Postal Service Clerks
43-5051.00Median wage $62,130/yr73,720 employed (US)Rank #81 of 923 scored · top 9% by substitution
Perform any combination of tasks in a United States Postal Service (USPS) post office, such as receive letters and parcels; sell postage and revenue stamps, postal cards, and stamped envelopes; fill out and sell money orders; place mail in pigeon holes of mail rack or in bags; and examine mail for correct postage. Includes postal service clerks employed by USPS contractors.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution50
Exposure49
Augmentation46
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
19 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
26%
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%51
panel mean rating 3.0/5 → substitution pressure 51/100
Technical feasibility todayw 20%47
panel mean rating 2.9/5 → substitution pressure 47/100
Cost vs. human wagew 15%54
panel mean rating 3.1/5 → substitution pressure 54/100
Adoption barriersw 20%inverted — strong barriers lower the score53
panel mean rating 2.9/5 (barrier strength) → substitution pressure 53/100
Sector adoption velocityw 10%38
panel mean rating 2.5/5 → substitution pressure 38/100
Task breakdown (19 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.
Weigh letters and parcels, compute mailing costs based on type, weight, and destination, and affix correct postage.
94CI 92–95 · exposure 100 · augmentation 75 · importance 4.6/5 · click for rater detail
Weigh letters and parcels, compute mailing costs based on type, weight, and destination, and affix correct postage.
94| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 5/5 | Postal services and package-handling retailers have already widely adopted automated weighing, metering, and postage systems in production. This represents decades-long, mature, deep adoption across the entire postal and shipping sector. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Self-service kiosks and online postage tools have been widely deployed for over a decade across postal and logistics sectors, though full clerk replacement lags due to counter service needs. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | Automated systems assist postal clerks by instantly computing rates, reducing calculation errors, and streamlining labeling, significantly raising throughput and accuracy. Even where humans remain in the loop for oversight, AI augmentation is highly productive. |
| Augmentation potential | claude-sonnet-5 | 4/5 | Where clerks remain, automated scales and rating software substantially speed up their per-transaction throughput and reduce pricing errors. |
| Task automatability | claude-haiku-4-5-20251001 | 5/5 | This task is entirely rule-based and data-driven: weight measurement (via scales connected to systems), cost lookup from postal rate tables, and label/postage generation are all fully automatable. Current systems can perform end-to-end processing with >50% time savings and higher accuracy than manual clerks. |
| Task automatability | claude-sonnet-5 | 5/5 | Weighing and rate computation is a deterministic, rule-based task already fully automated by postal scales, kiosks, and shipping software that compute postage instantly from weight, dimensions, and destination. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | No licensing or legal requirement mandates human performance of this task. Minimal barriers exist; some customer-service contexts may prefer human interaction, but the core task itself can be fully automated without regulatory constraint. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement exists for weighing/rating mail; some friction remains from customers preferring staffed counters and postal union/staffing rules, but the calculation itself faces no regulatory or professional barrier. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Automated postal systems (scale + software + label printer) have negligible marginal cost per item compared to clerk labor, especially considering integration with existing infrastructure. The cost per task is orders of magnitude lower than human wage for equivalent throughput. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Automated scales and rate-calculation software cost pennies per transaction versus a loaded clerk wage, an order-of-magnitude or greater cost advantage. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 5/5 | Production systems for postal automation (including integrated weighing, rate computation, and label generation) have been deployed at postal facilities and retail locations globally for years. Major postal carriers use mature, reliable end-to-end automation for this task. |
| Technical feasibility today | claude-sonnet-5 | 5/5 | Self-service postal kiosks, USPS.com/Click-N-Ship, and third-party shipping platforms (Pitney Bowes, Stamps.com, Shippo) reliably perform this exact function at scale in production today. |
Sort incoming and outgoing mail, according to type and destination, by hand or by operating electronic mail-sorting and scanning devices.
84CI 74–95 · exposure 87 · augmentation 38 · importance 4.4/5 · click for rater detail
Sort incoming and outgoing mail, according to type and destination, by hand or by operating electronic mail-sorting and scanning devices.
84| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 5/5 | Major postal services worldwide (USPS, Royal Mail, etc.) have already extensively automated mail sorting over decades; automation is standard practice in the industry, with most volume-processing facilities operating high-speed sorting machines. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Automated mail sorting has already been widely deployed across major postal operators for many years, representing mature, deep adoption in this specific subtask. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI offers minimal augmentation for human mail sorters because the task is already heavily mechanized and algorithmic; the role is primarily to feed machines or handle exceptions, leaving little room for AI to meaningfully enhance human productivity in the sorting process itself. |
| Augmentation potential | claude-sonnet-5 | 3/5 | Where manual sorting persists (irregular parcels, exceptions), scanning devices and handheld barcode readers assist clerks in verifying destinations and speeding handling. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Mail sorting by type and destination is largely algorithmic and mechanical; postal services already deploy automated sorting machines at scale for substantial portions of this workload. However, complex edge cases (illegible addresses, non-standard mail) still require human intervention, preventing a full 5-rating. |
| Task automatability | claude-sonnet-5 | 5/5 | Automated mail-sorting equipment using OCR and barcode scanning already performs this exact task at high speed and volume, meeting or exceeding the 50% time-saving threshold versus manual sorting. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | While no explicit licensing requirement protects human mail sorting, organizational infrastructure, union agreements in some jurisdictions, and the need for human oversight of exceptions create moderate adoption friction. Capital equipment requirements also slow transition in small facilities. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement blocks automation; postal unions and transition costs create some organizational friction, but the technology itself faces few regulatory or liability barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Automated mail sorting systems are already deployed because they are substantially cheaper than human labor at scale; once capital costs are amortized, the per-piece cost of machine sorting is orders of magnitude lower than paying clerks per item. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Automated sorting machines process thousands of pieces per hour at a fraction of the labor cost per item compared to manual sorting clerks. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Postal automation equipment is mature and in widespread production use globally; optical scanning and mechanized sorting are proven technologies deployed by major postal services. Minor limitations remain in handling unusual mail formats or damage, but the core task is reliably performed by existing systems. |
| Technical feasibility today | claude-sonnet-5 | 5/5 | Postal services worldwide (USPS, Royal Mail, etc.) have used automated sorting machines with optical character recognition in production for decades at massive scale. |
Answer questions regarding mail regulations and procedures, postage rates, and post office boxes.
