/authorized personnel to handle security screening decisions and physical searches, creating a strong regulatory and liability barrier to full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI screening equipment and integration costs are substantial, and the screening process still requires human operators present to monitor outputs and make decisions, limiting cost savings compared to the human wage for this security-critical role.
Cost vs. human wageclaude-sonnet-52/5Robotic baggage handling systems exist but require significant capital investment in conveyor/robotic infrastructure, and human oversight remains mandated, so cost savings versus a screener's wage are limited to detection assistance rather than full task replacement.
Technical feasibility todayclaude-haiku-4-5-202510013/5Automated screening machines with AI-assisted threat detection are deployed at airports, but they still require human operators to interpret alerts and make final decisions about bag handling; no system performs the complete task end-to-end without human oversight today.
Technical feasibility todayclaude-sonnet-52/5Automated X-ray/CT threat detection algorithms are deployed, but physical bag handling, sorting, and rescreening decisions still require human screeners; no product handles the full end-to-end task.

Direct passengers to areas where they can pick up their baggage after screening is complete.

26

CI 5–47 · exposure 20 · augmentation 25 · importance 3.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Airport security operations are heavily regulated and dominated by federal staff with limited technological adoption pressure. Passenger direction is a low-priority automation target in a sector where human presence is mandated and digitization is constrained by security and compliance rules.
Sector adoption velocityclaude-sonnet-52/5Airport security is a highly regulated, physically-oriented government function with slow technology adoption cycles compared to information-sector industries.
Augmentation potentialclaude-haiku-4-5-202510012/5Digital signage or wayfinding apps could assist by pre-displaying baggage claim locations, but AI adds minimal value to the human activity of verbally directing nearby passengers in real time. The assistance potential is limited to static information rather than dynamic task augmentation.
Augmentation potentialclaude-sonnet-52/5Digital signage, apps, or audio systems can supplement directions, but this narrow wayfinding task offers limited room for meaningful AI-driven productivity gains for the screener.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires real-time physical coordination, wayfinding in dynamic airport environments, and interpersonal interaction to direct moving passengers. Current AI systems cannot physically direct passengers or reliably navigate and manage fluid airport logistics to achieve meaningful time savings.
Task automatabilityclaude-sonnet-53/5Directing passengers is simple wayfinding that could be handled by signage, kiosks, or automated announcements, but real-time crowd management and verbal guidance still benefit from human presence in dynamic checkpoint environments.
Adoption barriersclaude-haiku-4-5-202510014/5Transportation Security Administration (TSA) regulations require human staff to operate screening facilities and direct passengers as part of security operations and customer service. Legal and regulatory mandates establish a hard barrier to substitution with AI or automation.
Adoption barriersclaude-sonnet-52/5No licensing requirement for this specific sub-task, though TSA screeners are federally authorized personnel and security checkpoints have organizational protocols that create some friction to full automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5Any viable automation would require significant robotics or IoT infrastructure (signage, navigation aids, staff coordination systems) that costs substantially more than the loaded wage of a security screener, with minimal offsetting labor reduction.
Cost vs. human wageclaude-sonnet-53/5Static signage or automated announcements are cheap, but handling exceptions (confused travelers, disabled passengers, congestion) still requires human labor, keeping costs roughly comparable overall.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product performs this task reliably in production. Wayfinding and passenger direction involve embodied navigation, real-time facility awareness, and responsive communication that current systems do not handle at scale in actual airports.
Technical feasibility todayclaude-sonnet-52/5While digital signage and automated announcement systems exist in some airports, no widespread deployed AI product autonomously directs individual passengers post-screening with the flexibility a human does.

Locate suspicious bags pictured in printouts sent from remote monitoring areas, and set these bags aside for inspection.

25

CI 23–28 · exposure 25 · augmentation 63 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Airport and transportation security sectors move slowly on automation due to regulatory oversight, unionized workforces, and high liability. Despite digitization, AI adoption in active screening duties remains minimal; pilots exist but production displacement in security screening roles is negligible.
Sector adoption velocityclaude-sonnet-53/5Transportation security has adopted AI-assisted image analysis and automated threat detection at a moderate pace, with pilots and some production deployments, but overall physical screening remains labor intensive.
Augmentation potentialclaude-haiku-4-5-202510013/5AI-assisted flagging and image highlighting could help screeners review printouts faster and reduce missed bags, improving their productivity. However, the augmentation is moderate because the screener must still verify and physically handle each bag, limiting the leverage AI provides.
Augmentation potentialclaude-sonnet-54/5AI-enhanced imaging and anomaly-detection overlays already help screeners flag and prioritize suspicious bags faster, meaningfully boosting throughput and detection accuracy while a human still makes the final call and retrieves the bag.
Task automatabilityclaude-haiku-4-5-202510012/5Current AI can detect objects in images with high accuracy, but locating bags in printouts and making threat judgments requires integration with physical retrieval workflows. The task has meaningful automation potential for image analysis but lacks the embodied action (physically setting bags aside) and contextual judgment that would achieve 50% time savings end-to-end today.
Task automatabilityclaude-sonnet-52/5Requires physically locating a specific bag among many on-site and physically setting it aside, which is a physical manipulation task not achievable end-to-end by current AI without robotics.the visual identification component could be AI-assisted but the retrieval action requires a human.
Adoption barriersclaude-haiku-4-5-202510014/5Transportation Security Administration (TSA) and security regulations require screened authorization and human responsibility for threat decisions and bag handling. Liability for missed threats creates high error-cost asymmetry, and regulatory requirements demand that a licensed human ultimately sign off on or perform the physical security action.
Adoption barriersclaude-sonnet-54/5Aviation security is heavily regulated (TSA/ICAO requirements) mandating trained, authorized personnel to physically handle and inspect flagged luggage, creating strong regulatory and liability barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Computer vision inference is cheap, but integration with security workflows, printout digitization, and human oversight add significant overhead. The loaded wage for a screener remains competitive with the total cost of a hybrid system, especially given liability and re-check costs.
Cost vs. human wageclaude-sonnet-52/5AI screening software adds cost on top of existing scanner infrastructure, but the physical bag retrieval still requires paid human labor, so total cost savings versus a human screener are limited.
Technical feasibility todayclaude-haiku-4-5-202510012/5Computer vision models can identify bags in images reliably, but deployed security systems typically flag anomalies for human review rather than autonomously making bag-segregation decisions. No mature product demonstrably performs the full task (detection + physical setaside) reliably in production airport or transit security at scale.
Technical feasibility todayclaude-sonnet-52/5Automated threat detection systems (CT scanners, ML-based image analysis) exist and are deployed in airports, but the specific task of matching a remote printout to a physical bag and pulling it aside is still a human/manual workflow step.

Watch for potentially dangerous persons whose pictures are posted at checkpoints.

