File Clerks
43-4071.00Median wage $43,600/yr73,440 employed (US)Rank #27 of 923 scored · top 3% by substitution
File correspondence, cards, invoices, receipts, and other records in alphabetical or numerical order or according to the filing system used. Locate and remove material from file when requested.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution65
Exposure63
Augmentation57
Substitution — the headline: capability discounted by cost, barriers and adoption.
Exposure — technical capability alone, regardless of whether anyone deploys it.
Augmentation — how much AI assists without replacing. High here + moderate substitution = a changing job, not a disappearing one.
Tasks on the substitution scale
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
47%
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%66
panel mean rating 3.6/5 → substitution pressure 66/100
Technical feasibility todayw 20%58
panel mean rating 3.3/5 → substitution pressure 58/100
Cost vs. human wagew 15%67
panel mean rating 3.7/5 → substitution pressure 67/100
Adoption barriersw 20%inverted — strong barriers lower the score76
panel mean rating 1.9/5 (barrier strength) → substitution pressure 76/100
Sector adoption velocityw 10%45
panel mean rating 2.8/5 → substitution pressure 45/100
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.
Sort or classify information according to guidelines, such as content, purpose, user criteria, or chronological, alphabetical, or numerical order.
97CI 97–97 · exposure 100 · augmentation 63 · importance 4.0/5 · click for rater detail
Sort or classify information according to guidelines, such as content, purpose, user criteria, or chronological, alphabetical, or numerical order.
97| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | File sorting and document classification have been priorities for RPA and AI-based document management in corporate and government offices for years. Adoption is well underway in digitized sectors, though legacy organizations still rely on manual processes. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Administrative and clerical functions in offices, healthcare records, and finance have widely adopted digital filing and automated classification systems, though some legacy paper-based environments lag. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can assist human clerks by auto-suggesting categories, pre-sorting batches, and highlighting edge cases for human review. This significantly raises human productivity while humans remain involved for exception handling and quality assurance. |
| Augmentation potential | claude-sonnet-5 | 3/5 | Where humans still handle exceptions or ambiguous cases, AI tools can pre-sort and flag items, improving throughput, though most of the core task is already fully automatable rather than merely augmented. |
| Task automatability | claude-haiku-4-5-20251001 | 5/5 | Sorting and classifying information by defined criteria (content, chronological, alphabetical, numerical order) is a fully structured, rule-based task. Modern AI systems and document processing tools can perform this end-to-end with >50% time savings at equal or superior quality. |
| Task automatability | claude-sonnet-5 | 5/5 | Sorting and classifying information by defined rules (alphabetical, numerical, chronological, categorical) is a well-structured, rule-based task that current AI/automation systems handle natively at scale with high accuracy. |
| Adoption barriers | claude-haiku-4-5-20251001 | 1/5 | No legal or regulatory barriers protect this task; no licensing or human sign-off is required. Organizational inertia may slow adoption, but nothing prevents substitution. |
| Adoption barriers | claude-sonnet-5 | 1/5 | There is no licensing, liability, or human-contact requirement tied to sorting information; it's a purely administrative function with minimal adoption friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | The cost of automation (API calls, RPA licensing, minimal oversight) is orders of magnitude cheaper than paying a human clerk's loaded wage to perform repetitive sorting and classification work. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Automated sorting via software/scripts costs a tiny fraction of a cent per record compared to hourly clerical wages, representing well over an order-of-magnitude savings. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 5/5 | Deployed products demonstrably perform document sorting, classification, and organization at scale today. RPA, ML-based document classifiers, and file management systems reliably execute this task in production across many organizations. |
| Technical feasibility today | claude-sonnet-5 | 5/5 | Document management systems, database software, and AI-based classification tools already perform automated sorting and filing in production across countless organizations today. |
Keep records of materials filed or removed, using logbooks or computers and generate computerized reports.
95CI 92–97 · exposure 100 · augmentation 50 · importance 4.4/5 · click for rater detail
Keep records of materials filed or removed, using logbooks or computers and generate computerized reports.
95| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | File clerk roles have been declining for decades as organizations digitize; automation of record-keeping and reporting is already widespread in finance, healthcare, government, and professional services sectors, with mature adoption in information-intensive industries. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Digital record-keeping and automated audit trails are already standard in most office environments, reflecting fast, broad adoption of software-based tracking over manual logbooks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist file clerks by auto-categorizing documents, suggesting filing locations, or flagging anomalies in records, but since the core task (logging and reporting) is already highly automatable, augmentation is secondary and less transformative than in more judgment-heavy roles. |
| Augmentation potential | claude-sonnet-5 | 3/5 | Where clerks still perform manual filing, software can assist by auto-generating reports and reminders, though the underlying task is largely already automated rather than merely augmented. |
| Task automatability | claude-haiku-4-5-20251001 | 5/5 | This task is a near-perfect candidate for end-to-end automation: recording file movements and generating reports are inherently digital processes that current RPA and database systems can perform reliably at far less than half the time required by human clerks, with higher accuracy and no quality loss. |
| Task automatability | claude-sonnet-5 | 5/5 | Tracking file movement and generating reports is a structured data-entry and reporting task that off-the-shelf document management and database systems can fully automate with automatic logging and report generation. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | While some organizations have legacy systems requiring integration work and some may prefer human oversight for compliance, there are no legal licensing requirements, liability asymmetry, or regulatory mandates that a human must perform record-keeping; adoption friction is primarily organizational and technical, not legal. |
| Adoption barriers | claude-sonnet-5 | 1/5 | No licensing, liability, or human-contact requirements attach to internal record-keeping logs; organizations can freely automate this without regulatory constraint. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Automating record-keeping and report generation via RPA or integrated document management systems costs pennies per transaction or operation, while even a low-wage file clerk costs $15–$25/hour; the cost ratio is orders of magnitude in favor of automation. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Automated logging via software incurs negligible marginal cost compared to a human manually maintaining logbooks or updating tracking spreadsheets. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 5/5 | Document management, workflow automation, and reporting software are mature, widely deployed products used in organizations of all sizes today; vendors like OpenText, FileNet, and basic database systems with logging already perform these exact functions in production at scale. |
| Technical feasibility today | claude-sonnet-5 | 5/5 | Enterprise content management and records management systems (e.g., SharePoint, DMS platforms) already log check-in/check-out events and auto-generate audit reports in production at scale. |
Input data, such as file numbers, new or updated information, or document information codes into computer systems to support document and information retrieval.
