Government Property Inspectors and Investigators
13-1041.04Median wage $80,730/yr417,070 employed (US)Rank #612 of 923 scored · top 66% by substitution
Investigate or inspect government property to ensure compliance with contract agreements and government regulations.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution23
Exposure25
Augmentation53
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
13 rated tasks, binned by substitution score.
Position among all scored occupations
Distribution of 923 occupation scores; the marker is this occupation.
Tasks with substitution ≥ 70
0%
Run 1.0.0-draft.1 · computed 2026-08-05 · rater panel: claude-sonnet-5, claude-haiku-4-5-20251001 · intervals span rater disagreement.
Why this score
The five weighted dimensions of the composite, averaged across this occupation's tasks (importance-weighted, panel mean). Exact weights and formulas: /api/v1/methodology.
Task automatabilityw 35%27
panel mean rating 2.1/5 → substitution pressure 27/100
Technical feasibility todayw 20%21
panel mean rating 1.9/5 → substitution pressure 21/100
Cost vs. human wagew 15%27
panel mean rating 2.1/5 → substitution pressure 27/100
Adoption barriersw 20%inverted — strong barriers lower the score18
panel mean rating 4.3/5 (barrier strength) → substitution pressure 18/100
Sector adoption velocityw 10%17
panel mean rating 1.7/5 → substitution pressure 17/100
Task breakdown (13 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.
Examine records, reports, or other documents to establish facts or detect discrepancies.
55CI 43–67 · exposure 62 · augmentation 75 · importance 4.1/5 · click for rater detail
Examine records, reports, or other documents to establish facts or detect discrepancies.
55| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Government and compliance-heavy sectors show moderate adoption of document automation tools in pilots and limited rollouts, but full-scale production deployment remains slower than private finance or tech due to procurement cycles and risk-aversion. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Government sectors are typically slower adopters of AI due to procurement cycles, data sensitivity, and compliance requirements, resulting in limited production deployment compared to private-sector professional services. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI tools significantly augment human inspectors by rapidly scanning large document sets, flagging anomalies, and organizing findings—allowing inspectors to focus on higher-judgment tasks like interpreting discrepancies and making enforcement decisions. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can meaningfully assist by pre-screening documents, flagging anomalies, and summarizing records, allowing investigators to focus attention more efficiently while retaining final judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Document review and discrepancy detection is largely automatable using current AI systems for classification, extraction, and anomaly detection. End-to-end automation of fact-checking and cross-referencing documents can achieve >50% time savings, though complex contextual judgment may still require human review in some cases. |
| Task automatability | claude-sonnet-5 | 3/5 | AI document analysis tools can flag discrepancies and extract facts from structured or semi-structured records, but government inspection often involves varied, unstructured, or legally nuanced documents requiring contextual judgment that current systems handle imperfectly. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Government procurement rules, audit requirements, and the need for human sign-off on official findings create organizational friction and oversight requirements. However, no legal mandate requires a human to perform the initial document examination, only to validate conclusions. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Government investigators often require statutory authority, chain-of-custody rules, and accountability for findings used in legal or regulatory actions, creating strong institutional and legal barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-driven document review costs are substantially lower than human document analysts—inference and integration costs are typically 10-20% of a fully-loaded government inspector's wage for equivalent throughput. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI can cut review time significantly on digitized records, but integration, data security requirements, and human verification for legal/regulatory accuracy keep overall costs closer to comparable rather than order-of-magnitude cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Mature OCR, document classification, and data extraction products are deployed in production at scale in finance, compliance, and legal sectors. Anomaly detection and pattern-matching tools reliably flag inconsistencies, though integration with complex government workflows and legacy systems remains challenging. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Document review and anomaly-detection products (e.g., audit software, AI-assisted compliance tools) are deployed in some agencies, but broad reliable use across the diversity of government inspection records is still limited and error-prone. |
Prepare correspondence, reports of inspections or investigations, or recommendations for action.
53CI 46–60 · exposure 58 · augmentation 88 · importance 4.2/5 · click for rater detail
Prepare correspondence, reports of inspections or investigations, or recommendations for action.
53| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Government agencies adopt AI writing tools slowly due to regulatory compliance, audit requirements, and institutional conservatism; while some forward-looking agencies pilot AI assistance, widespread production deployment remains limited. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Government agencies, especially at state/local levels, tend to be slower adopters of AI tools due to procurement processes, data sensitivity, and legacy systems compared to private-sector professional services. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can substantially accelerate report and correspondence drafting by generating templates and initial text from inspection notes, allowing inspectors to focus on verification and judgment rather than clerical writing. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI is highly effective as a drafting and summarization aid, letting inspectors input raw findings and receive polished report language while retaining full control over final content and judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can draft inspection reports and correspondence from structured data and notes with significant time savings, but requires human review for accuracy, legal compliance, and sign-off on findings that may carry legal weight. |
| Task automatability | claude-sonnet-5 | 4/5 | Drafting correspondence and inspection reports from structured findings/notes is well within current LLM capability, and can achieve significant time savings when a human reviews and finalizes the output.rats |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Government inspection records often have legal standing and liability implications; many jurisdictions require a licensed or authorized inspector to certify findings, creating a hard requirement that a human must review and sign off on the final document. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Government reports often require the inspector's personal certification, signature, and legal accountability for accuracy, creating moderate liability and procedural barriers even though drafting itself isn't restricted. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI inference and integration costs are substantially lower than the loaded wage of a government inspector, especially when drafting reports that constitute bulk clerical output. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | AI drafting assistance costs a small fraction of an inspector's loaded hourly wage for producing report drafts, though human review and fact-verification still add cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Commercial writing tools and LLMs can generate inspection reports and correspondence, but real-world production use varies widely across government agencies; many still rely on manual drafting due to liability and regulatory concerns. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Government agencies increasingly use AI drafting tools and templates, but adoption in this specific investigative/inspection reporting context is uneven and often informal rather than a mature deployed product for this niche use case. |
Recommend legal or administrative action to protect government property.
37CI 20–55 · exposure 41 · augmentation 75 · importance 3.6/5 · click for rater detail
Recommend legal or administrative action to protect government property.
37| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Government agencies adopt AI slowly due to procurement, security, and oversight constraints; property inspection automation lags commercial sectors. Pilots exist, but production displacement remains limited. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Government inspection and investigation functions are typically slow-adopting sectors with cautious AI integration due to compliance and public accountability concerns. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can draft recommendation memos, flag relevant precedents, and surface policy gaps, substantially reducing the investigator's research burden while they focus on judgment and stakeholder coordination. Productivity gains for the human expert are significant. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully assist by summarizing case files, drafting recommendation language, and referencing relevant regulations, significantly speeding up the inspector's work while they retain final judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | AI can analyze property condition data, legal precedents, and regulatory requirements to generate actionable recommendations with minimal human oversight, saving 60–80% of the time spent on research and initial drafting. However, the final judgment call on proportionality and legal strategy typically requires human sign-off, preventing a full 5. |
| Task automatability | claude-sonnet-5 | 2/5 | This requires synthesizing investigative findings, legal context, and judgment calls about appropriate government action, which current AI cannot reliably do end-to-end without heavy human oversight.", |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Government procurement rules, legal liability for incorrect recommendations, and the requirement that an authorized official typically must sign off on formal legal/administrative actions create substantial friction. Delegation is constrained by authority and accountability frameworks. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Recommending legal or administrative action often requires accountability, authority, and liability tied to a credentialed government official, creating strong institutional and legal barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI inference and legal document processing are cheap relative to the billable time of a government attorney or senior investigator; end-to-end cost per recommendation is likely 1/5 to 1/10 of the human equivalent once systems are integrated. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | While AI can cheaply draft text, the actual value-add—defensible legal/administrative recommendations—still requires expert human review, keeping effective cost comparable to or only modestly cheaper than human labor. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Legal research tools and document analysis systems exist in production (e.g., contract analyzers, compliance checkers), but applying them specifically to government property protection recommendations is narrower and less proven at scale. Error rates on novel legal interpretations remain material. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously recommends legal/administrative actions for property protection in government inspection contexts; this remains a human judgment task supported at most by generic drafting tools. |
Investigate alleged license or permit violations.