25

CI 20–30 · exposure 30 · augmentation 63 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Despite technological capability, TSA and airport operators have adopted face recognition primarily as an *assistive* tool (pre-screening, alerting), not as a replacement for human screeners. Regulatory, liability, and public-safety concerns have kept adoption slow and limited.
Sector adoption velocityclaude-sonnet-52/5Aviation security is a slow-moving, highly regulated government sector where AI/biometric adoption is proceeding via limited pilots rather than widespread deployment.
Augmentation potentialclaude-haiku-4-5-202510014/5Real-time facial recognition watchlist alerts can meaningfully assist human screeners by flagging priority cases and reducing search time. This augmentation is already in limited deployment and significantly raises screener productivity without removing human judgment from the final call.
Augmentation potentialclaude-sonnet-53/5AI-based facial recognition and alert systems can meaningfully assist screeners by flagging potential matches, improving vigilance and reducing missed detections, while humans remain responsible for verification and action.
Task automatabilityclaude-haiku-4-5-202510012/5Face recognition can identify individuals against watchlists, but the task involves nuanced behavioral assessment ('potentially dangerous persons'), threat context evaluation, and real-time decision-making under uncertainty. Current systems excel at facial matching but cannot reliably infer dangerousness from appearance alone, requiring substantial human judgment.
Task automatabilityclaude-sonnet-52/5Facial recognition/watchlist matching against posted images is technically automatable, but the task as performed by TSA screeners involves real-time visual vigilance integrated with physical screening duties, limiting full end-to-end automation today.
Adoption barriersclaude-haiku-4-5-202510015/5TSA screening is a federally regulated security function; only authorized personnel are legally permitted to make final disposition decisions on potential threats. Privacy law, liability for false negatives, and statutory requirements for human oversight create hard barriers to autonomous automation.
Adoption barriersclaude-sonnet-54/5Security screening is a federally regulated function with strict legal authority requirements; human TSA officers must be present and authorized to act on identifications, creating strong regulatory and liability barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Face recognition infrastructure has high upfront costs (cameras, databases, integration) and ongoing maintenance, while TSA screener labor is relatively low-cost in absolute terms. All-in per-checkpoint cost is comparable to or exceeds human screening labor.
Cost vs. human wageclaude-sonnet-52/5Biometric surveillance systems require significant camera infrastructure, integration, and human oversight for false positives, making near-term cost savings modest rather than order-of-magnitude cheaper.
Technical feasibility todayclaude-haiku-4-5-202510013/5Facial recognition systems are deployed in some airports for watchlist matching, but they operate with material false positive/negative rates and typically require human verification. No fully autonomous end-to-end system reliably performs the threat assessment component in production.
Technical feasibility todayclaude-sonnet-52/5Facial recognition systems exist and are piloted at some checkpoints, but reliable, widely deployed production systems performing this specific watch-and-match duty across all screeners are not yet standard.

Contact leads or supervisors to discuss objects of concern that are not on prohibited object lists.

24

CI 0–47 · exposure 28 · augmentation 38 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5TSA operations are highly regulated government functions with strong institutional resistance to automation of security decisions. Current deployment favors human judgment in flagged-item scenarios, and adoption of AI-driven escalation triggering is minimal to non-existent in production security screening.
Sector adoption velocityclaude-sonnet-51/5Airport security is a highly regulated, low-digitization-for-decision-authority environment with minimal AI agent deployment for judgment calls and human escalation processes.
Augmentation potentialclaude-haiku-4-5-202510013/5AI could assist screeners by automatically drafting communication templates, organizing concern details, and suggesting relevant supervisor contacts, meaningfully reducing composition time while human judgment remains central to the escalation decision.
Augmentation potentialclaude-sonnet-52/5AI-assisted imaging systems can flag anomalies to aid screener judgment, offering some support, but the actual interpersonal escalation conversation itself is not augmented by current AI tools.
Task automatabilityclaude-haiku-4-5-202510014/5AI could draft emails, identify concern patterns, and prompt appropriate escalation with high consistency, easily achieving 50% time savings. The task involves structured communication with clear triggering conditions, which LLMs handle reliably, though final message composition may benefit from human review.
Task automatabilityclaude-sonnet-51/5This task is inherently interpersonal communication requiring real-time judgment escalation and human authority; AI cannot substitute for the act of contacting and discussing with a supervisor.'
Adoption barriersclaude-haiku-4-5-202510014/5Transportation Security Administration rules and liability frameworks require human security personnel to make judgment calls on ambiguous items and sign off on escalations. A licensed human screener must ultimately validate and authorize the contact, creating a legal/regulatory barrier to full automation.
Adoption barriersclaude-sonnet-55/5Security screening is heavily regulated (TSA), requires authorized personnel with legal authority to make security determinations and escalate anomalies, precluding AI substitution.
Cost vs. human wageclaude-haiku-4-5-202510014/5AI inference and integration costs for classification and draft communication are orders of magnitude cheaper than the loaded wage of a screener or supervisor time spent composing messages and deciding escalations, once integration overhead is amortized.
Cost vs. human wageclaude-sonnet-51/5There is no AI product performing this task, so no cost comparison favors AI; human labor is the only viable option today.
Technical feasibility todayclaude-haiku-4-5-202510012/5No production system reliably handles this end-to-end in security screening contexts. While email drafting and classification tools exist, the sensitivity of flagging security concerns and the requirement to contact actual supervisors means current AI products are insufficient for autonomous deployment without significant human oversight.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs this escalation-communication task; it requires a human screener to identify ambiguity and a human supervisor to weigh in.

Test baggage for any explosive materials, using equipment such as explosive detection machines or chemical swab systems.

23

CI 20–25 · exposure 30 · augmentation 75 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Deployment of detection equipment is mature and widespread at major airports, but the model remains human-supervised screening with machine assistance. True automation (unattended baggage analysis without human clearance) has seen minimal adoption due to regulatory and safety constraints.
Sector adoption velocityclaude-sonnet-52/5Security screening is a highly regulated, safety-critical government function with slow-moving procurement and certification cycles, resulting in incremental rather than fast AI adoption.
Augmentation potentialclaude-haiku-4-5-202510014/5Modern explosive detection equipment substantially augments human screeners by automating the scan-and-flag process, reducing manual chemical testing needs, and enabling faster throughput while humans focus on judgment calls and anomalies.
Augmentation potentialclaude-sonnet-54/5AI-enhanced detection algorithms (e.g., automated threat recognition in CT scanners) already assist screeners by flagging suspicious items and improving accuracy, meaningfully boosting screener throughput and detection rates.
Task automatabilityclaude-haiku-4-5-202510012/5While explosive detection machines can automatically scan baggage and flag anomalies, human judgment is essential to interpret results, handle ambiguous findings, and make final clearance decisions. The task requires verification of machine outputs rather than fully autonomous end-to-end execution.
Task automatabilityclaude-sonnet-52/5Physical operation of detection equipment and interpreting ambiguous scan results still requires a human operator on-site, though the machines themselves already do the sensing autonomously; full end-to-end automation without human oversight is not yet deployed at scale.
Adoption barriersclaude-haiku-4-5-202510015/5Federal law (TSA regulations) requires a human screener to conduct or authorize baggage screening for security purposes; liability for missed explosives creates strong legal and procedural barriers against full automation independent of human oversight.
Adoption barriersclaude-sonnet-55/5Aviation security screening is heavily regulated (e.g., TSA/ICAO requirements) and mandates certified human screeners and government oversight, making unsupervised automation legally prohibited.
Cost vs. human wageclaude-haiku-4-5-202510012/5Detection equipment is capital-intensive (machines cost tens of thousands of dollars, maintenance required) and still requires human operators for interpretation and decision-making, making the all-in cost per task comparable to or higher than the human wage for screening.
Cost vs. human wageclaude-sonnet-52/5The detection hardware itself is expensive and highly specialized, and human screeners are still required for operation, loading, and alarm resolution, so cost savings versus a human-operated system are limited.
Technical feasibility todayclaude-haiku-4-5-202510012/5Detection equipment exists and is deployed at scale in airports, but these are tools that assist rather than replace human screeners; they generate alerts that humans must evaluate and act upon. No production system performs the full task—human authorization and judgment—without human involvement.
Technical feasibility todayclaude-sonnet-53/5Explosive detection machines and chemical swab systems are mature, widely deployed products, but they still require human screeners to run swabs, position bags, and interpret alerts, so the task isn't fully automated in production.