89CI 81–97 · exposure 87 · augmentation 75 · importance 4.3/5 · click for rater detail
Input data, such as file numbers, new or updated information, or document information codes into computer systems to support document and information retrieval.
89| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Organizations across finance, healthcare, legal, and government sectors are actively adopting document automation and RPA for data entry. Adoption is accelerating in information-intensive industries, though lagging in small firms and non-digitized sectors. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Clerical/administrative functions are adopting RPA and IDP steadily, but many organizations still run legacy systems and manual processes, so adoption is moderate rather than fast and universal. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI assists file clerks by auto-populating fields, flagging anomalies, and suggesting corrections, substantially raising productivity even when humans remain in the loop. Augmentation is strong and widely deployed in document management systems today. |
| Augmentation potential | claude-sonnet-5 | 4/5 | Even where full automation isn't achieved, AI-assisted data capture (auto-fill, validation, OCR suggestions) significantly speeds up remaining human-in-the-loop data entry work. |
| Task automatability | claude-haiku-4-5-20251001 | 5/5 | Data entry into computer systems is highly automatable with OCR, RPA, and structured extraction tools that can achieve >50% time savings. Existing systems like document capture platforms and form-filling bots already handle file numbers, codes, and standardized information input at scale. |
| Task automatability | claude-sonnet-5 | 4/5 | Data entry from structured or semi-structured documents into computer systems is well within the capability of OCR plus AI/RPA pipelines, meeting the 50% time-saving bar for most instances though some messy source documents still need human correction. |
| Adoption barriers | claude-haiku-4-5-20251001 | 1/5 | No licensing, authorization, or legal requirement mandates human involvement. Data entry is routine clerical work with no regulatory barrier to automation, and error costs, while present, are easily managed through verification workflows. |
| Adoption barriers | claude-sonnet-5 | 1/5 | There is no licensing, liability, or human-contact requirement tied to file/data entry; it's a purely administrative back-office task with minimal regulatory protection. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI-driven document capture and RPA cost far less than human data entry labor when amortized across volume. A single document processing solution can handle thousands of records, making per-task cost 10–100x lower than manual entry. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Automated document processing pipelines cost a small fraction per document compared to a human clerk's loaded wage for equivalent data entry volume. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 5/5 | Mature, deployed products (document processing software, RPA platforms, OCR engines) reliably perform this task in production across many organizations. Enterprise systems like UiPath, Automation Anywhere, and cloud-based document intake services are proven and widely adopted. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Mature OCR, intelligent document processing, and RPA products (e.g., ABBYY, UiPath, Azure Form Recognizer) are deployed at scale in enterprises for exactly this kind of data indexing and entry task. |
Assign and record or stamp identification numbers or codes to index materials for filing.
89CI 81–97 · exposure 87 · augmentation 63 · importance 3.9/5 · click for rater detail
Assign and record or stamp identification numbers or codes to index materials for filing.
89| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | While document automation is growing, file clerk roles remain prevalent in sectors with lower digitization (small firms, government, healthcare records). Adoption is happening but remains uneven; this is not yet as rapid as in finance or high-tech. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Records management and back-office administrative functions across industries have rapidly adopted digital indexing and automated classification tools over the past decade. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can significantly assist clerks by pre-assigning codes and flagging uncertain classifications for human review, or auto-populating index fields. This keeps humans in quality-control while boosting throughput substantially. |
| Augmentation potential | claude-sonnet-5 | 3/5 | Where physical files or legacy systems remain, AI-assisted scanning and metadata suggestion tools speed up human indexing, though many environments are already fully automated rather than human-augmented. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | This task involves reading documents, identifying content, assigning standardized codes/numbers, and recording them—all activities that current AI systems handle well. OCR + classification + database entry can achieve well over 50% time savings with minimal human oversight on routine materials, though some edge cases may require human judgment. |
| Task automatability | claude-sonnet-5 | 5/5 | Assigning and recording identification codes to documents is a rules-based, structured data-entry task that OCR/document-management systems automate fully today with far less time than manual stamping and logging. |
| Adoption barriers | claude-haiku-4-5-20251001 | 1/5 | No licensing requirement, regulatory mandate, or legal obligation requires a human to assign filing codes. Organizational inertia may exist, but nothing prevents automation—indexing systems are not legally protected work. |
| Adoption barriers | claude-sonnet-5 | 1/5 | No licensing, liability, or human-contact requirement applies to indexing and coding documents; it's a purely administrative function with minimal friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | OCR, classification, and database recording are among the cheapest AI tasks to perform at scale; inference cost is minimal and integration into existing filing systems is straightforward. This easily achieves an order of magnitude cost advantage over human clerks. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Automated indexing via scanning/OCR software costs a small fraction of a per-document basis compared to a human clerk's hourly wage for the same volume. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed document processing systems (Tesseract, commercial RPA platforms, document AI services) reliably perform document classification and number assignment in production. Some variability exists with poor-quality scans or ambiguous documents, but core functionality is mature and production-proven. |
| Technical feasibility today | claude-sonnet-5 | 5/5 | Enterprise document management and records systems (e.g., SharePoint, Iron Mountain digital, DMS with auto-indexing) already assign IDs, barcodes, and metadata automatically at scale in production. |
Scan or read incoming materials to determine how and where they should be classified or filed.
77CI 76–79 · exposure 75 · augmentation 63 · importance 4.0/5 · click for rater detail
Scan or read incoming materials to determine how and where they should be classified or filed.
77| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Document automation and RPA tools are widely adopted in finance, healthcare, and administrative functions; many large organizations already deploy document classification in production workflows. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Document processing automation is spreading steadily across administrative and clerical functions, though many smaller organizations still rely on manual filing processes.on |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist clerks by pre-classifying and flagging high-confidence documents, reducing manual review burden and speeding triage, though the human remains the decision-maker for uncertain cases. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-assisted classification tools significantly speed up human review and sorting by pre-sorting and flagging materials, even when final judgment or exception-handling remains human.on |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | AI can reliably classify and route documents using OCR and text-classification models, achieving >50% time savings on incoming material triage. Most of the scanning and routing logic is highly automatable, though complex edge cases may require human review. |
| Task automatability | claude-sonnet-5 | 4/5 | Document classification via OCR and ML-based classification/routing is a well-established capability that can handle most standard document types with high accuracy, meeting the time-saving threshold for routine cases.ed |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few regulatory barriers exist for document classification automation itself, though organizations may have internal policies preferring manual handling for sensitive materials or requiring human oversight for certain filetypes. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement for filing; main friction is ensuring accuracy for sensitive or ambiguous documents and integration with legacy systems, but no legal mandate for human filing.on |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Cloud-based OCR and classification APIs cost pennies per document; total inference and integration is orders of magnitude cheaper than human clerk labor for high-volume scanning and routing. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Automated OCR/classification pipelines cost fractions of a cent per document versus the loaded wage of a human clerk performing the same sorting task.on |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed document classification and OCR systems exist in production (e.g., RPA platforms, enterprise content management tools). Error rates on well-defined categories are low, though performance degrades on ambiguous or handwritten content. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Production document management systems widely use automated classification and routing (e.g., intelligent document processing platforms) in insurance, legal, and enterprise settings today.on |
Complete general financial activities, such as processing accounts payable, reviewing invoices, collecting cash payments, or issuing receipts.
76CI 72–79 · exposure 75 · augmentation 63 · importance 4.2/5 · click for rater detail
Complete general financial activities, such as processing accounts payable, reviewing invoices, collecting cash payments, or issuing receipts.
76| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Finance and accounting sectors have rapidly adopted AP automation, RPA, and document-processing AI. Major companies and mid-market firms are in active production; adoption is fast and measurable in cost-savings reports and headcount reduction. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Finance back-office functions show moderate-to-strong AI/RPA adoption, but many smaller organizations still rely on manual or semi-manual invoice and cash-handling processes. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist clerks by pre-filling invoice fields, flagging anomalies, and summarizing cash transactions, raising accuracy and speed. However, the task is routine enough that augmentation is secondary to full automation in most implementations. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools substantially speed up invoice matching, data entry, and receipt generation, letting clerks focus on exceptions and reconciliation while remaining in the workflow. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Current AI can automate most of these routine financial activities through RPA and document processing: invoice review via OCR/ML, accounts payable processing with structured data extraction, payment receipt generation, and basic cash collection tracking are all well-established. Some tasks like exception handling may still require human judgment, but the core workflow easily achieves >50% time savings. |
| Task automatability | claude-sonnet-5 | 4/5 | Invoice processing, AP workflows, and receipt generation are highly structured, rules-based tasks that modern AI/RPA systems (e.g., AP automation platforms) can largely execute with significant time savings, though cash handling requires physical presence. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few regulatory or legal barriers exist for automating these routine clerical tasks; no licensing requirement mandates human signing. Main friction is internal organizational change management and integration with legacy systems, not external compliance. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement for this clerical work, though internal controls, segregation-of-duties requirements, and audit/compliance expectations create some organizational friction around automating financial transactions. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI-driven document processing and RPA cost pennies per transaction compared to a file clerk's fully-loaded wage (typically $35–45k annually, ~$20/hour). Processing thousands of invoices monthly via automation is orders of magnitude cheaper than manual entry. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated invoice/AP processing software costs a small fraction of a clerk's loaded wage per transaction once implemented, though integration and exception-handling oversight add some cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed products (Kofax, UiPath, Automation Anywhere, specialized AP automation platforms) reliably handle invoice processing, data extraction, and receipt generation at scale in production environments. Error rates on routine invoices are low, though complex or malformed documents may still need human review. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Mature AP automation and OCR-based invoice processing products (Bill.com, SAP Concur, Stampli) are deployed at scale in production for invoice review and payables, though cash collection remains manual. |
Answer questions about records or files.
76CI 67–84 · exposure 75 · augmentation 88 · importance 4.0/5 · click for rater detail
Answer questions about records or files.
76| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Adoption is accelerating rapidly in professional services, healthcare, and finance where document volume is high and ROI is clear. Enterprise search and document AI tools are now standard in digitized organizations. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Administrative and clerical functions are adopting AI search/chat tools at a moderate pace, with pilots common but full-scale deployment inconsistent across sectors relying on file clerks (e.g., government, healthcare records). |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI powerfully augments file clerks by instantly surfacing relevant documents and providing summaries, allowing humans to focus on judgment calls, complex queries, and exception handling rather than manual search. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can significantly speed up locating and summarizing file information for a human clerk who verifies or contextualizes results. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Current AI systems can reliably retrieve and answer questions about records or files through semantic search, RAG (retrieval-augmented generation), and document understanding. With proper indexing and setup, this achieves well over 50% time savings at equal or better quality compared to manual file searching. |
| Task automatability | claude-sonnet-5 | 4/5 | If records are digitized and searchable, AI systems (chatbots, RAG-based retrieval) can answer most routine questions about file contents or locations with substantial time savings. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few hard legal barriers exist; however, organizations face some friction from data access policies, sensitivity of records, and preference for human verification on sensitive queries. Liability concerns around incorrect answers add modest friction. |
| Adoption barriers | claude-sonnet-5 | 2/5 | Some organizational and privacy/compliance concerns exist around records access, but no licensing requirement mandates a human answer such questions. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI inference cost for document retrieval and QA is orders of magnitude cheaper than paying a file clerk's loaded wage ($35k–$45k annually) to search and answer the same queries manually. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Once digitized, AI-based retrieval and Q&A costs (cloud inference, indexing) are far lower per query than a human clerk's wage for repetitive lookups. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 5/5 | Deployed products (document QA systems, enterprise search tools, LLM-based retrieval agents) perform this task reliably in production at scale across law firms, healthcare organizations, and administrative departments. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Deployed enterprise search and chatbot tools handle document Q&A in many organizations, but many file clerk environments still involve legacy paper records or unindexed systems that limit reliability. |
Add new material to file records or create new records as necessary.