25CI 25–25 · exposure 25 · augmentation 50 · importance 4.2/5 · click for rater detail
Investigate alleged license or permit violations.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Government agencies lag private-sector AI adoption; most inspectorates remain process-heavy and resistant to automation due to budgeting, change management, and established workflows. Pilots for document automation exist but widespread production deployment across jurisdictions is minimal. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Public sector inspection and enforcement functions are historically slow to adopt AI due to procurement cycles, legal constraints, and limited digitization of enforcement workflows. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist inspectors by organizing complaint data, flagging regulatory precedents, prioritizing cases, and summarizing documentation—raising efficiency without replacing the investigator's core judgment and legal authority. This augmentation is practical and partially deployed. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help investigators search records, summarize case files, cross-reference permit databases, and draft reports, meaningfully speeding up portions of the investigative process. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Investigating alleged violations requires gathering evidence, interviewing subjects, interpreting context-dependent regulations, and applying discretionary judgment. AI can assist with documentation review and flag patterns, but cannot independently conduct site visits, interviews, or make enforcement determinations that typically require human authority and judgment. |
| Task automatability | claude-sonnet-5 | 2/5 | Investigation requires gathering evidence, interviewing parties, site visits, and judgment calls about intent and context that current AI cannot reliably perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Government agencies typically require licensed or appointed inspectors/investigators to conduct official investigations due to legal authority, liability, and evidentiary requirements. Regulations often mandate human inspectors sign investigations, and the public nature of enforcement creates accountability and legal standing barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Government investigations typically require authorized personnel with legal standing to issue findings, subpoena records, or testify, and due-process/liability concerns necessitate human accountability. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI-assisted document review and flagging could reduce some administrative overhead, but investigations require specialized government personnel (inspectors, investigators) earning professional salaries. AI cost savings on partial automation do not offset the full cost of human investigators needed for legal authority and judgment. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI can cheaply screen records or flag discrepancies, but the bulk of investigative work (interviews, site visits, evidentiary judgment) still requires human labor, keeping overall cost comparable to or higher than human-only work when factoring integration and oversight. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Document scanning and initial triage are feasible, but no deployed product reliably handles the full investigation workflow: collecting site evidence, interviewing witnesses, interpreting ambiguous regulatory compliance, and recommending legal action. Pilots exist for record analysis but not end-to-end investigation deployment. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some products assist with document review or flagging anomalies in permit databases, but no deployed system conducts full violation investigations independently in production. |
Inspect government property, such as construction sites or public housing, to ensure compliance with contract specifications or legal requirements.
25CI 25–25 · exposure 25 · augmentation 50 · importance 4.0/5 · click for rater detail
Inspect government property, such as construction sites or public housing, to ensure compliance with contract specifications or legal requirements.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Government agencies are slow adopters of field automation; most inspections still rely on manual site visits and paper/basic digital documentation. Pilot programs exist, but production-scale AI deployment in public-sector compliance inspection remains minimal. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Government and construction sectors are traditionally slow adopters of AI, with pilots for site monitoring emerging but production deployment for compliance inspection still rare. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by analyzing photos, flagging potential defects, and generating draft reports, raising efficiency on data synthesis and preliminary assessment. However, the human inspector remains essential for on-site judgment, safety assessment, and legal accountability. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-powered image analysis, document review, and report drafting can meaningfully assist inspectors in preparing for and documenting site visits, though the physical inspection itself remains human-led. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Inspection requires physical presence at construction sites and detailed visual/tactile assessment of complex, context-dependent conditions. While AI vision systems can flag some defects in imagery, they cannot reliably walk jobsites, assess hidden compliance issues, or contextualize findings against contract specifications without substantial human guidance and follow-up. |
| Task automatability | claude-sonnet-5 | 2/5 | Physical site inspection requires on-site presence, sensory judgment, and interpretation of physical conditions against specifications, which current AI cannot perform end-to-end; AI can assist with documentation and analysis but not the core inspection act. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Government property inspection is heavily regulated; inspectors often must be licensed/certified, and their findings carry legal weight and liability. Contract compliance and code enforcement require human authority and sign-off, creating strong legal barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Government inspections often require certified/authorized personnel to make legally binding compliance determinations, creating liability and regulatory barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Inspectors command loaded wages of $60k–$80k+ annually; AI-assisted workflows still require human site visits, decision-making, and legal sign-off. Integration, drone/sensor hardware, and oversight costs mean AI assistance does not yet undercut human inspection labor on a per-task basis. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI tools (drones, computer vision) can reduce some inspection costs but still require human inspectors for judgment, verification, and legal sign-off, keeping overall costs comparable to or only modestly below human-only inspection. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI tools can analyze photos and documents but no deployed product reliably conducts end-to-end property inspections independently. Current systems in pilot phases (drone imagery analysis, automated report generation) require heavy human verification and cannot handle the full scope of legal/contractual compliance assessment at scale. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Products exist for image-based defect detection or drone-assisted site surveys, but no deployed system reliably performs full compliance inspection and legal determination in production at scale. |
Inspect manufactured or processed products to ensure compliance with contract specifications or legal requirements.