Monitor passenger flow through screening checkpoints to ensure order and efficiency.

23

CI 20–25 · exposure 25 · augmentation 50 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5TSA adoption of autonomous monitoring is minimal despite digitization pressure; most deployments remain pilots in video review and anomaly flagging. Federal procurement, security constraints, and entrenched staffing models slow meaningful production adoption.
Sector adoption velocityclaude-sonnet-52/5Airport security is a highly regulated, safety-critical government sector where AI adoption for people management remains at the pilot stage, not production-scale replacement.
Augmentation potentialclaude-haiku-4-5-202510013/5AI-powered queue analytics and real-time congestion alerts can help screeners allocate resources and adjust checkpoint configuration, improving productivity when screeners remain in control of dispatch and passenger direction.
Augmentation potentialclaude-sonnet-53/5AI-driven crowd analytics and queue-monitoring dashboards can help screeners and supervisors better manage checkpoint flow and staffing decisions.
Task automatabilityclaude-haiku-4-5-202510012/5Current AI can passively monitor video feeds and flag anomalies, but directing passenger flow requires real-time decision-making, physical presence, and responsive communication that AI agents cannot yet execute end-to-end at security standards. No deployed system saves >50% of screener labor here.
Task automatabilityclaude-sonnet-52/5This requires real-time physical presence, crowd management, and situational authority over people in a security-sensitive space, which current AI cannot execute end-to-end even with camera-based monitoring assistance.ed.rating and rationale. This task requires physical presence and real-time human judgment to redirect people, manage lines, and respond to disruptions, which AI cannot autonomously execute today. Ratings below. Automatability=2, feasibility=2, cost_ratio=2, barriers=4, adoption_velocity=2, augmentation=3. Now rationale text finalized. Final answer below. Overall low automatability, low feasibility, moderate barriers due to security/regulatory context, slow sector adoption, some augmentation via computer vision flow analytics. Proceeding to output JSON. Rationale text will be concise per field. Automatability rationale: physical monitoring and crowd control require human presence and judgment; AI vision tools can flag congestion but cannot manage flow end-to-end. Feasibility rationale: some crowd analytics deployed in airports but not replacing human screeners' oversight function. Cost ratio rationale: sensor/camera systems plus human oversight still needed, so cost savings are limited compared to a screener's wage. Barriers rationale: TSA screeners are federally regulated/certified; security functions have high liability and regulatory barriers. Adoption velocity rationale: government security screening is a slow-adopting, highly regulated sector for full automation. Augmentation rationale: AI-based crowd density and flow analytics can help supervisors optimize checkpoint staffing and queue management. Finalizing JSON output now. Here is final.
Adoption barriersclaude-haiku-4-5-202510015/5TSA screeners are federal employees subject to strict regulatory authority, security clearance, and human accountability requirements. Law and regulation mandate human personnel responsible for checkpoint operations; automation faces hard legal barriers to substitution.
Adoption barriersclaude-sonnet-54/5TSA screening operations are federally regulated with certified personnel required for security-sensitive functions, creating strong regulatory and liability barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Camera systems and analytics software are relatively cheap, but 24/7 monitoring infrastructure plus integration into existing checkpoint workflows and human oversight makes total cost comparable to or higher than a screener's wage for equivalent output.
Cost vs. human wageclaude-sonnet-52/5Deploying sensors/cameras plus required human oversight for security compliance keeps costs comparable to or higher than staffing screeners for this function.
Technical feasibility todayclaude-haiku-4-5-202510012/5Existing computer vision systems detect congestion and queue length, but no production deployment fully monitors passenger flow and maintains checkpoint order without human oversight. Proof-of-concept exists; reliable autonomous deployment for security operations does not.
Technical feasibility todayclaude-sonnet-52/5Some computer-vision-based queue/crowd analytics exist in airports, but no product autonomously manages passenger flow and order at checkpoints without human screeners present.

Inspect carry-on items, using x-ray viewing equipment, to determine whether items contain objects that warrant further investigation.

21

CI 18–25 · exposure 30 · augmentation 63 · importance 4.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5TSA operations are slow to adopt new technologies due to regulatory lock-in, security requirements, and the need for extensive pilot testing. No measurable displacement of screeners by AI agents has occurred.
Sector adoption velocityclaude-sonnet-52/5Airport security is a highly regulated, safety-critical physical-world sector with slow, cautious technology rollout cycles compared to information-sector adoption speeds.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can highlight suspicious regions in x-ray images or flag statistical anomalies, assisting screeners in directing attention and potentially speeding review. However, the human must retain full decision authority, limiting augmentation upside.
Augmentation potentialclaude-sonnet-54/5AI-assisted image analysis and automated threat-detection algorithms meaningfully help screeners flag suspicious items faster and more consistently, improving throughput while humans retain final judgment.
Task automatabilityclaude-haiku-4-5-202510012/5While AI vision systems can detect certain anomalies in x-ray imagery, the task requires real-time judgment about threat levels, contextual decision-making, and handling edge cases that current systems struggle with. Achieving 50% time savings at equal quality is not demonstrated in production settings.
Task automatabilityclaude-sonnet-52/5Computer vision can flag suspicious x-ray images, but final determination and handling of ambiguous or adversarial cases still requires human judgment and physical follow-up, limiting full end-to-end automation today.
Adoption barriersclaude-haiku-4-5-202510015/5This task is heavily regulated by TSA and DHS; a licensed, authorized human screener is legally required to make final determinations. Liability for missed threats is asymmetrically costly, creating a hard barrier to full automation.
Adoption barriersclaude-sonnet-55/5Aviation security is heavily regulated (e.g., TSA/ICAO rules) and mandates human screeners or human sign-off for final security decisions, creating a hard legal barrier to full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Current AI systems require significant infrastructure, specialized training, and human oversight to match screener performance. The integration and liability costs are high relative to the wage of a single TSA screener, making it not yet cost-competitive.
Cost vs. human wageclaude-sonnet-52/5Security-grade x-ray AI systems require expensive certified hardware, integration, and continuous human oversight, so cost savings versus a screener's wage are moderate at best, not order-of-magnitude cheaper.
Technical feasibility todayclaude-haiku-4-5-202510012/5Research prototypes and limited trials exist for AI-assisted x-ray screening, but no deployed product reliably performs this task autonomously at airport scale. Material error rates and the high cost of false negatives prevent production deployment.
Technical feasibility todayclaude-sonnet-53/5Automated threat detection systems (e.g., algorithms integrated into airport x-ray scanners) are deployed in some airports to assist screeners, but they operate as decision-support tools rather than fully autonomous replacements and still have notable false positive/negative rates.

Decide whether baggage that triggers alarms should be searched or should be allowed to pass through.