73CI 67–79 · exposure 70 · augmentation 63 · importance 4.1/5 · click for rater detail
Add new material to file records or create new records as necessary.
73| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Organizations with significant back-office operations—finance, healthcare, government, legal—have already adopted RPA and document automation for these exact tasks in production. Adoption is measurable and rapid in digitized sectors, though small or less-digitized organizations lag. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Digitization and document management automation are progressing steadily in offices and administrative sectors, but full replacement of file clerks remains uneven due to legacy paper systems and small-office lag. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist clerks by auto-populating fields, suggesting correct filing categories, and flagging suspicious or anomalous records for manual review, raising productivity. However, the task is primarily mechanical, so augmentation potential is meaningful but not transformative compared to full automation. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools significantly speed up sorting, classification, and data entry for file clerks, letting them focus on exceptions and quality control rather than manual filing. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | File clerks adding material to records or creating new records involves largely structured data entry and basic categorization. Current AI systems (e.g., OCR + RPA) can automate the majority of this workflow, including document scanning, field extraction, and filing into existing systems, delivering meaningful time savings, though human judgment on edge cases may still be needed. |
| Task automatability | claude-sonnet-5 | 4/5 | Filing and record-creation is a structured, rules-based data entry task that current AI/automation systems (OCR + document classification + database integration) can perform largely end-to-end for digital records, though physical filing still requires human action. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few regulatory or legal barriers exist for automating file creation and record-adding tasks in most sectors. The main friction is organizational inertia and legacy system integration; there is no licensing requirement or mandatory human sign-off for this clerical function. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing or legal requirement mandates human filing; some organizational friction exists around data governance, security, and legacy paper records but no hard regulatory barrier for most industries. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI-driven document processing and RPA operate at a tiny fraction of the cost of a human file clerk, including the overhead of training, benefits, and continuous availability. Inference and integration costs are negligible compared to even minimum-wage labor for repetitive filing tasks. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated indexing and filing software costs a small fraction of a clerk's hourly wage once implemented, though initial setup and integration costs offset some savings versus simple manual filing. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Document management and RPA solutions deployed at scale in enterprises routinely handle record creation and filing. Products like UiPath, Blue Prism, and integrated AI-powered document systems reliably perform this task in production, though implementation complexity and integration with legacy systems can create friction. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Document management systems with AI-assisted classification and auto-filing exist and are deployed in many organizations, but reliability varies with document quality, edge cases, and legacy paper-based workflows still requiring manual handling. |
Find, retrieve, and make copies of information from files in response to requests and deliver information to authorized users.
71CI 62–79 · exposure 70 · augmentation 75 · importance 4.1/5 · click for rater detail
Find, retrieve, and make copies of information from files in response to requests and deliver information to authorized users.
71| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Large organizations and information-heavy sectors (finance, healthcare, legal) have deployed document management and RPA systems for file retrieval at scale. Adoption is mature in digital-forward enterprises, though small firms and paper-heavy operations lag. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Clerical and administrative support roles are seeing moderate digitization and automation adoption, but many file clerk environments (small offices, government records) still lag in full deployment. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI dramatically assists file clerks through rapid search, indexing, and automated retrieval suggestions, allowing humans to focus on authorization decisions and complex queries. Productivity gains are substantial when AI handles the locate-and-copy workflow while clerks manage exceptions and permissions. |
| Augmentation potential | claude-sonnet-5 | 4/5 | Search tools, OCR, and automated indexing significantly speed up a clerk's ability to locate and deliver requested information, even when a human remains involved for verification and exceptions. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Current AI systems can reliably locate files by metadata/content, retrieve documents, and generate copies with high accuracy. End-to-end automation of routine retrieval and copying meets the ≥50% time-saving threshold, though authorization verification and unusual file-structure edge cases may require some human oversight. |
| Task automatability | claude-sonnet-5 | 4/5 | For digitized files, retrieval and copying can largely be automated via search/indexing systems and document management software, though physical files and edge cases requiring judgment reduce full automation.4 |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Authorization controls and access management add friction; many organizations require human sign-off on sensitive file delivery. Regulatory compliance (HIPAA, financial records) and organizational policies around data handling create meaningful but surmountable adoption barriers. |
| Adoption barriers | claude-sonnet-5 | 2/5 | Some friction exists around authorization verification and data privacy/security requirements, but no licensing requirement mandates a human perform simple retrieval and copying. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Document retrieval and copying via software is orders of magnitude cheaper than manual file clerk labor once systems are deployed. Inference and integration costs are minimal compared to the loaded wage of a clerk performing these repetitive physical and clerical tasks. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Once files are digitized and indexed, automated retrieval costs are far lower than paying a human clerk per request, though initial digitization and system setup carry upfront costs. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed document management systems, RPA tools, and AI-powered retrieval solutions (including OCR and semantic search) handle file retrieval and copying at scale in production environments. Reliability is high for standard document types, though some legacy or unstructured filing systems may require hybrid approaches. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Enterprise document management systems with search and automated retrieval are deployed today, but many organizations still rely on manual retrieval for legacy paper files or poorly indexed systems. |
Perform periodic inspections of materials or files to ensure correct placement, legibility, or proper condition.
64CI 52–76 · exposure 62 · augmentation 63 · importance 3.7/5 · click for rater detail
Perform periodic inspections of materials or files to ensure correct placement, legibility, or proper condition.
64| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Adoption is uneven: large enterprises and government agencies invest in document automation, but many small and mid-size file management operations remain manual. Broader uptake is held back by legacy workflows and lower digitization in some sectors, placing adoption in the middling range. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Clerical/administrative support roles are typically slower adopters of AI tooling compared to finance or professional services, though digital record-keeping systems are gradually incorporating automated checks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI-powered inspection dashboards significantly enhance clerk productivity by flagging issues for human review, prioritizing work, and reducing manual scanning time. The human can focus on judgment-heavy edge cases while AI handles routine detection, creating strong assistive value. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI tools like OCR and metadata validators can help flag illegible or misplaced digital files, assisting clerks in prioritizing manual review, though physical inspection still needs human judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Computer vision and document analysis systems can reliably detect misplacement, illegibility, and material damage at scale. A significant portion of the inspection workflow (scanning, flagging issues, organizing results) can be fully automated, easily meeting the 50% time-saving threshold with current OCR and image-classification tools. |
| Task automatability | claude-sonnet-5 | 3/5 | AI/OCR systems can check digital file placement, legibility, and metadata correctness at scale, but inspecting physical files or materials still requires human or robotic handling not fully automated today.atabase |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few regulatory or licensing barriers exist for automating file inspection itself; organizations can deploy without human sign-off requirements. Minor friction points include integration with existing file systems and staff training, but nothing structurally prevents substitution. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing or legal requirement mandates human inspection; the main friction is practical necessity for physical materials rather than regulatory or liability barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Automated scanning and AI-based inspection of documents costs a fraction of a human clerk's loaded wage for equivalent output volume, especially in high-throughput settings. The per-unit cost of AI inspection is substantially cheaper once infrastructure is in place. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | For digitized records, automated scanning/validation is cheaper than manual review, but many file clerk environments still involve physical files where AI offers no cost advantage, making the average roughly comparable. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed document scanning, OCR, and quality-assurance platforms (including AI-powered defect detection) are in production use across many organizations. Error rates on legibility and basic condition checks are low enough for routine deployment, though human verification of edge cases remains common. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Document management systems with automated audit/validation features exist and are deployed, but they mainly cover digital records; physical file inspection still relies on manual work in most organizations. |
Perform general office activities, such as typing, answering telephones, operating office machines, processing mail, or securing confidential materials.
58CI 49–67 · exposure 58 · augmentation 63 · importance 4.4/5 · click for rater detail
Perform general office activities, such as typing, answering telephones, operating office machines, processing mail, or securing confidential materials.
58| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Adoption is primarily in larger firms with high-digitization office environments; most small firms and government agencies retain file clerks, indicating slow and uneven market penetration. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Office administrative functions are being augmented steadily with AI tools (virtual assistants, document management), but full replacement in file clerk roles remains at pilot/moderate stage in most firms. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist with document sorting, indexing, and mail categorization, improving clerk productivity on routine tasks, though phone answering and confidential material handling still require human judgment and oversight. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI substantially aids drafting, transcription, call handling, and document organization, letting a human file clerk cover more volume with the same effort. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | Typing, mail processing, and document handling can be partially automated with RPA and document processing AI, but answering phones, determining confidentiality levels, and securing materials require judgment and context that limit full end-to-end automation to roughly half the work. |
| Task automatability | claude-sonnet-5 | 4/5 | Typing, mail processing, and answering standardized calls are largely automatable with current AI (voice agents, OCR, document routing), though some physical tasks (securing materials, operating machines) still need a human presence. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few legal or licensing barriers exist, but organizational inertia, customer preference for human contact in phone answering, and confidentiality handling policies create moderate friction to adoption. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement, but confidential material handling can trigger internal policy/security constraints and some preference for human accountability in sensitive information handling. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | While automation of typing and mail sorting is cheap per unit, the integration, oversight, and exception-handling overhead, combined with the low wage of file clerks, means full-stack cost savings remain modest and sector-dependent. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Cloud-based transcription, virtual receptionists, and automated mail sorting are cheap per unit compared to a loaded clerical wage, though integration and oversight costs reduce the savings somewhat. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Products exist for specific subtasks (optical character recognition for mail, chatbots for phone screening, RPA for data entry) but no single system reliably handles the full scope of general office activities with consistent quality across all components. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | AI transcription, chatbots, and mailroom automation systems exist in production, but many organizations still use humans for phone answering and physical security tasks, so reliability across the full bundle is mixed. |
Modify or improve filing systems or implement new filing systems.
56CI 44–67 · exposure 53 · augmentation 75 · importance 3.3/5 · click for rater detail
Modify or improve filing systems or implement new filing systems.
56| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Mid-market and large organizations are adopting intelligent document management and AI-driven file classification pilots, but full production deployment remains inconsistent. Financial services and healthcare lead adoption; smaller organizations lag due to integration costs and legacy system constraints. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Clerical/administrative functions in many organizations are slower to adopt AI-driven system redesigns compared to core information-work functions like finance or software. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI significantly assists file clerks by automatically analyzing file patterns, suggesting optimal categorization schemes, and flagging inconsistencies, allowing clerks to focus on validation, exception handling, and domain-specific judgment. The human-in-the-loop model raises productivity substantially on this task. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can significantly assist in analyzing document volumes, suggesting taxonomies, and drafting new filing protocols, substantially speeding up the design phase even if humans oversee implementation. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | File system design, data organization, and implementation of new classification schemes can largely be automated through AI-driven analysis of existing filing patterns, metadata extraction, and optimization algorithms. Current systems can exceed 50% time savings by automatically categorizing files, identifying redundancies, and recommending structural improvements, though some domain expertise validation typically remains. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can analyze existing file structures and propose taxonomies or metadata schemas, but implementing physical or hybrid filing systems and integrating with legacy processes still requires human judgment and execution.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | There are minimal regulatory or legal barriers to automating file system design and implementation; no licensing requirement exists for this task. The main friction is organizational inertia, data sensitivity concerns, and staff resistance to system changes, but these are adoption frictions rather than hard blockers. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement, but organizational inertia, need for buy-in from multiple departments, and data governance/compliance concerns create moderate friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-based document classification and file system optimization tools cost substantially less than manual labor for large-scale implementations. Once deployed, the per-file cost of automated organization is orders of magnitude below the loaded wage of a file clerk performing the same analysis and reorganization work. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Designing and implementing a new filing system requires analysis, stakeholder input, and change management that AI tools reduce but don't eliminate, keeping costs closer to human-led efforts than an order of magnitude cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Products exist for automated file organization, metadata tagging, and data governance, but they require significant configuration, human review of categorization decisions, and integration with existing organizational systems. Deployed solutions handle well-scoped scenarios reliably but struggle with ambiguous, context-dependent classification decisions across heterogeneous legacy systems. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some document management platforms offer AI-assisted auto-categorization and tagging, but full redesign and rollout of filing systems in production is still largely a human-led consulting/IT task. |
Track materials removed from files to ensure that borrowed files are returned.