82CI 76–87 · exposure 83 · augmentation 75 · importance 4.3/5 · click for rater detail
Answer questions regarding mail regulations and procedures, postage rates, and post office boxes.
82| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | USPS and postal-adjacent organizations have already deployed chatbots and AI-driven customer service systems. The information-heavy nature and digital infrastructure of postal services enable rapid adoption relative to other government sectors. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Postal services are government-affiliated and traditionally slower adopters, but self-service kiosks and website chatbots for basic inquiries have already been rolled out, indicating moderate progress. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can assist clerks by instantly retrieving accurate regulations and rates, reducing lookup time and human error. A clerk can use AI as a knowledge base while handling nuanced follow-up questions or exceptions, significantly raising throughput. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-powered knowledge bases and chat assistants significantly speed up clerks' ability to look up regulations and rates, improving accuracy and response time even when a human remains involved. |
| Task automatability | claude-haiku-4-5-20251001 | 5/5 | This task is highly structured information retrieval and explanation. Current AI systems can reliably answer questions about postal regulations, procedures, rates, and box policies using fine-tuned models or retrieval-augmented generation over USPS documentation, achieving >50% time savings over human lookup. |
| Task automatability | claude-sonnet-5 | 4/5 | Answering questions about mail regulations, postage rates, and PO box procedures is largely factual Q&A that current chatbots and knowledge-base-driven AI can handle with high accuracy, though edge cases still need human escalation.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | No licensing barrier exists for AI to answer postal questions, and customer preference for human service is weak in this context. Some organizational friction around integration and the desire to maintain human touchpoints for complex issues present modest barriers. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement exists for answering these informational questions, though some customers still prefer human interaction and complex regulatory nuances may require staff correction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI inference cost per query is negligible (cents), while a postal clerk's fully loaded cost per customer interaction is $20–40. The cost ratio favors AI by an order of magnitude. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | A chatbot or automated FAQ system answering routine postage/regulation questions costs a small fraction of a clerk's wage per interaction, especially at scale. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed chatbots and AI customer service systems already handle postal inquiries in production at scale (including USPS-adjacent services). While edge cases and policy updates require monitoring, mainstream questions about rates and procedures are solved reliably by existing systems. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | USPS and other postal services already deploy chatbots and virtual assistants (e.g., USPS's website chatbot) that answer these exact types of questions in production, though complex or unusual queries still route to humans. |
Register, certify, and insure letters and parcels.
72CI 65–79 · exposure 75 · augmentation 63 · importance 4.5/5 · click for rater detail
Register, certify, and insure letters and parcels.
72| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Major postal services (USPS, Royal Mail, Deutsche Post) have already deployed automated registration and certification systems in high-volume facilities. Adoption is accelerating as logistics companies integrate AI-driven sorting and tracking; this is a digital, process-heavy sector with strong efficiency incentives. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Postal services are historically slow-moving, unionized, and physically-oriented organizations; kiosk adoption has been steady but gradual rather than fast and deep. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI significantly enhances clerk productivity by auto-populating forms, suggesting appropriate insurance levels, flagging compliance issues, and generating documents in real-time. Clerks using AI-assisted systems process substantially more items per hour while maintaining accuracy and handling exceptions. |
| Augmentation potential | claude-sonnet-5 | 3/5 | Software systems assist clerks with fee calculation, label printing, and tracking number generation, meaningfully speeding up the transaction even when a human remains involved for handling and customer service. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Current AI systems can reliably capture package/letter metadata, validate against shipping rules, and generate certifications and insurance documents with minimal human intervention. The task involves structured data entry, rule-based validation, and document generation—all well within AI capabilities—though human verification of high-value items may remain necessary for full end-to-end automation. |
| Task automatability | claude-sonnet-5 | 4/5 | Registering, certifying, and insuring mail is largely a structured data-entry and transaction process (weighing, classifying, fee calculation, printing labels/receipts) that can be automated via kiosks, scales, and software with minimal human intervention.dimensional physical handling of parcels still requires some human/automated hardware interaction. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Postal services are regulated but do not legally require a human clerk to perform registration itself; however, organizational inertia, customer expectations (some prefer human interaction), and the need for in-person identity verification for some certified services create moderate friction against full automation. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement for a human to perform this task, but there is some regulatory oversight around mail security, chain-of-custody for insured/certified items, and customer preference for human assistance with disputes or complex packages. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI-driven automation (barcode scanning, OCR, rule-based certification, insurance calculation) costs a fraction of a penny per transaction, while a postal clerk's loaded wage (salary + benefits) is $25–40/hour. The cost per registered item is orders of magnitude lower with automation. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated kiosks and software cost far less per transaction than a staffed clerk once deployed, though machine maintenance and physical infrastructure add some ongoing cost compared to pure software automation. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed postal and logistics software already automate significant portions of registration, certification, and insurance processing. Computer vision systems can read addresses and barcodes; rule engines validate shipping rules; digital signatures and certificate generation are mature. Production systems handle millions of transactions daily, though some edge cases and fraud detection still require human oversight. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Self-service postal kiosks and automated parcel machines already handle registration, insurance selection, and certification printing in production at USPS and other postal services worldwide, though staffed counters remain common for edge cases. |
Order retail items and other supplies for office use.
72CI 65–79 · exposure 70 · augmentation 75 · importance 3.7/5 · click for rater detail
Order retail items and other supplies for office use.
72| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | E-procurement and supply-chain automation are advancing rapidly in government and corporate sectors; the USPS and similar large organizations have already deployed digital ordering systems, and further AI integration is progressing at a fast pace. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Postal services are traditionally slow-moving, unionized, government-affiliated organizations with lower digitization and AI adoption rates compared to private-sector logistics or retail firms. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can substantially assist clerks by instantly surfacing optimal suppliers, comparing prices, predicting quantities, and auto-populating orders—freeing humans to focus on exception handling, cost optimization, and relationship management rather than routine order entry. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-based inventory and procurement tools can meaningfully assist clerks by flagging low stock, suggesting order quantities, and automating routine reordering while a human retains final approval. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Ordering supplies is a highly structured, rule-based task involving catalog lookup, quantity selection, and form completion. Current AI agents can reliably handle vendor research, price comparison, SKU matching, and purchase order generation, delivering at least 50% time savings for routine orders with minimal human oversight. |
| Task automatability | claude-sonnet-5 | 4/5 | Ordering supplies is a structured, repetitive procurement task (checking inventory, placing reorders) that off-the-shelf AI/automation tools (procurement software, agentic assistants) can handle with high time savings, though occasional human approval remains typical. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few legal or regulatory barriers exist to automating routine supply ordering; internal approval workflows and vendor relationships provide modest friction, but no licensing requirement or human-contact mandate constrains substitution. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing or legal requirement mandates a human clerk perform this task; the main friction is organizational habit and integration with existing government procurement systems. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Automated ordering via AI agents costs pennies per transaction in inference and integration, vastly cheaper than human time spent researching vendors, comparing prices, filling forms, and managing approvals—easily an order of magnitude savings. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated procurement systems cost far less per transaction than staff time spent manually tracking and ordering supplies, especially at scale, though integration with existing postal systems adds some overhead. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Procurement automation is mature and widely deployed in enterprise settings. E-procurement platforms and AI-assisted ordering systems are in production use across organizations; errors are low for standard items, though some complex or non-standard orders may still require human review. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Procurement and inventory management software with automated reordering exists and is deployed widely in retail/office settings, but postal service workflows often still rely on manual or semi-manual ordering processes tied to legacy systems. |
Keep money drawers in order, and record and balance daily transactions.
59CI 48–70 · exposure 62 · augmentation 75 · importance 4.6/5 · click for rater detail
Keep money drawers in order, and record and balance daily transactions.
59| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Post offices and financial institutions have been automating transaction processing and cash management for decades; modern postal automation systems and banking platforms are mature and widely deployed, showing deep adoption in the sector. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Postal service is a low-digitization, unionized, government sector with slow technology adoption relative to private financial services. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI-assisted reconciliation, anomaly detection, and real-time balance alerts substantially amplify a clerk's ability to maintain accuracy and catch discrepancies faster than manual review alone, keeping the human in oversight while raising throughput. |
| Augmentation potential | claude-sonnet-5 | 4/5 | Modern point-of-sale and accounting software already significantly assists clerks in tracking, recording, and reconciling daily transactions, reducing manual calculation effort. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Recording and balancing daily transactions is highly automatable via accounting software and OCR systems; money drawer management (physical handling and verification) has a non-automatable component, but modern POS and cash-handling systems can automate 60–80% of the workflow, meeting the ≥50% time-saving threshold. |
| Task automatability | claude-sonnet-5 | 3/5 | Recording and balancing transactions is highly structured and could largely be automated via POS/accounting systems, but physical cash drawer handling and reconciliation still requires human/physical steps. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Internal controls, audit requirements, and regulatory compliance (federal banking/postal rules) require human sign-off on balances and discrepancies, creating oversight obligations; however, these are procedural rather than absolute legal bars to automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Handling government funds involves accountability, audit trails, and some regulatory/organizational controls, but no formal licensing requirement mandates a human specifically for balancing a drawer. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | Accounting and transaction-management software cost per task is a small fraction of clerk wages; automated cash-handling systems (sorters, counters) have upfront capital costs but amortize to far below human labor cost per transaction. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | Software-based transaction recording is cheap, but the task also involves physical drawer management and oversight, keeping overall cost comparable to human labor rather than dramatically cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed products (enterprise accounting software, automated cash-handling machines, bank reconciliation tools) reliably perform transaction recording and balancing in production; physical cash drawer oversight remains partly manual but well-supported by automated exception reporting. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | POS and cash-management software already automate much of the recording/balancing function in retail and postal settings, though integration with legacy USPS systems and physical cash handling limits full automation today. |
Sell and collect payment for products such as stamps, prepaid mail envelopes, and money orders.
57CI 40–75 · exposure 55 · augmentation 38 · importance 4.7/5 · click for rater detail
Sell and collect payment for products such as stamps, prepaid mail envelopes, and money orders.
57| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | USPS has deployed self-service kiosks nationwide and expanded online sales; private shipping and postal operators widely use automated systems. Adoption is measurable and accelerating in the information/logistics sector. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Postal services are slow-moving public sector organizations with historically low digitization and automation adoption rates compared to private sector retail or finance. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist clerks by suggesting products based on customer needs, verifying postage rates, or processing refunds, improving productivity without full automation. The augmentation is useful but modest compared to the task's simplicity. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI offers limited direct assistance to a clerk performing point-of-sale transactions, though inventory and payment systems provide minor digital support. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | A self-service kiosk or online system can handle ~90% of this task (product selection, payment processing, transaction recording) with minimal time investment compared to a clerk interaction. Only edge cases (damaged products, refunds, customer disputes) require human intervention, meeting the ≥50% time-saving threshold. |
| Task automatability | claude-sonnet-5 | 2/5 | Physical handling of cash, products, and customers at a counter cannot be replaced end-to-end by current AI; self-service kiosks automate some of this but are hardware-based automation, not AI-driven task substitution.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few hard legal barriers exist: stamps and envelopes are not regulated products requiring human sign-off. Some USPS locations maintain clerk roles for customer service and foot traffic, but regulatory or liability constraints do not prevent self-service alternatives. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement for selling postage, but cash handling, fraud prevention for money orders, and government employment structures create moderate organizational friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | A fully automated kiosk or online system costs cents per transaction in amortized infrastructure and payment processing, versus a clerk wage (fully loaded ~$25–35/hour) for 5–10 minute interactions; the cost ratio heavily favors automation. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Kiosk/automation hardware has significant upfront capital and maintenance costs, and human clerks remain necessary for exceptions, so all-in cost savings are moderate rather than order-of-magnitude. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed products exist and work reliably: USPS kiosks, online stamp/envelope sales, and payment-processing systems are in wide production use. Error rates are low for standard transactions, though some integration and oversight overhead remains. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Self-service postal kiosks and vending machines already sell stamps and process payments reliably in many locations, though they don't fully replace clerks for money orders and complex transactions. |
Feed mail into postage canceling devices or hand stamp mail to cancel postage.