21

CI 18–25 · exposure 25 · augmentation 63 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Adoption in TSA and airport security is slow because security screening is a federally regulated, risk-critical function with strict protocols and human accountability requirements. Despite decades of automation opportunity, screeners remain the decision-maker; early automation rollouts have been limited and heavily supervised.
Sector adoption velocityclaude-sonnet-52/5Aviation security is a highly regulated, safety-critical sector with slow, cautious technology adoption cycles, relying on incremental upgrades to detection algorithms rather than autonomous decision-making.
Augmentation potentialclaude-haiku-4-5-202510013/5AI-driven threat detection and image enhancement (e.g., 3D X-ray reconstruction, anomaly highlighting) meaningfully assist screeners by surfacing suspicious items and reducing operator fatigue, thereby improving accuracy and throughput. However, the AI does not handle the full decision-making burden, so augmentation is significant but partial.
Augmentation potentialclaude-sonnet-54/5AI-enhanced imaging and anomaly detection significantly aid screeners by highlighting suspicious items and reducing false negatives, improving decision speed and accuracy while humans remain in control.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can flag anomalies in screening imagery and sensor data, the decision to search baggage or allow passage involves risk assessment, judgment calls on ambiguous cases, and accountability that current AI systems cannot reliably automate end-to-end. Human TSA officers still make the final determination, and replacing this requires >50% time savings at equal quality—difficult given liability and false-positive costs.
Task automatabilityclaude-sonnet-52/5This decision requires integrating machine alarm output with contextual judgment, risk assessment, and legal authority to search, which current AI cannot fully perform end-to-end."},
Adoption barriersclaude-haiku-4-5-202510014/5Strong regulatory and legal barriers exist: TSA protocol requires a screener to make the search/pass decision, security policy mandates human accountability for false negatives (security threats), and replacing human judgment requires federal regulatory approval and liability reassignment. Substitution is not merely organizationally difficult but legally constrained.
Adoption barriersclaude-sonnet-55/5TSA regulations and federal law require certified human screeners to make search decisions and physically conduct searches, creating a hard legal and liability barrier to full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Current TSA screening labor is relatively low-cost per passenger ($5–15 loaded wage for 2–3 minute screening). AI systems require upfront capital, integration, maintenance, and human oversight of edge cases, making the all-in cost comparable to or higher than the marginal screener wage, especially accounting for liability and false-positive handling.
Cost vs. human wageclaude-sonnet-52/5Screening equipment with AI-assisted detection is costly to deploy and still requires human screeners for final decisions, so cost savings versus human labor are limited at this stage.
Technical feasibility todayclaude-haiku-4-5-202510012/5Deployed products exist for automated threat detection in X-ray and baggage screening, but they identify suspicious items rather than make the final pass/search decision. No production system today reliably makes this binary decision autonomously; human screeners interpret AI alerts and decide disposition, so the task itself is not yet fully performed by AI at scale.
Technical feasibility todayclaude-sonnet-52/5AI-assisted detection algorithms flag anomalies in scanners today, but the final search/pass decision is made by human screeners in deployed airport security systems, not by AI autonomously.

Inspect checked baggage for signs of tampering.

20

CI 20–20 · exposure 25 · augmentation 50 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Adoption of AI-assisted baggage screening in airports is limited and mostly experimental; legacy infrastructure, regulatory inertia, and the critical nature of the task slow deployment. Most airports still rely predominantly on human screeners and conventional scanners.
Sector adoption velocityclaude-sonnet-52/5Aviation security is a highly regulated, slow-moving sector with cautious technology adoption cycles tied to government procurement and certification processes, resulting in limited AI deployment for this specific task.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can flag suspicious regions in x-ray images or highlight potential anomalies to speed human review, modestly improving screener efficiency. However, the augmentation is incremental; the human remains the decision-maker and final arbiter of tampering risk.
Augmentation potentialclaude-sonnet-53/5AI-enhanced X-ray and CT imaging systems assist screeners by flagging anomalies and suspicious items, improving detection speed and consistency while humans retain final judgment.
Task automatabilityclaude-haiku-4-5-202510012/5Current AI can detect some visual anomalies in baggage imagery (e.g., x-ray images), but cannot reliably identify subtle signs of tampering, assess structural integrity, or make the nuanced security judgments required. End-to-end automation with 50% time savings at equal quality is not demonstrated.
Task automatabilityclaude-sonnet-52/5Physical inspection of checked baggage for tampering signs requires visual/tactile assessment and handling that current AI systems cannot fully replace end-to-end; some image analysis exists but doesn't cover the full task including physical handling and decision-making.'
Adoption barriersclaude-haiku-4-5-202510015/5Transportation security is heavily regulated; TSA and equivalent agencies legally mandate human screener involvement and responsibility for security decisions. Liability for missed tampering is asymmetric and severe, making substitution without human approval infeasible.
Adoption barriersclaude-sonnet-55/5TSA screening is federally regulated and mandates certified human screeners for security decisions, with strict liability and legal requirements preventing full automation of security judgment calls.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI systems for baggage screening require significant infrastructure (hardware, integration with existing scanners, maintenance), and must operate alongside human screeners; the all-in cost does not yet undercut human labor when oversight and integration overhead are included.
Cost vs. human wageclaude-sonnet-52/5Machine vision systems require significant capital investment (scanners, integration, maintenance) and still need human oversight, making costs comparable to or higher than existing screener labor for this specific function.
Technical feasibility todayclaude-haiku-4-5-202510012/5Computer vision systems exist for baggage scanning and anomaly detection, but deployed TSA and airport systems rely heavily on human screeners interpreting x-ray images; AI serves as a narrow assist tool rather than a reliable autonomous performer of the full tampering-inspection task.
Technical feasibility todayclaude-sonnet-52/5X-ray image analysis AI tools are deployed for threat detection, but tampering-specific inspection (locks, seals, exterior damage) still relies heavily on human screeners; no mature product fully replaces this specific sub-task.

View images of checked bags and cargo, using remote screening equipment, and alert baggage screeners or handlers to any possible problems.

19

CI 18–20 · exposure 25 · augmentation 63 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5TSA and airport security operate in a heavily regulated, risk-averse sector with slow innovation cycles. Adoption of AI-assisted screening remains minimal in production; pilots are ongoing but deployment at scale is negligible.
Sector adoption velocityclaude-sonnet-52/5Aviation security is a highly regulated, safety-critical government sector that adopts new detection technology slowly and only after extensive certification and testing cycles.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can highlight suspicious items or anomalies to assist screeners in focusing attention, raising throughput modestly, but human judgment and legal authority remain central to the task. Augmentation potential is moderate rather than transformative.
Augmentation potentialclaude-sonnet-54/5AI-assisted threat detection algorithms already highlight suspicious items and organic materials on X-ray images, meaningfully speeding up and improving screener accuracy while humans retain final decision authority.
Task automatabilityclaude-haiku-4-5-202510012/5While computer vision can detect some anomalies in X-ray imagery, TSA-grade threat detection requires distinguishing weapons, explosives, and contraband with extremely high precision and low false-positive rates—thresholds current AI falls short of reliably. The task cannot be fully automated to meet the 50%-time-saving bar without unacceptable security risk.
Task automatabilityclaude-sonnet-52/5Computer vision can flag anomalies in X-ray images, but final threat determination and alerting still require certified human judgment under strict regulatory protocols, limiting true end-to-end automation today.'
Adoption barriersclaude-haiku-4-5-202510015/5TSA certification and federal regulations explicitly require human authorization of screening decisions; liability for missed threats creates asymmetric error costs; and security protocols mandate human accountability. These are hard legal and regulatory barriers to full automation.
Adoption barriersclaude-sonnet-55/5This is a government-mandated, licensed security function (e.g., TSA) with strict regulatory requirements that a certified human make final threat calls, creating hard legal and liability barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Computer vision infrastructure, model maintenance, integration with screening workflows, and mandatory human oversight costs are substantial and competitive with or exceed the loaded cost of a full-time TSA screener (~$45–60k/year all-in), especially when error penalties are factored in.
Cost vs. human wageclaude-sonnet-52/5Specialized security screening hardware and certified detection algorithms carry high costs, and human oversight remains mandatory, so cost savings versus a screener's wage are modest, not order-of-magnitude.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI products for baggage screening exist in research and limited pilot deployments, but TSA and major airports continue to rely on human screeners for primary decisions. No mature commercial system has demonstrably replaced human judgment at scale in actual airport operations due to liability and regulatory requirements.
Technical feasibility todayclaude-sonnet-52/5Automated threat detection algorithms (ATR) are deployed in some airport scanners as decision-support, but they augment rather than replace human screeners, and false positive/negative rates require human verification.