52CI 35–70 · exposure 55 · augmentation 63 · importance 4.2/5 · click for rater detail
Track materials removed from files to ensure that borrowed files are returned.
52| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | File clerk roles remain concentrated in large institutions (hospitals, legal firms, government) that move slowly on automation; many smaller organizations and departments still use manual tracking. Adoption of systematic automation is gradual and often limited to digitization of new files rather than full replacement of human oversight. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | File clerk roles are concentrated in smaller offices, legal, healthcare, and government settings with slower digitization and physical record-keeping, resulting in middling-to-slow adoption despite available technology. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | Automated tracking systems and email/SMS reminders can assist file clerks by flagging overdue returns and generating reports, reducing manual checking time. However, the improvement is incremental rather than transformative, as the core task of verifying returns and following up still requires human judgment and persistence. |
| Augmentation potential | claude-sonnet-5 | 4/5 | Even where physical files remain, software tools significantly assist clerks by automating logging, sending reminders, and flagging overdue items, improving accuracy and speed. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Tracking file removal and borrowing could be partially automated through barcode/RFID systems and database logging, but the task requires human judgment about whether returns are overdue and follow-up actions—and many organizations still use manual checkout systems. Current AI cannot reliably handle the full end-to-end workflow without significant human oversight and manual entry. |
| Task automatability | claude-sonnet-5 | 4/5 | Tracking checked-out files is a structured data-logging and reminder task well-suited to database/workflow software and simple automation, meeting the time-saving bar for most of the workflow. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Organizations have established policies and procedures for file management, and switching to new systems requires staff retraining and process redesign. Regulatory requirements in some sectors (legal, healthcare) mandate specific audit trails and accountability, which may limit full automation. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing or legal requirement mandates a human perform this tracking; main friction is organizational inertia and integration with legacy physical filing systems. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Implementing and maintaining automated file-tracking infrastructure (RFID, integrated databases, alerts) can be capital-intensive and requires ongoing integration with legacy filing systems. For small organizations, the total cost of ownership often exceeds the wages of a single part-time file clerk. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated tracking systems (barcode/RFID scanning, database logging, automated reminders) cost far less per transaction than manual clerk tracking, though initial setup and integration add some cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Document management systems and library-style check-out software exist and perform basic tracking in production, but they are often fragmented, require manual data entry, and rely on human compliance. No end-to-end AI system reliably enforces return accountability across diverse organizational settings. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Document management and library/records systems already provide check-out tracking, due-date alerts, and automated follow-ups in production use across many organizations. |
Eliminate outdated or unnecessary materials, destroying them or transferring them to inactive storage, according to file maintenance guidelines or legal requirements.
52CI 43–62 · exposure 58 · augmentation 63 · importance 4.0/5 · click for rater detail
Eliminate outdated or unnecessary materials, destroying them or transferring them to inactive storage, according to file maintenance guidelines or legal requirements.
52| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Adoption is slow in many sectors; legacy organizations still rely on manual file management, legal/compliance teams resist full automation due to liability concerns, and small to mid-size firms have low digitization. Large enterprises and finance/tech firms show faster adoption, but overall velocity remains below mainstream. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Records/document management sectors show moderate AI adoption for classification and retention tagging, but full pipeline automation including physical disposal remains pilot-stage in most organizations. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by flagging candidates for retention decisions, generating compliance reports, and accelerating the identification of outdated materials, allowing clerks to focus on judgment calls and legal exceptions rather than manual review. The human remains necessary for context and authorization. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-driven document classification and retention-schedule matching significantly speeds up identification of eligible materials, letting clerks focus on verification and execution. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Most of this task can be automated: AI systems can identify outdated materials using metadata, date rules, and retention schedules; robotic systems can physically move files to storage or destroy them; and workflows can apply legal/compliance rules. Setup and oversight are needed, but time savings and quality should exceed the 50% threshold in well-structured filing systems. |
| Task automatability | claude-sonnet-5 | 3/5 | AI/software can identify records meeting retention criteria and flag or route them for disposal, but physical destruction and final verification against legal holds still require human action or robotic integration not yet standard.br 50% time savings plausible for digital records but less so for physical files.br |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strong regulatory and legal barriers exist: record retention laws, litigation holds, compliance audits, and liability for improper destruction or transfer create high stakes. Most organizations require documented approval processes and human sign-off before destruction, making wholesale automation legally risky without substantial governance. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Legal/regulatory retention requirements and liability for improper destruction create moderate oversight burdens, though no license is required to perform the task itself. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | Automated document lifecycle management and robotic file handling are significantly cheaper than full-time file clerk labor once deployed, especially at scale. Cost per file processed is orders of magnitude lower than human handling, though initial system setup and legal review add overhead. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | For digital archives, automated retention flagging is cheap, but oversight, legal review, and physical destruction of paper records keep overall costs comparable to human clerks in many settings. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Document management systems and workflow automation products exist and perform parts of this task in production (metadata-based deletion, transfer workflows), but they typically require manual review for edge cases, legal holds, and context-dependent judgments about what is truly 'unnecessary' or compliant with retention rules. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Records management software with retention-rule automation exists and is deployed in many organizations for digital files, but comprehensive automated purging that reliably handles legal/regulatory nuance across mixed physical-digital archives is still narrow in scope. |
Place materials into storage receptacles, such as file cabinets, boxes, bins, or drawers, according to classification and identification information.