57CI 18–97 · exposure 58 · augmentation 13 · importance 3.8/5 · click for rater detail
Feed mail into postage canceling devices or hand stamp mail to cancel postage.
57| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Postal services are laggard adopters of advanced automation due to union protections, decentralized facility operations, and the physical, handling-intensive nature of mail work. Adoption velocity for AI-driven automation of clerk tasks remains minimal in practice. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Postal services have already broadly adopted automated canceling equipment, though some manual hand-stamping persists for special/philatelic mail or backup processes. |
| Augmentation potential | claude-haiku-4-5-20251001 | 1/5 | This task is primarily mechanical sorting and stamping with little scope for AI to meaningfully assist a human operator. The work does not involve analysis, judgment, or information processing that AI tools could enhance. |
| Augmentation potential | claude-sonnet-5 | 2/5 | Since the task is largely already mechanized, there's little additional AI augmentation benefit for a human performing manual hand-stamping tasks. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Feeding mail into physical canceling devices requires mechanical handling and alignment of variable-sized pieces in 3D space—a task current general-purpose robotics struggle with at scale. Hand-stamping is even less automatable due to precise placement and pressure requirements. Narrow automation of high-volume standardized mail is conceivable but does not achieve the 50% time-saving threshold for the full task diversity postal clerks encounter. |
| Task automatability | claude-sonnet-5 | 5/5 | Feeding mail into canceling machines is already highly mechanized and repetitive; automated equipment (not even 'AI' in the modern sense) can fully handle this with high throughput and consistent quality. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Postal service is heavily unionized (NALC, APWU) with strong labor agreements protecting clerk positions; USPS operates under federal oversight and statutory service obligations that create organizational friction against displacement. Regulatory and contractual barriers are substantial. |
| Adoption barriers | claude-sonnet-5 | 1/5 | No licensing or legal requirement mandates human hand-canceling; postal operators have full discretion to mechanize this task. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Specialized robotics for mail handling remain expensive to deploy and maintain relative to low-wage postal clerk labor, particularly when accounting for integration, recalibration for mail variability, and downtime costs. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Automated canceling machines process vastly more mail per hour than manual stamping at a fraction of the per-unit labor cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI systems reliably perform physical mail feeding or hand-stamping operations in postal facilities today. This requires specialized mechanical systems beyond current off-the-shelf automation, and no production systems demonstrate this at postal-facility scale. |
| Technical feasibility today | claude-sonnet-5 | 5/5 | Postal services worldwide have used automated canceling and sorting machinery in production for decades, reliably processing millions of pieces of mail. |
Check mail to ensure correct postage and that packages and letters are in proper condition for mailing.
52CI 50–55 · exposure 50 · augmentation 63 · importance 4.5/5 · click for rater detail
Check mail to ensure correct postage and that packages and letters are in proper condition for mailing.
52| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | USPS and parcel carriers are actively piloting automated sorting and inspection, but full deployment remains uneven; many smaller facilities and routes still rely primarily on manual checking. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Postal and logistics sectors have adopted automated sorting and scanning technology broadly at the back-end, but customer-facing counter tasks remain a slower-adopting segment due to public-sector pace and physical infrastructure needs. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can assist clerks by automatically flagging suspicious or damaged items and highlighting postage issues, allowing human staff to focus their attention on exception handling and marginal cases, substantially raising review throughput. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-powered scales, barcode scanners, and postage verification tools already assist clerks by speeding up checks and flagging issues, meaningfully improving efficiency while humans still handle exceptions and customer interaction. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | Current computer vision systems can detect postage presence and flag obvious damage (tears, stains), achieving meaningful time savings on routine mail. However, judging whether postage is 'correct' requires address parsing and rate calculation across many rule sets, and determining 'proper condition' involves nuanced judgment about what will survive transit—parts requiring human override. |
| Task automatability | claude-sonnet-5 | 3/5 | Computer vision and weighing/sensor systems can automatically check postage and package condition, but exceptions (damaged items, ambiguous cases, customer interaction) still require human judgment, so only partial time savings are realistic without heavy setup. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | No hard legal barrier prevents AI inspection, but USPS standards, liability for lost/damaged mail, and organizational preference to maintain human quality gates create moderate friction to full automation. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement for this specific check, though postal regulations and chain-of-custody/security concerns (e.g., detecting prohibited or hazardous items) create some institutional caution around full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 3/5 | AI vision infrastructure and integration costs are falling, but the need for human oversight to handle edge cases and verify system decisions keeps total cost roughly comparable to a clerk reviewing mail at a moderate pace. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | Industrial scanning/weighing equipment is cost-effective at high volume but requires significant capital investment; for lower-volume retail counter checks, cost parity with a clerk is closer to even. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Vision systems and package inspection tools exist and are piloted at scale in some postal operations, but error rates on subtle damage assessment and postage correctness remain material. Deployed systems typically flag ambiguous items for human review rather than making final decisions autonomously. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Automated mail sorting and postage verification systems exist at scale (USPS/FedEx sorting centers use machine vision and scales), but front-counter clerk checks of individual customer mail are still largely manual in retail settings. |
Receive letters and parcels, and place mail into bags.