Patrol work areas to detect any suspicious items.

16

CI 7–25 · exposure 13 · augmentation 50 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5TSA and airport security operations remain human-centric and heavily regulated; adoption of AI-driven automation for active patrol has been slow despite investments in screening technology. Pilots exist but production replacement of human screeners is minimal due to regulatory constraints and liability concerns.
Sector adoption velocityclaude-sonnet-52/5Security and physical patrol functions are adopting AI slowly, mainly through supplementary camera analytics and sensors, not replacing human patrol officers at scale.
Augmentation potentialclaude-haiku-4-5-202510013/5AI-assisted alerting systems and anomaly detection can help screeners by flagging suspicious items or behavioral patterns for human review, modestly raising screening efficiency. However, the task remains primarily human-driven, and augmentation is limited to supporting existing workflows rather than transforming them.
Augmentation potentialclaude-sonnet-53/5AI-enabled camera analytics, computer vision alerts, and anomaly detection tools can assist screeners by flagging areas of interest, improving situational awareness during patrols.
Task automatabilityclaude-haiku-4-5-202510012/5Computer vision can detect some object anomalies in controlled settings, but real-world patrol requires contextual judgment, threat assessment, and rapid physical response that current AI systems cannot reliably perform end-to-end. The subjective nature of 'suspicious' and the need for human intuition across complex, variable environments prevents meeting the 50% time-saving threshold.
Task automatabilityclaude-sonnet-51/5This requires physical presence, visual scanning of environments, and situational judgment in dynamic real-world spaces; no current AI system can autonomously patrol and detect threats end-to-end without human execution.
Adoption barriersclaude-haiku-4-5-202510014/5Transportation Security Administration (TSA) regulations and Department of Homeland Security protocols require human screeners to perform security functions, with legal liability and human-contact requirements for screening operations. Licensed security personnel are mandated for this role, creating substantial regulatory and authorization barriers to full automation.
Adoption barriersclaude-sonnet-54/5Security screening at airports and similar facilities is regulated (e.g., TSA requirements) and mandates trained, authorized personnel for security-critical detection tasks, creating strong legal and liability barriers.
Cost vs. human wageclaude-haiku-4-5-202510012/5Current AI surveillance hardware, software, integration, and mandatory human oversight costs are comparable to or exceed the loaded hourly wage of security screeners, especially when accounting for the liability of false negatives in high-stakes environments.
Cost vs. human wageclaude-sonnet-51/5Deploying robotic patrol or comprehensive sensor networks with equivalent detection capability would cost far more than a human screener's wage given current hardware and integration costs.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI surveillance and anomaly detection products exist in limited forms (fixed-camera monitoring, baggage scanning), but deployed systems for active patrol work with reliable threat detection remain immature and are not yet in production at scale in transportation security contexts. Error rates remain material and scope is narrow.
Technical feasibility todayclaude-sonnet-51/5While fixed CCTV analytics and stationary scanners exist, there are no deployed autonomous systems that physically patrol security work areas to detect suspicious items in place of a human screener.

Notify supervisors or other appropriate personnel when security breaches occur.

13

CI 0–25 · exposure 13 · augmentation 38 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Transportation security is a laggard sector for AI deployment, with strict regulatory requirements, unionized workforces, and high liability sensitivity. Adoption of autonomous breach notification is negligible in practice.
Sector adoption velocityclaude-sonnet-52/5Aviation security is a highly regulated, safety-critical sector with cautious technology adoption cycles, and automated breach detection/notification systems are still in limited pilot or narrow-scope use.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist by flagging potential anomalies or organizing information for a screener's review, but current systems offer limited meaningful assistance on the core task of identifying and evaluating security breaches in real time at checkpoints.
Augmentation potentialclaude-sonnet-53/5AI-enabled scanning and anomaly detection systems can help screeners identify potential breaches faster, improving situational awareness, though the notification and escalation process remains human-managed.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires detecting a security breach (itself a complex task with high stakes) and then making a human judgment call about notification timing, escalation path, and personnel selection. Current AI cannot reliably identify novel security threats in real time with sufficient confidence to autonomously notify supervisors; human judgment on threat severity is essential.
Task automatabilityclaude-sonnet-52/5While AI can detect anomalies and trigger alerts, the decision to notify supervisors and the judgment about what constitutes a genuine security breach still requires human verification and communication in most current deployments.dynamics.
Adoption barriersclaude-haiku-4-5-202510015/5Transportation security is heavily regulated by TSA and DHS; security personnel are required by law to perform security functions and breach response. Liability for missed breaches or false alarms creates strong legal and organizational barriers to full automation. A trained, authorized human must remain in the decision loop.
Adoption barriersclaude-sonnet-54/5TSA screening involves federal regulatory oversight, legally mandated human authority for security response, and high liability for missed breaches, creating strong institutional and legal barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of implementing and maintaining AI systems to detect breaches, coupled with the requirement for human oversight and the liability risk of false negatives, exceeds the cost of a security screener making a notification decision themselves.
Cost vs. human wageclaude-sonnet-52/5Deploying and maintaining integrated detection-and-alert systems requires significant infrastructure investment, and human screeners still must remain in the loop, so cost savings are modest relative to labor costs.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI system in production at security checkpoints autonomously detects security breaches and independently notifies supervisors without human validation. Breach detection remains primarily human-driven, and notification decisions require human authority and accountability.
Technical feasibility todayclaude-sonnet-52/5Some airport security systems use automated alerting for detected threats, but the full task of assessing and notifying appropriate personnel is still largely human-driven with AI as a supplementary detection tool.

Ask passengers to remove shoes and divest themselves of metal objects prior to walking through metal detectors.

10

CI 0–20 · exposure 8 · augmentation 13 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5No adoption of AI for this task is occurring or feasible because regulatory requirements, security liability, and the mandate for human screeners create structural barriers that prevent any AI substitution pathway.
Sector adoption velocityclaude-sonnet-52/5Aviation security is a highly regulated, safety-critical physical sector with slow AI adoption for frontline screening interactions, despite some tech piloting in scanning equipment.
Augmentation potentialclaude-haiku-4-5-202510011/5AI offers no meaningful assistance to this task. The role is inherently interpersonal communication and real-time compliance monitoring; there is no relevant data entry, document drafting, or decision support that AI could augment in this context.
Augmentation potentialclaude-sonnet-52/5Automated signage, recorded messages, or digital displays can supplement verbal instructions, but they only marginally reduce the screener's ongoing verbal and enforcement role.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires face-to-face instruction and compliance monitoring of individual passengers in a security context. Current AI has no capability to physically interact with or instruct human passengers, and the task depends entirely on human-to-human communication in a real-time, in-person setting.
Task automatabilityclaude-sonnet-52/5This is a brief, repetitive verbal instruction paired with physical passenger management, but it occurs within a live security checkpoint requiring situational awareness and authority that current AI cannot fully replicate end-to-end.atable systems could handle scripted announcements but not the interactive gatekeeping role.'},
Adoption barriersclaude-haiku-4-5-202510015/5This task is protected by multiple hard barriers: TSA regulations mandate human security personnel presence, the task requires legal authority to direct passengers, and federal security screening requirements explicitly require human screeners to perform and take responsibility for security decisions.
Adoption barriersclaude-sonnet-54/5TSA screening involves federal authorization, security regulations, and human authority to enforce compliance and handle non-compliant passengers, creating strong regulatory and liability barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI deployment would be exponentially more expensive than a human screener because it would require humanoid robotics, on-site infrastructure, and sophisticated real-time perception—far exceeding the cost of a trained human performing the task.
Cost vs. human wageclaude-sonnet-52/5Automated signage or recorded announcements are cheap, but full replacement requiring compliance verification and physical presence still requires human staffing, keeping costs comparable to human labor.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI system can perform this task in any meaningful way. The task requires physical presence, real-time auditory/visual communication with passengers, and the ability to respond to diverse passenger behaviors and questions—all outside current AI capabilities.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously manages this passenger-facing instruction-and-compliance task at security checkpoints today; it remains a human screener function.