51CI 39–62 · exposure 53 · augmentation 25 · importance 3.9/5 · click for rater detail
Place materials into storage receptacles, such as file cabinets, boxes, bins, or drawers, according to classification and identification information.
51| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Adoption of robotic filing systems remains slow and concentrated in large enterprises, financial institutions, and government archives with high-volume standardized documents. Most small and mid-market organizations still rely on manual filing, indicating laggard sector characteristics for this task. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Clerical/administrative physical tasks are being replaced more by digitization than robotics, and physical file clerk roles are a shrinking, low-digitization niche with slow robotic adoption. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI assistance for file clerks is minimal; optical character recognition can help classify documents before filing, but the core task—placing items into receptacles—offers limited scope for human-AI collaboration that would materially improve clerk productivity without full automation. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can assist with classification/labeling decisions and inventory tracking, but offers little assistance for the physical placement action itself. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Physical robotics can reliably perform document placement into file cabinets and storage boxes given proper setup; however, the task requires some spatial reasoning and object handling that introduces minor complexity. Current robotic systems with vision and manipulation can achieve >50% time savings in warehouse/large-scale filing scenarios, though corner cases (unusual item shapes, tight spaces) may require human intervention. |
| Task automatability | claude-sonnet-5 | 3/5 | The classification/identification logic could be automated via digital systems, but the physical act of placing materials into physical receptacles requires robotic manipulation not yet reliable off-the-shelf; digital equivalents (e-filing) are highly automatable but this task as stated is physical. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | File clerks typically work in organizations with established workflows and human oversight norms; automating this task faces moderate friction from workflow integration, quality verification requirements, and organizational inertia rather than hard legal barriers. Some sectors (regulated document retention) may impose oversight requirements that slow adoption. |
| Adoption barriers | claude-sonnet-5 | 1/5 | No licensing or legal requirement mandates a human perform physical filing; it's a low-stakes clerical task with no regulatory protection. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current robotic systems for document filing require significant capital investment (hardware, integration, maintenance), and the per-task inference cost plus infrastructure overhead often exceeds a file clerk's loaded wage for routine filing. The economics favor automation only in high-volume, standardized scenarios. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Robotic solutions for this narrow physical task require capital investment in hardware and maintenance, often exceeding the cost of low-wage human labor for small-scale filing operations. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Robotic document handling and filing systems exist in production at scale in some logistics and archival settings, but they typically require controlled environments and well-structured inputs. General-purpose deployment across typical office filing scenarios remains narrow; most real-world implementations handle standardized document types and locations. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Warehouse/robotic sorting systems exist for standardized bins but general physical filing into cabinets/drawers with varied materials is not a mature deployed product for office file clerk contexts. |
Design forms related to filing systems.
45CI 34–56 · exposure 33 · augmentation 63 · importance 3.5/5 · click for rater detail
Design forms related to filing systems.
45| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | File clerk roles remain largely in traditional, lower-digitization sectors; pilot projects for AI-assisted form design exist but adoption in production remains rare outside large organizations with dedicated process improvement teams. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | File clerk roles are typically in low-digitization administrative settings with slow AI tool adoption for such niche, infrequent design tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist a clerk by suggesting form field structures, identifying redundancies, and proposing layout improvements, thereby improving design productivity while the clerk retains final judgment on applicability and compliance. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can quickly generate draft form layouts, field suggestions, and formatting options that a clerk can then refine, meaningfully speeding up the design process. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Form design requires creative judgment, understanding of workflow requirements, and iterative refinement—tasks where AI can suggest layouts and templates but rarely produces end-to-end solutions that meet the 50% time-saving threshold without substantial human revision. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can generate form layouts and field structures from a description, but aligning them to an organization's specific filing taxonomy, physical constraints, and compliance needs still requires human review and iteration.5 |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Organizational policies on documentation standards and the requirement for human sign-off on forms used across business units create moderate friction, though no strict legal licensing requirement binds this task to humans. |
| Adoption barriers | claude-sonnet-5 | 1/5 | There is no licensing, legal, or safety requirement mandating a human specifically design filing forms; it's a low-stakes administrative task. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 3/5 | An AI system for form design would require integration, training data, and human review, likely matching the cost of a clerk's time spent on this task rather than substantially undercutting it. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Using AI to draft a form template is very cheap compared to a human clerk's time spent designing and iterating on layouts, though some human review cost remains. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can generate form templates and suggest structural improvements, no deployed product reliably designs complete, context-appropriate filing system forms without human oversight; most applications are research-stage or narrow proof-of-concept tools. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | General-purpose generative and document-design tools can draft form templates, but no widely deployed product specifically automates filing-system form design as a reliable end-to-end workflow. |
Operate mechanized files that rotate to bring needed records to a particular location.
41CI 24–57 · exposure 41 · augmentation 13 · importance 3.6/5 · click for rater detail
Operate mechanized files that rotate to bring needed records to a particular location.
41| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | File clerk roles are concentrated in lower-digitization, smaller organizations (healthcare back-office, legal firms, small offices) that typically adopt automation slowly; large-scale digital document management has already displaced many file clerk roles rather than automating the task itself. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | File clerk roles and this specific mechanized retrieval task are in a low-digitization, declining physical process area with minimal AI adoption momentum; such carousel systems are largely legacy technology. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | Mechanized file systems can assist clerks by reducing retrieval time, but the core task is inherently manual/physical and offers limited opportunity for AI to augment human judgment or decision-making. |
| Augmentation potential | claude-sonnet-5 | 1/5 | AI offers no meaningful assistance to the physical act of operating a mechanized rotating file system; this is a manual/mechanical task outside AI's typical productivity-enhancing scope. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Operating mechanized files involves straightforward physical movement (rotating to retrieve records) based on location identifiers. AI-integrated robotic systems can reliably perform this retrieval at significantly higher speed and consistency than humans, meeting the 50% time-saving threshold with minimal setup in controlled environments. |
| Task automatability | claude-sonnet-5 | 2/5 | This is a physical operation task involving mechanized carousel/rotary filing equipment; AI software cannot directly operate physical machinery without robotics integration, so end-to-end automation via generally available AI is not feasible today.imit |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Physical task requiring no licensing; main barriers are organizational infrastructure constraints and preference to avoid capital expenditure rather than regulatory or legal prohibition of automation. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing or regulatory barrier exists, but the physical nature of operating mechanized equipment creates a practical barrier since it requires physical presence and interaction, not software substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Initial capital costs for mechanized retrieval automation are high, and integration expenses may outweigh labor savings for small to mid-size operations, making the all-in cost comparable to or exceeding human file clerk wages in many contexts. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Without a viable AI/robotic substitute for this physical machine-operation task, comparing cost is moot—any hypothetical robotic solution would be far more expensive than a human clerk performing this simple mechanical action. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Robotic filing systems and automated retrieval solutions exist in production environments (libraries, medical records, warehouses), but deployment typically requires specialized hardware integration and performance varies by file organization system complexity. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product operates mechanized rotary file systems; this remains a manual/mechanical interaction task with no commercial AI-driven robotic solution in production. |
Retrieve documents stored in microfilm or microfiche and place them in viewers for reading.
26CI 24–28 · exposure 16 · augmentation 13 · importance 3.6/5 · click for rater detail
Retrieve documents stored in microfilm or microfiche and place them in viewers for reading.
26| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Microfilm/microfiche use is declining across sectors; organizations are digitizing archives rather than automating legacy systems, placing this task in laggard, low-digitization contexts with minimal automation investment. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Microfilm/microfiche systems are a legacy technology in a declining niche with minimal digitization investment or AI adoption activity. |
| Augmentation potential | claude-haiku-4-5-20251001 | 1/5 | AI and automation offer minimal productivity assistance for manual microfilm retrieval and viewer loading; digital document management systems bypass the task entirely rather than enhancing it. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can assist in indexing, cataloging, or digitizing scanned microfilm content after retrieval, but offers little help with the physical retrieval and viewer-loading steps themselves. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Retrieving and loading microfilm/microfiche requires physical manipulation of fragile media and precise mechanical handling that current robotics struggle with in varied real-world conditions. While image capture and digital conversion are automatable, the core retrieval and placement task remains largely manual, offering only partial automation potential. |
| Task automatability | claude-sonnet-5 | 2/5 | This is a physical retrieval and placement task requiring locating a physical reel/card and inserting it into a machine, which current AI cannot perform without robotic embodiment.The information-lookup portion could be aided but the core physical act is not automatable today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few legal or regulatory barriers exist, but organizational inertia is moderate: institutions maintain microfilm systems for archival/compliance reasons and may lack incentive or capital to automate retrieval in aging workflows. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing or legal barrier exists, but physical infrastructure and low economic incentive to automate a shrinking, low-volume task create practical friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The specialized hardware required to robotically handle and load microfilm (precision positioning, gentle gripping, viewer interfacing) would be expensive to deploy and maintain, likely exceeding the wage cost of file clerks who perform this task infrequently in legacy workflows. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI-driven substitute for the physical handling involved, so any AI solution would require costly robotic infrastructure exceeding the cost of a human file clerk. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed commercial products reliably perform end-to-end retrieval and viewer-loading of microfilm/microfiche at scale. This is specialized physical manipulation with low market demand due to declining microfilm use, leaving only research or bespoke solutions. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product retrieves physical microfilm/microfiche and loads viewers; this remains a manual physical task with no robotic automation in production for this niche. |
Gather materials to be filed from departments or employees.
25CI 15–35 · exposure 13 · augmentation 38 · importance 4.2/5 · click for rater detail
Gather materials to be filed from departments or employees.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Adoption of autonomous material-gathering systems in file clerk roles is minimal; most organizations still rely on human staff for this task, reflecting the immaturity of embodied automation in office environments. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Clerical file management is a low-digitization, often physical task in sectors that have been slower to adopt AI-driven workflow automation compared to fully digital-native industries. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | Current AI offers minimal assistance for material gathering itself, though digital inventory or document-tracking systems can help humans locate what to gather—a peripheral rather than core augmentation of the gathering task. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-powered workflow tools and reminders can help track what needs to be gathered and from whom, improving efficiency, though the physical/manual collection step still requires human effort. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Gathering materials from physical departments and employees requires autonomous navigation, physical manipulation, and real-time human interaction in varied office environments. Current AI systems lack embodied robotics at scale and cannot reliably perform this end-to-end task across diverse organizational spaces. |
| Task automatability | claude-sonnet-5 | 2/5 | This requires physical collection of materials from various people and locations, or in digital contexts, requires integration into disparate systems, which current off-the-shelf AI cannot do end-to-end without significant infrastructure. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | While no formal license is required for AI to gather materials, there are practical barriers: physical infrastructure dependency, workplace safety concerns, and the need for human coordination to identify and locate materials reduce straightforward substitution. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing or legal barriers exist, but organizational friction (employees need to hand over materials, workflows vary by department) creates some practical resistance to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Autonomous robots capable of navigating offices and gathering materials remain significantly more expensive to deploy and maintain than paying file clerks; per-task cost is far higher than human labor. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | For physical gathering, AI offers no cost advantage since a human or robot must still physically retrieve items; for digital gathering, software automation may be cheaper but requires setup costs comparable to human labor initially. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No commercially deployed AI product reliably gathers physical materials from multiple departments autonomously. This requires integrated robotics, mapping, and human coordination—capabilities not yet in production at organizational scale. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some document management systems automate digital intake, but physically gathering materials from departments/employees is not something deployed AI products handle; it remains largely a human logistical task. |
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.