50CI 13–87 · exposure 50 · augmentation 13 · importance 4.3/5 · click for rater detail
Receive letters and parcels, and place mail into bags.
50| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Large postal services and major logistics operators (UPS, FedEx, DHL) have extensively deployed automated sorting and bagging systems; adoption is deep in the logistics sector where scale justifies investment, though small regional facilities lag. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Postal services are a slow-moving, unionized, physically-oriented sector with limited AI/robotics adoption for hands-on sorting and handling tasks at retail counters. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI/automation in this domain is primarily substitutive rather than assistive; once a human is present, robotic systems do not meaningfully enhance their ability to perform the core bagging task itself. |
| Augmentation potential | claude-sonnet-5 | 1/5 | AI offers essentially no direct assistance to the physical act of receiving and bagging mail, though software might help with logistics elsewhere in the job. |
| Task automatability | claude-haiku-4-5-20251001 | 5/5 | Modern mail sorting and bagging automation systems (conveyor-based sorters, robotic systems) can handle the full task of receiving, sorting, and placing mail into bags with >50% time savings and equivalent quality; this is already standard in large postal facilities worldwide. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a physical manipulation task requiring receiving, sorting, and bagging mail items of varying shapes and sizes; current AI (software-based) cannot perform physical handling, and robotics for this specific unstructured task are not deployed at scale. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Capital equipment requirements and integration complexity create moderate adoption barriers, but no legal licensing requirement exists; some facilities retain human staff for exception handling or customer preference reasons. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing requirement specifically for bagging mail, but physical infrastructure, customer interaction, and chain-of-custody/security concerns around mail create moderate organizational friction against automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Industrial mail sorting systems have amortized costs far below the loaded wage of human postal clerks; once deployed, per-piece handling cost is orders of magnitude cheaper than manual labor. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI-only solution for this physical task, so any comparison would require robotics with high capital and maintenance costs likely exceeding human labor costs for this narrow task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 5/5 | Automated mail sorting and bagging systems are deployed at scale in postal services globally (USPS, Royal Mail, etc.) and operate reliably in production environments handling millions of pieces daily. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product performs receiving and bagging mail at postal counters; automated sorting exists in backend facilities but not this specific clerk-facing physical task. |
Complete forms regarding changes of address, or theft or loss of mail, or for special services such as registered or priority mail.
49CI 23–76 · exposure 50 · augmentation 50 · importance 4.4/5 · click for rater detail
Complete forms regarding changes of address, or theft or loss of mail, or for special services such as registered or priority mail.
49| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | USPS adoption of AI automation is slow; postal services are traditional, geographically distributed, and subject to federal employment protections and union contracts that inhibit rapid workforce displacement. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Postal services have adopted digital self-service for routine transactions but remain a traditionally slow-moving, unionized public sector environment with continued reliance on physical counters. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by auto-populating forms from customer input, flagging suspicious requests, or suggesting service options, meaningfully raising clerk productivity without full replacement—a reasonable augmentation use case. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-assisted form-filling, chatbots, and address-verification tools help clerks process requests faster and reduce errors, though human review remains common for fraud or dispute cases. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Form completion for address changes and special mail services is partially automatable with structured data extraction, but requires human judgment on fraud detection, policy interpretation, and exception handling—most workflows would still need significant human oversight and cannot achieve 50% time savings end-to-end. |
| Task automatability | claude-sonnet-5 | 4/5 | Form completion for address changes, mail theft/loss reports, and special service requests is a structured, rules-based data entry task well-suited to automation via online forms, chatbots, and workflow systems, though some cases require human judgment on complex claims. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | USPS is a federal agency with strict regulatory oversight; customer identity verification, fraud prevention, and legal accountability for mail handling create hard barriers requiring authorized human review and sign-off rather than pure automation. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement blocks automation, though some verification steps (identity proofing for change of address, fraud prevention) create moderate friction requiring oversight or hybrid processes. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Postal clerk wages are modest, and implementing AI systems with necessary compliance, integration, and human oversight would likely cost more than the direct labor savings, especially given low task complexity and high error-cost sensitivity in mail systems. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Automated online form processing costs a small fraction of a clerk's loaded wage per transaction, given high volume and minimal marginal cost of digital form submission and processing. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While OCR and basic form-filling systems exist, no mature production system reliably handles the full scope of postal forms (change-of-address, theft/loss reports, special service requests) with the accuracy and fraud-detection capability required by USPS standards. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | USPS and other postal services already deploy online self-service portals (e.g., USPS.com change-of-address, mail theft reporting) that handle much of this volume without clerk involvement, though in-person counter service persists for some transactions. |
Rent post office boxes to customers.
41CI 23–60 · exposure 45 · augmentation 50 · importance 4.2/5 · click for rater detail
Rent post office boxes to customers.
41| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | The Postal Service is a large, conservative, publicly managed organization with entrenched labor agreements and limited incentive to rapidly automate routine clerk work. Adoption of autonomous box-rental automation remains minimal. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Postal services are a slow-moving, unionized, government-run sector with uneven digitization; self-service kiosks exist but widespread automation of this specific task is still limited. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | Digital forms, payment systems, and customer lookup tools can assist a clerk by reducing manual data entry and speeding payment processing, but AI does not substantially transform the productivity of the core rental transaction itself. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-assisted forms, ID verification, and payment processing can speed up the clerk's portion of the transaction, though the task is simple enough that augmentation gains are modest. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While some elements (payment processing, documentation) could be automated, the task involves customer interaction, identity verification, and in-person box assignment that current AI cannot fully handle end-to-end. AI cannot currently meet a ≥50% time-saving threshold for the complete rental workflow. |
| Task automatability | claude-sonnet-5 | 4/5 | Renting a PO box is a structured transactional process (identity verification, form completion, payment) that self-service kiosks and online portals already handle for most of the workflow. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Postal Service operations are government-regulated; box rental involves identity verification, signature authorization, and legal accountability for key issuance that typically require a licensed postal employee to execute or directly oversee. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Identity verification requirements (often government ID checks) for box rental create some regulatory/compliance friction, though this is increasingly handled digitally with document-scanning verification. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current AI solutions (chatbots, CRM systems) require significant integration and oversight costs, while the task itself is simple, repetitive work that human clerks perform at modest wage rates. Full automation cost does not yet undercut human labor all-in. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | An automated kiosk or online system costs far less per transaction than a staffed clerk once deployed, though initial verification hardware and compliance systems add some cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed AI system reliably performs the full customer-facing rental transaction autonomously. Chatbots and online portals exist but do not eliminate the need for human verification, authorization, and physical key/box assignment in actual postal operations. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | USPS and other postal services offer online PO box rental and self-service kiosks, but many customers still complete this in person with a clerk, especially for ID verification, so deployment is partial. |
Provide customers with assistance in filing claims for mail theft, or lost or damaged mail.
34CI 25–43 · exposure 33 · augmentation 63 · importance 4.1/5 · click for rater detail
Provide customers with assistance in filing claims for mail theft, or lost or damaged mail.
34| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | The Postal Service is a large, traditional government organization with conservative adoption patterns. While some digital claim portals exist, widespread AI-driven claim assistance has not been adopted at scale, and the sector lags in AI pilot-to-production transition. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Postal service and public-sector logistics organizations tend to be slower adopters of AI relative to finance or tech, with online claim tools rolling out unevenly and pilot-stage AI use rather than widespread deployment. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI could usefully assist clerks by drafting claim templates, organizing customer information, retrieving relevant policies, or flagging common issues—helping structure the intake process. However, the human must remain central to judgment, fraud detection, and customer communication, limiting transformative potential. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully assist clerks by pre-filling forms, retrieving tracking data, summarizing claim history, and guiding customers through the process, improving speed and consistency while a human still finalizes decisions. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | This task requires nuanced customer interaction, understanding of claim procedures, and judgment about case-specific circumstances. While AI could draft templated claim forms, handling diverse customer situations, resolving ambiguities, and providing personalized guidance falls short of the 50% time-saving threshold for reliable end-to-end automation. |
| Task automatability | claude-sonnet-5 | 3/5 | Claim filing involves structured form completion and status lookups that chatbots and AI-assisted workflows can handle, but investigating theft/damage claims often requires judgment, verification, and exception handling that still need human involvement. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Postal Service operations are regulated and customer-facing; claims involve potential fraud assessment and legal liability. Organizational policies strongly favor human interaction for customer-facing claim handling, and errors in claim processing carry high reputational and legal costs that create organizational friction against full automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing requirement, but claims often involve verifying identity, ownership, and evidence, and there are regulatory/liability considerations around compensation determinations that create moderate friction for full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Integration costs for claim management systems, ongoing oversight to catch errors, and liability exposure would make AI assistance expensive relative to a clerk's wage. The need for human verification of claim details keeps total cost near or above human labor equivalence. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | Automated claim intake could be cheap to run, but oversight, fraud checks, and exception resolution still require paid staff time, keeping overall costs roughly comparable to human-only processing. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed product reliably handles the full scope of claim assistance—intake, customer communication, documentation requests, and case evaluation—at production scale. Chatbots exist for basic FAQs, but actual claim filing requires human judgment and verification that AI systems cannot yet perform reliably. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some postal-adjacent chatbots and online claim portals exist, but they mostly guide customers through forms rather than fully resolving claims, and complex cases still route to human clerks. |
Transport mail from one work station to another.
30CI 25–35 · exposure 25 · augmentation 13 · importance 4.2/5 · click for rater detail
Transport mail from one work station to another.
30| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Postal services are traditionally conservative, digitization is slow, and most facilities still rely on manual mail transport; automation adoption in this sector lags significantly behind information and finance sectors. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Postal service physical operations are a low-digitization, labor-intensive sector where robotic automation adoption for internal transport has been slow and uneven, mostly limited to major processing centers. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | While route-planning software could assist a clerk, AI offers limited meaningful productivity enhancement for the core manual task of physically moving mail between stations. |
| Augmentation potential | claude-sonnet-5 | 1/5 | AI offers essentially no direct assistance to a human physically carrying mail between workstations; this is a manual logistics task outside AI's typical augmentation scope. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While conveyor systems and sorting machines can move mail between stations, this task often requires navigating variable office layouts, handling exceptions, and making real-time decisions about mail routing that current AI systems cannot reliably do end-to-end without substantial human oversight. |
| Task automatability | claude-sonnet-5 | 2/5 | Physical transport of mail between workstations requires robotics or human labor; general-purpose AI systems (LLMs, software agents) cannot perform physical movement, though automated conveyor/sorting systems exist as specialized hardware, not 'AI' in the software sense.rated low for text/vision AI automation.rated for general AI capability. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Union agreements, labor regulations, and existing infrastructure investments create substantial organizational friction; postal services are heavily regulated and unionized, making substitution of human workers difficult regardless of technical capability. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing or legal requirement mandates a human perform this specific physical transport task, though facility redesign, safety regulations, and capital costs create moderate practical friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Existing automated systems require significant capital investment, maintenance, and integration costs that often exceed the wage cost of a postal clerk for many mail volumes. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Deploying robotic transport systems requires significant capital investment in facility infrastructure, making it costly relative to a clerk simply carrying mail short distances, unless at very large scale. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Some postal facilities use automated conveyor systems, but these are narrow-scope solutions requiring manual setup and handling of irregular items; no deployed AI product independently transports mail at scale without human assistance. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Automated material handling systems (conveyors, AGVs) exist in some large postal facilities, but these are specialized robotics/logistics equipment, not generally available AI products, and adoption is limited to a subset of large facilities. |
Put undelivered parcels away, retrieve them when customers come to claim them, and complete any related documentation.