Inform passengers of how to mail prohibited items to themselves, or confiscate these items.

4

CI 0–9 · exposure 8 · augmentation 25 · importance 3.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5TSA screening is a government function with strict regulatory controls and no observed pilot adoption of autonomous agents. Sector adoption of AI for core screening decisions is negligible.
Sector adoption velocityclaude-sonnet-51/5Airport security is a highly regulated, physical, government-controlled sector with minimal AI-driven task automation deployed for passenger-facing confiscation duties.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist by suggesting regulatory guidance or mailing instructions to display to passengers, but the screener must retain decision authority and interpersonal judgment, limiting meaningful productivity gains.
Augmentation potentialclaude-sonnet-52/5AI could provide screeners with quick-reference guides or scripted information about mailing prohibited items, offering minor assistance, but the core interaction and physical handling remain manual.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires understanding TSA regulations, making discretionary decisions about confiscation vs. mailing options, and interpersonal communication with passengers. Current AI cannot reliably perform the full decision-making loop end-to-end with legal and safety accountability.
Task automatabilityclaude-sonnet-52/5The physical confiscation, communication with anxious travelers, and situational judgment about ambiguous items cannot be done end-to-end by current AI; only some information delivery could be scripted or automated.'
Adoption barriersclaude-haiku-4-5-202510015/5Federal law requires a licensed TSA officer to inspect baggage and make confiscation decisions. Legal liability, regulatory mandate, and the requirement for an authorized human to sign off create hard barriers to automation.
Adoption barriersclaude-sonnet-55/5TSA screening is a federally regulated, licensed function requiring authorized human officers to handle physical searches, confiscation, and chain-of-custody procedures, creating hard legal barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5A TSA screener performs this in-person as part of their workflow; the marginal cost of an AI system to replace this interaction would exceed the cost of the human labor involved, given integration and oversight overhead.
Cost vs. human wageclaude-sonnet-51/5AI cannot physically confiscate items or handle the transaction, so the human screener's cost is unavoidable, making AI substitution cost inapplicable/more expensive when factoring necessary human presence.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI system performs TSA screening decisions independently. Regulatory and liability requirements mandate human screeners make these determinations; no production system automates this task.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs the physical seizure or in-person passenger communication involved; this remains a human task at checkpoints today.

Inform other screeners when baggage should not be opened because it might contain explosives.

1

CI 0–3 · exposure 0 · augmentation 38 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5The transportation security sector has very low adoption of autonomous AI for explosive threat detection because of strict regulatory requirements, security protocols, and the critical safety implications. Screeners remain the human decision-makers in all major airports.
Sector adoption velocityclaude-sonnet-52/5Aviation security is a highly regulated, safety-critical physical-world sector with slow, cautious AI adoption limited mainly to scanning hardware, not decision communication tasks.
Augmentation potentialclaude-haiku-4-5-202510012/5Current AI offers limited assistance—automated detection systems can flag suspicious items for human review, but AI does not meaningfully enhance screener productivity in making the judgment to avoid opening a baggage item due to explosive risk. The task is inherently human-judgment-based with minimal augmentation potential.
Augmentation potentialclaude-sonnet-53/5AI-enhanced imaging and detection algorithms already help flag suspicious items, indirectly supporting screeners' decisions to alert colleagues, though the alerting task itself remains human-driven.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires real-time threat assessment based on baggage screening results and judgment about explosive risk—a domain where current AI has no demonstrated capability to autonomously make life-safety decisions that replace human expertise. The task also requires communication and coordination with team members in a high-stakes security context, which is not automatable end-to-end today.
Task automatabilityclaude-sonnet-51/5This is a real-time, high-stakes verbal/procedural communication decision requiring physical presence, judgment on explosive threat indicators, and immediate coordination with colleagues; no current AI system performs this end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5This task is heavily protected by regulatory barriers: TSA personnel must be certified, and federal law governs baggage screening procedures. Liability for a missed explosive threat is severe, and human screeners are legally required to perform and sign off on these decisions. The human-contact requirement and critical safety nature create hard barriers to substitution.
Adoption barriersclaude-sonnet-55/5TSA screening involves federally mandated, licensed personnel with legal authority and liability for security decisions; explosive threat handling has strict regulatory and safety protocols requiring human authorization.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of deploying AI systems to make autonomous explosive threat assessments, combined with the liability and oversight infrastructure required, far exceeds the cost of a human screener performing this judgment-based safety task.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing this specific communication task, so cost comparison favors the human screener who is already required on-site for security operations.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably performs autonomous explosive threat detection and risk communication in airport security screening. Current TSA screening relies on human screeners interpreting X-ray and other sensor data; AI systems exist as detection aids but not as autonomous decision-makers for not opening baggage due to explosive risk.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously identifies explosive threats in baggage and issues verbal warnings to fellow screeners; explosive detection systems (CT scanners) assist but the interpersonal alerting task itself is not automated by any product.

Contact police directly in cases of urgent security issues, using phones or two-way radios.

1

CI 0–3 · exposure 0 · augmentation 38 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5This task involves direct emergency contact with government agencies and carries legal accountability—sectors like transportation security are highly regulated and unlikely to automate emergency reporting away from trained human operators.
Sector adoption velocityclaude-sonnet-52/5Transportation security is a highly regulated, safety-critical physical environment with slow AI adoption for decision-critical actions, though sensor-assisted detection tools are emerging.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist a screener by flagging anomalies or summarizing threat information to support their decision, but the communication itself must remain human-directed and authorized, limiting augmentation value in this narrow task.
Augmentation potentialclaude-sonnet-53/5AI-enabled detection systems (e.g., threat imaging, anomaly detection) can alert screeners to potential issues faster, helping them decide when to escalate to police, though the contact itself remains manual.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires real-time judgment to determine urgency, direct human-to-human communication with law enforcement, and accountability that cannot be delegated to AI. Current AI cannot reliably assess security threat severity or initiate emergency contact without human authorization.
Task automatabilityclaude-sonnet-51/5This requires real-time human judgment to assess urgent security threats and directly communicate with police; no AI system autonomously makes this call and initiates emergency contact end-to-end today.It's an inherently human decision-action loop.
Adoption barriersclaude-haiku-4-5-202510015/5Legal and regulatory barriers are severe: law enforcement agencies require authorized human personnel to initiate reports; liability for false alarms or missed threats falls on the screener; and emergency dispatch systems are designed only to accept calls from identified individuals responsible for their accuracy.
Adoption barriersclaude-sonnet-55/5Security screening involves legal authority, liability, and regulatory requirements (e.g., TSA mandates) that require a certified human screener to initiate contact with law enforcement in urgent situations.
Cost vs. human wageclaude-haiku-4-5-202510011/5A screener making an emergency call takes seconds; the integration and liability costs of AI attempting this task independently would far exceed the trivial cost of human-initiated contact, making AI economically inferior.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this exact task, so cost comparison favors the human who is already embedded in the security workflow and legally responsible.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product autonomously contacts emergency services or police based on security assessments. Legal liability, authentication, and the requirement for human authority over emergency dispatch make this task not feasible for AI to perform independently in any production system.
Technical feasibility todayclaude-sonnet-51/5No deployed product independently detects security threats and contacts police on behalf of a screener; this remains a human responsibility in aviation/transit security operations.

Search carry-on or checked baggage by hand when it is suspected to contain prohibited items such as weapons.

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-4-5-202510011/5TSA operates in a highly regulated, centralized government environment with slow technology adoption. Baggage screening remains human-centric despite decades of investment in supporting technology; there is no measurable shift toward automation of the actual hand-search task in any airport today.
Sector adoption velocityclaude-sonnet-51/5Aviation security screening is a highly regulated, physically-oriented government function with minimal AI adoption for hands-on physical search tasks.
Augmentation potentialclaude-haiku-4-5-202510012/5Current AI (CT scanning, threat-detection algorithms, imaging) helps screeners decide which bags to search and prioritize, but does not augment the physical search act itself. Once a screener decides to hand-search, AI provides minimal real-time assistance during the manual inspection process.
Augmentation potentialclaude-sonnet-52/5AI-enhanced imaging (e.g., CT scanners with automated threat detection) can flag suspicious items to guide where screeners should search, but the physical search itself remains unassisted by AI.
Task automatabilityclaude-haiku-4-5-202510011/5Hands-on physical search of baggage for prohibited items requires tactile discrimination, spatial reasoning in confined spaces, and real-time decision-making about ambiguous objects. Current AI cannot perform this end-to-end; vision systems alone cannot reliably detect contraband by external appearance, and manipulation of luggage contents requires dexterous robotics that cannot yet operate safely in this unstructured environment.
Task automatabilityclaude-sonnet-51/5Manual hand-searching of physical bags requires physical dexterity, judgment, and tactile inspection that current AI systems and robots cannot perform end-to-end.'
Adoption barriersclaude-haiku-4-5-202510015/5TSA is a federal law-enforcement function with strict regulatory authority (49 CFR Part 1540). Federal regulation requires that screening be conducted by authorized personnel; liability for missed weapons or explosives falls on the organization, and no regulatory pathway currently permits autonomous or unsupervised machine performance of this duty. A human screener must legally conduct or directly oversee the search.
Adoption barriersclaude-sonnet-55/5This task is federally mandated (TSA) to be performed by trained, authorized security personnel under strict regulatory and legal frameworks, making substitution by AI legally and physically infeasible.
Cost vs. human wageclaude-haiku-4-5-202510011/5Building a robotic system to reliably search baggage, including manipulation, object recognition, and decision logic, would cost orders of magnitude more per search than the loaded wage of a TSA screener (~$45k–$55k annually, or ~$25–$30/hour burdened). AI automation here is economically infeasible.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute for physical manual search, so AI cost comparison is not applicable; the human remains the only cost-effective option.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI system performs baggage hand-search autonomously or at production scale. While X-ray and threat-detection algorithms assist human screeners, the actual manual search—opening luggage, feeling contents, identifying prohibited items—remains entirely human-performed in all operational settings.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs physical hand-searches of luggage; this remains an inherently manual, physical task performed by human screeners.

Perform pat-down or hand-held wand searches of passengers who have triggered machine alarms, who are unable to pass through metal detectors, or who have been randomly identified for such searches.

0

CI 0–0 · exposure 0 · augmentation 25 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Zero adoption of AI or automation in this task across the aviation security sector. TSA operations remain highly regulated and conservative; any change would require federal rulemaking and pilot programs that have not materialized.
Sector adoption velocityclaude-sonnet-51/5Physical security screening is a highly regulated, low-digitization physical task with essentially no AI/robotic adoption for the hands-on search component itself.
Augmentation potentialclaude-haiku-4-5-202510012/5Limited augmentation exists today. Wand sensors provide data readouts, but AI does not materially enhance the screener's ability to perform or interpret pat-downs beyond existing handheld metal detector feedback, which is already standard.
Augmentation potentialclaude-sonnet-52/5AI-enhanced imaging and detection systems can flag areas of concern to guide where a screener focuses a pat-down, but this offers only marginal assistance to the physical search task itself.
Task automatabilityclaude-haiku-4-5-202510011/5Physical pat-down and wand searches require direct bodily contact and tactile sensing that current AI and robotic systems cannot perform safely or legally. The task fundamentally involves human judgment in interpreting sensor readings and making security decisions with intimate personal contact, which remains beyond automation feasibility.
Task automatabilityclaude-sonnet-51/5This requires physical human contact and manual dexterity to search a person's body, which current AI systems cannot physically perform at all; robotics for this application does not exist in deployable form.
Adoption barriersclaude-haiku-4-5-202510015/5Hard regulatory and legal barriers exist: TSA regulations explicitly require human screeners to perform physical searches, and there are strict protocols around bodily contact, consent, and liability. Additionally, passengers have a right to refuse searches or request same-gender personnel, requirements that embed human discretion and legal authority into the task.
Adoption barriersclaude-sonnet-55/5Security screening pat-downs are legally mandated to be performed by authorized, trained human personnel under strict government regulations (e.g., TSA), with significant liability, privacy, and legal authority requirements precluding automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5Even if conceptually automatable, the cost of purpose-built humanoid robotics capable of performing tactile security searches would vastly exceed the loaded wage of a TSA screener ($40–60k annually), by orders of magnitude.
Cost vs. human wageclaude-sonnet-51/5There is no AI system capable of performing this physical task, so no cost comparison is meaningful other than infinite AI cost relative to a human screener.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI or robotic system performs actual pat-down or hand-held wand searches in TSA or airport operations today. While metal detectors and some imaging systems exist, the hands-on search component remains exclusively human-staffed and has no production AI equivalent.
Technical feasibility todayclaude-sonnet-51/5No product exists that performs physical pat-downs or wand searches autonomously; this remains entirely a human physical task with no research-stage robotic alternative in security contexts.

Follow those who breach security until police or other security personnel arrive to apprehend them.

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-4-5-202510011/5This task operates in a highly regulated, human-centric security environment with explicit legal requirements for human judgment and accountability. Adoption of autonomous systems for this function remains negligible across the transportation security sector.
Sector adoption velocityclaude-sonnet-51/5Physical security enforcement functions in transportation security are low-digitization, human-presence-dependent tasks with essentially no AI agent deployment for the pursuit/apprehension component.
Augmentation potentialclaude-haiku-4-5-202510012/5AI can marginally assist by detecting anomalies or alerting screeners to potential breaches, but cannot meaningfully augment the core function of physically following and monitoring a subject until law enforcement arrival.
Augmentation potentialclaude-sonnet-52/5AI-powered camera tracking or surveillance systems could help alert or guide a screener to a breach location, but they don't materially assist the act of physically following and apprehending someone.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires real-time physical presence, situational judgment, and direct intervention in dynamic security scenarios that current AI cannot perform. No current system can autonomously follow, monitor, and engage with a human actor in an uncontrolled physical environment.
Task automatabilityclaude-sonnet-51/5This requires physical pursuit, real-time judgment, and human presence to track and contain an individual; no AI system can perform physical apprehension-support tasks today.
Adoption barriersclaude-haiku-4-5-202510015/5This task is legally and operationally restricted to licensed security personnel and law enforcement. TSA and airport security have strict regulatory requirements for who can pursue and detain individuals, creating hard legal barriers to automation.
Adoption barriersclaude-sonnet-55/5This involves physical security enforcement, potential use of force, and legal authority to detain/apprehend, which requires an authorized human officer and carries high liability and regulatory constraints.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of deploying autonomous surveillance and physical security systems, including liability infrastructure and oversight, far exceeds the cost of deploying trained human security personnel for this critical function.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing this physical task, so cost comparison is moot—human presence is mandatory and cannot be replaced by cheaper inference costs.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product can autonomously perform physical surveillance and apprehension support in real airport security settings. This requires embodied autonomous agents that do not exist in production.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs physical following/tracking of suspects in place of a human security officer; this remains entirely human-executed.