29CI 23–35 · exposure 20 · augmentation 38 · importance 4.3/5 · click for rater detail
Put undelivered parcels away, retrieve them when customers come to claim them, and complete any related documentation.
29| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Postal services are traditionally slower to adopt automation compared to tech and finance sectors; most facilities still rely predominantly on manual sorting and retrieval. Pilot automation programs exist but widespread production deployment remains limited. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Postal services are a slow-moving, low-digitization physical-operations sector with minimal AI agent deployment for parcel handling tasks.yy |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | Inventory management systems and barcode/RFID tracking augment clerk productivity by reducing search time and automating documentation, though the core manual handling and customer interaction remain human-dependent. Modest productivity gains are realized through better tracking systems. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can help with digital documentation, tracking lookups, and customer identity verification, but offers little assistance for the physical storage/retrieval component.yy |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While retrieving and shelving parcels could theoretically be automated with robotics and inventory systems, the task requires physical manipulation in varied storage environments, customer interaction for claim verification, and handling of exceptions. Current AI+robotics systems can handle structured warehouse tasks but struggle with the full end-to-end workflow including documentation and customer service components. |
| Task automatability | claude-sonnet-5 | 2/5 | This is a physical retrieval and storage task involving locating parcels, verifying customer identity, and handling paperwork; current AI cannot perform the physical handling and only partially assists documentation.yy |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Postal services face some organizational friction and customer-preference factors favoring human service, though no hard legal licensing requirement exists. Union contracts and operational standardization in USPS and similar agencies provide moderate friction against full automation. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No strict licensing barrier, but chain-of-custody, identity verification, and liability for lost/misdelivered parcels create moderate friction against pure automation.yy |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Automated systems for this task (robotics, vision systems, integration with inventory software) carry high upfront capital and maintenance costs that typically exceed the loaded wage of a postal clerk, especially given the low-volume, variable-demand nature of parcel retrieval and customer interactions. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Automating physical retrieval would require robotics and facility redesign, which is currently more costly than a human clerk performing the same task.yy |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Some automated parcel sorting and basic inventory tracking systems exist in large postal operations, but no mature end-to-end system reliably handles the combined retrieval, customer claims verification, and documentation at production scale. Most deployments remain human-centric with limited automation. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs physical parcel storage/retrieval in a postal setting; some warehouse robotics exist but not integrated into postal clerk claim workflows.yy |
Provide assistance to the public in complying with federal regulations of Postal Service and other federal agencies.
28CI 25–30 · exposure 25 · augmentation 50 · importance 4.3/5 · click for rater detail
Provide assistance to the public in complying with federal regulations of Postal Service and other federal agencies.
28| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | The Postal Service and government agencies are historically slow to adopt AI due to regulatory constraints, procurement processes, and risk aversion; most adoption remains in back-office automation rather than customer-facing compliance assistance. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | USPS and government services are traditionally slow adopters of AI-driven customer service compared to fast-moving private-sector information industries. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI could assist clerks by drafting regulatory summaries, organizing relevant rules, or flagging potential compliance issues, reducing lookup time and improving consistency, though a human clerk must ultimately communicate and take responsibility for the guidance. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist clerks with quick lookup of regulations, form requirements, and answers to common questions, improving speed and consistency while the clerk remains the primary point of contact. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can retrieve and summarize postal regulations and general compliance information, the task requires explaining nuanced federal rules to diverse members of the public with different circumstances, which demands contextual judgment and clarification—capabilities current AI struggles with reliably without significant human oversight. |
| Task automatability | claude-sonnet-5 | 2/5 | This involves in-person interaction, physical document handling, and nuanced regulatory guidance that current AI cannot fully replicate end-to-end at equal quality without significant human oversight.chapter |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Postal Service is a federal agency with strict regulatory requirements; giving incorrect compliance advice carries legal liability, and there is a strong organizational and customer expectation that federal regulatory guidance be provided by authorized personnel who can be held accountable. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No strict licensing requirement, but there is organizational friction, customer preference for human clerks, and liability concerns around misinforming the public about federal regulations. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI infrastructure and oversight costs are comparable to or exceed a clerk's wage when factoring in the need for human review of compliance guidance, customer dissatisfaction recovery, and liability mitigation. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | While simple chatbot deployments are cheap, the need for accurate, liability-sensitive regulatory guidance and physical counter interaction keeps human staffing largely necessary, so all-in AI cost savings are limited. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Chatbots and rule-based systems exist to answer postal FAQs, but no production system reliably handles the full scope of federal compliance assistance across multiple agencies with acceptable error rates; errors carry liability risk, limiting real-world deployment at scale. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Chatbots and FAQ systems exist for basic postal inquiries, but no deployed product reliably handles the full range of regulatory compliance assistance across federal agencies in production at scale. |
Respond to complaints regarding mail theft, delivery problems, and lost or damaged mail, filling out forms and making appropriate referrals for investigation.
26CI 23–30 · exposure 25 · augmentation 50 · importance 4.3/5 · click for rater detail
Respond to complaints regarding mail theft, delivery problems, and lost or damaged mail, filling out forms and making appropriate referrals for investigation.
26| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | USPS is a government agency with slow digitization and conservative adoption patterns; most complaint handling remains manual and human-centered, with minimal AI-agent deployment in production roles. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Postal service is a slow-moving, unionized, government-affiliated sector with limited AI deployment in customer-facing complaint handling. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI could assist clerks by auto-populating forms, suggesting complaint classifications, and flagging missing information, meaningfully raising clerk efficiency without removing human oversight of complaint assessment and referral decisions. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with automated form completion, categorization of complaint types, and routing suggestions, improving clerk efficiency without replacing the judgment-based referral process. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI could assist with form-filling and basic triage, the task requires judgment about complaint validity, severity assessment, and appropriate referral decisions that often involve nuance and investigation discretion. Current systems cannot reliably perform the full task end-to-end with 50% time savings at equal quality. |
| Task automatability | claude-sonnet-5 | 2/5 | Complaint intake and form-filling could be partially automated via chatbots, but investigating lost/stolen mail, verifying claims, and making judgment-based referrals require human handling of ambiguous physical-world evidence and customer interaction. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | USPS operates under federal oversight and postal regulations; complaint handling and investigation referrals have legal and liability implications requiring documented human accountability, particularly for mail theft and loss claims that may trigger investigations or compensation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No strict licensing requirement, but mail theft is a federal matter needing accountability trails, human judgment for fraud detection, and customer trust in in-person clerks creates moderate friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Integration costs for complaint intake automation, oversight of AI triage decisions, and correction of misrouted cases would likely approach or exceed the cost of a clerk handling complaints directly, given error consequences in mail investigations. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI could cheaply handle basic intake, but the referral, verification, and escalation aspects still require human labor, keeping blended costs close to or above current staffing costs for full task completion. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed postal service system today automates complaint intake and investigation referral end-to-end. While chatbots can collect information and fill templates, actual complaint assessment and routing require human judgment that isn't yet reliably automated in production at scale. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some postal services deploy chatbots/online forms for complaint submission, but end-to-end handling including referral decisions and dispute resolution is not reliably automated in production today. |
Cash money orders.
16CI 14–18 · exposure 16 · augmentation 25 · importance 3.8/5 · click for rater detail
Cash money orders.
16| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Postal services remain largely traditional and slow to digitize core operations. Money order cashing is a legacy service with limited sector-wide automation adoption, especially in smaller postal facilities. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Postal and retail cash-handling functions are low-digitization, physical-service sectors with minimal AI agent adoption in production for this specific task.' |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist with rapid verification of money order validity and fraud detection, but the physical cash handling and customer interaction components limit the breadth of augmentation available today. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can support fraud detection or transaction logging in the background, but offers little direct productivity boost to the physical act of cashing a money order.' |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Cashing money orders requires physical handling (accepting, validating, dispensing cash), customer interaction, and fraud detection. While AI could assist with verification and record-keeping, the physical cash transaction and in-person customer interface prevent end-to-end automation meeting the 50% time-saving threshold today. |
| Task automatability | claude-sonnet-5 | 2/5 | Cashing money orders requires physical handling of cash, identity verification, and in-person transaction processing that current AI systems cannot perform end-to-end without robotic/physical infrastructure.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Cash handling is heavily regulated; federal and postal regulations typically require a licensed/bonded human employee to handle and disburse cash, creating legal and liability barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Handling cash and verifying identity for financial transactions involves regulatory compliance (e.g., anti-fraud, KYC rules) and liability concerns that typically require an authorized human employee.' |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The combination of required hardware (cash handling, document scanning, customer interaction systems) and ongoing oversight makes AI automation more expensive than the relatively low wage of a postal clerk performing this routine task. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | While kiosks/ATMs exist for some cash transactions, adapting AI to fully replace clerk-mediated money order cashing would require costly specialized hardware, making near-term cost savings limited.' |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI system reliably performs the full cash transaction workflow end-to-end in postal environments. Cash handling and fraud detection require human judgment and physical presence that current systems cannot reliably deliver. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product autonomously cashes money orders; this remains a manual, in-person financial transaction handled by human clerks at counters.' |
Obtain signatures from recipients of registered or special delivery mail.
3CI 0–5 · exposure 0 · augmentation 13 · importance 4.3/5 · click for rater detail
Obtain signatures from recipients of registered or special delivery mail.
3| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Postal services remain heavily dependent on human carriers for physical delivery and signature collection. Adoption of automation in this sector is minimal, with no evidence of AI agents in production replacing signature-collection duties. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Postal delivery and physical signature collection is a low-digitization, physically-bound task with minimal AI adoption in this specific function. |
| Augmentation potential | claude-haiku-4-5-20251001 | 1/5 | Current AI offers no meaningful assistance in the core task of obtaining signatures from recipients. The task is entirely dependent on in-person interaction and does not benefit from AI tools or prompts. |
| Augmentation potential | claude-sonnet-5 | 2/5 | Electronic signature pads and tracking software offer minor digital assistance in recording and logging signatures, but the core task of physically obtaining them is unaffected by AI. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Obtaining signatures requires direct physical interaction with a human recipient, which current AI systems cannot perform. Even with robotic agents, the task involves verifying identity and ensuring recipient consent—capabilities that remain beyond practical deployment today. |
| Task automatability | claude-sonnet-5 | 1/5 | Obtaining a physical signature from a recipient requires physical presence and interaction at a delivery point, which current AI systems cannot perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Postal delivery requires government authorization; registered and special delivery mail have legal requirements for proof of delivery and signature authentication that mandate human verification and accountability. Liability for lost or misdelivered mail creates hard barriers to substitution. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Physical presence and often identity verification are required, and legal/chain-of-custody requirements for registered mail create strong barriers to any non-human substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Any attempt to automate would require robotics or remote verification infrastructure far exceeding the cost of a human postal clerk performing the task. The capital and operational overhead makes AI orders of magnitude more expensive than human labor. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute for the physical act of collecting a signature, so AI cost comparison is not applicable and the human remains necessary and cost-effective for this step. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product performs signature collection from mail recipients in production. The task fundamentally requires physical presence and human-to-human verification, which autonomous systems do not reliably achieve in real-world postal operations. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product independently obtains signatures from recipients; this remains a physical, in-person task performed by human clerks or carriers with electronic capture devices operated by humans. |
Related occupations — Office & Administrative Support
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.