Close entry areas following security breaches or reopen areas after receiving notification that the airport is secure.

0

CI 0–0 · exposure 0 · augmentation 25 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Airport security is a regulated sector with strict federal oversight and resistance to removing human decision-makers from critical security functions. Adoption of automation in this domain is minimal and unlikely to accelerate.
Sector adoption velocityclaude-sonnet-51/5Aviation security screening is a highly regulated, physically-grounded government/quasi-government function with minimal AI-driven automation of operational decision-making of this kind.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could theoretically assist by monitoring sensors and alerting screeners to conditions requiring re-closure, but the core task—judging when to reopen and physically managing barriers—remains human-centric with limited augmentation opportunity.
Augmentation potentialclaude-sonnet-52/5AI could assist with notification systems, logging, or coordinating communications about breach status, but offers minimal help with the core physical and authorization-based actions.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires physical gate/door closure operations and real-time security assessment judgment tied to human safety decisions. Current AI cannot perform the physical component reliably, and the authorization to declare an area secure must come from human security personnel, not autonomous systems.
Task automatabilityclaude-sonnet-51/5This requires physical presence to close/reopen barriers, coordinate with on-site personnel, and make judgment calls tied to security clearance decisions; no AI system can perform this end-to-end today.
Adoption barriersclaude-haiku-4-5-202510015/5Airport security operations are heavily regulated by TSA and federal law; a human security screener must legally authorize and validate security decisions. Liability for security failures rests on human judgment, creating a hard legal barrier to full automation.
Adoption barriersclaude-sonnet-55/5This is a federally regulated security function requiring authorized, credentialed personnel (e.g., TSA) to make and execute the decision, with high liability for errors—hard legal and physical barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5The task includes minimal, quick physical actions (closing doors/barriers) combined with human judgment and accountability. Automating only the physical component while still requiring a human to authorize decisions offers no cost advantage and adds system complexity.
Cost vs. human wageclaude-sonnet-51/5AI has no viable way to perform this physical, authorization-dependent action, so there is no cost-competitive AI substitute; human labor is the only option.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI system today reliably decides when to reopen security zones or performs the physical blocking of entry areas. This is inherently a human-supervised security function with strict protocols and liability; research-stage automation only.
Technical feasibility todayclaude-sonnet-51/5No deployed products physically manage checkpoint closures/reopenings or hold authority to declare an area secure; this remains a human-executed, procedural task.

Challenge suspicious people, requesting their badges and asking what their business is in a particular areas.

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-4-5-202510011/5Transportation security remains one of the most regulated and human-centric sectors, with little evidence of AI-driven replacement of core screening functions. Adoption of supporting technologies (cameras, analytics) is slow and always paired with human decision-making.
Sector adoption velocityclaude-sonnet-51/5Physical security and law enforcement-adjacent screening roles show minimal AI-driven displacement; adoption in this specific interpersonal confrontation task is essentially absent.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist with flag detection or anomaly alerts in video feeds, but the core task of challenging people and making judgment calls remains almost entirely human-dependent. Limited potential for meaningful productivity enhancement given the irreducible role of human authority.
Augmentation potentialclaude-sonnet-52/5AI-based surveillance, facial recognition, or anomaly detection can flag suspicious individuals for human follow-up, but it doesn't materially transform the confrontation task itself.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires real-time human judgment to assess suspicion, read social cues, and make security decisions based on context—capabilities far beyond current AI. No AI system can reliably perform the gatekeeping function of challenging individuals and making threat assessments in dynamic physical environments.
Task automatabilityclaude-sonnet-51/5This requires real-time physical presence, judgment about human behavior/intent, and direct interpersonal confrontation that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Federal law and TSA regulations require that transportation security be performed by authorized human personnel with legal standing to challenge and detain. The human-contact requirement and mandatory licensing/authorization create hard legal barriers to full automation.
Adoption barriersclaude-sonnet-55/5Security screening authority, use of force/detention implications, and legal authorization to challenge individuals require a human officer with specific credentials and liability accountability.
Cost vs. human wageclaude-haiku-4-5-202510011/5An AI system capable of autonomous physical screening (with cameras, audio, decision logic, and escalation protocols) would require substantial infrastructure investment, whereas human screeners are a direct wage expense. The all-in cost of a deployed autonomous system would exceed human labor costs.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this physical, confrontational task, so cost comparison favors the human by default.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product performs this task independently. Physical security screening requires human presence, legal authority, and discretionary judgment that current AI systems cannot exercise in real-world airport or facility settings.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously challenges suspicious individuals and verifies credentials in physical security contexts; this remains a human security function.

Confiscate dangerous items and hazardous materials found in opened bags and turn them over to airlines for disposal.

0

CI 0–0 · exposure 0 · augmentation 25 · importance 3.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Transportation security is a federally regulated, high-compliance sector where automation of core screening duties faces legal and safety-first culture. No evidence of AI-driven confiscation deployment; human screeners remain legally mandated.
Sector adoption velocityclaude-sonnet-51/5Airport security screening is a highly regulated, physically-oriented, low-digitization task with essentially no AI-driven displacement of the physical confiscation step.
Augmentation potentialclaude-haiku-4-5-202510012/5AI vision systems can assist screeners in identifying potentially dangerous items more quickly, but the actual confiscation, handling, and disposal remains a human task. Augmentation is limited to detection support, not the core confiscation workflow.
Augmentation potentialclaude-sonnet-52/5AI-based scanning (e.g., CT-based baggage screening) can help detect items to flag for human inspection, but this specific act of confiscating and transferring items to airlines is not meaningfully augmented by AI.
Task automatabilityclaude-haiku-4-5-202510011/5Confiscating items and handling hazardous materials requires physical manipulation, real-time risk judgment of novel items, and compliance with TSA protocols that demand human authority and liability. Current AI cannot reliably identify all dangerous items, make context-dependent decisions about confiscation, or physically interact with materials.
Task automatabilityclaude-sonnet-51/5Physically retrieving and confiscating items from bags and handing them to airline staff requires manipulation and judgment calls in a real-world environment that current AI systems cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5TSA screeners operate under federal aviation security law; confiscation and disposal authority is legally vested in licensed federal employees. Regulatory mandate, liability for improper handling of hazardous materials, and the requirement for human authority make this task protected by hard legal barriers.
Adoption barriersclaude-sonnet-55/5TSA screening requires federally authorized, trained personnel with legal authority to confiscate items, and chain-of-custody/liability requirements make this a hard regulatory and legal barrier.
Cost vs. human wageclaude-haiku-4-5-202510011/5Even if partial automation of item identification were feasible, the liability, physical manipulation, and regulatory compliance costs of automating the confiscation step itself would exceed the cost of human screening labor, which is relatively low-wage.
Cost vs. human wageclaude-sonnet-51/5There is no AI system that performs this physical handling task, so no viable cost comparison exists; a human must be paid to do it.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI system performs end-to-end confiscation and disposal coordination in production TSA checkpoints. While computer vision can assist in detection, the actual seizure, handling, documentation, and transfer to airlines remains a human-performed, liability-bearing task.
Technical feasibility todayclaude-sonnet-51/5No deployed product physically confiscates items and hands them to airlines; this remains entirely a human physical and procedural task.

Related occupations — Protective Service

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.