25CI 25–25 · exposure 25 · augmentation 50 · importance 3.9/5 · click for rater detail
Inspect manufactured or processed products to ensure compliance with contract specifications or legal requirements.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Government agencies and regulated industries are slow to adopt automated compliance systems due to liability concerns, procurement processes, and regulatory inertia. Adoption remains limited to pilot programs and narrow use cases rather than widespread deployment. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Public sector inspection and compliance functions are historically slow to adopt AI due to regulatory caution, procurement cycles, and physical/on-site requirements. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist inspectors through automated visual defect detection, documentation aggregation, and flagging of anomalies, thereby raising inspection speed and consistency. However, the human inspector must retain judgment over compliance determinations and legal sign-off. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist inspectors by flagging anomalies, cross-referencing specifications, and speeding documentation review, improving efficiency while the human retains final judgment and legal responsibility. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with documentation review and anomaly detection in images or sensor data, the task requires subjective judgment about compliance, legal interpretation, and discretionary decision-making that current systems cannot reliably perform end-to-end. Manual inspection of manufactured products and the verification against complex specifications remain substantially human-dependent. |
| Task automatability | claude-sonnet-5 | 2/5 | While AI vision systems can assist with detecting some defects or specification mismatches, this task involves complex legal compliance judgment, contract interpretation, and often physical inspection that current AI cannot fully replace end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Government regulatory frameworks typically require a licensed or authorized human inspector to certify compliance and bear legal responsibility. Liability, contractual authority, and statutory requirements for human sign-off create strong legal barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Government property inspection typically requires authorized, often sworn, personnel to certify compliance with legal/contract requirements, creating a strong liability and authorization barrier against full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI systems for visual inspection (cameras, anomaly detection) have significant upfront costs and require continuous human oversight and validation. The loaded cost of government inspectors is often lower than the combined infrastructure, integration, and human review needed for AI-assisted compliance systems. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI vision tools have some cost advantage for repetitive visual defect checks, but the need for human oversight, legal judgment, and liability for government sign-off keeps overall costs comparable to human inspectors. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Computer vision systems exist for defect detection in manufacturing, but production deployment for regulatory compliance inspection is narrow and requires human oversight. No mature product reliably performs the full legal/contractual interpretation and sign-off duties that government inspectors perform. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Machine vision and defect-detection products exist in manufacturing QA contexts, but deployed systems for full legal/contract compliance inspection with government authority are narrow and not widely used for this specific regulatory function. |
Inspect government-owned equipment or materials in the possession of private contractors to ensure compliance with contracts or regulations or to prevent misuse.
21CI 18–25 · exposure 20 · augmentation 50 · importance 3.7/5 · click for rater detail
Inspect government-owned equipment or materials in the possession of private contractors to ensure compliance with contracts or regulations or to prevent misuse.
21| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Government agencies typically move slowly on automation, and property compliance inspections involve federal contracts and regulations that create institutional inertia and risk aversion against substituting human judgment. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Government inspection functions are typically slow-moving, bureaucratic, and physically grounded, with limited AI agent deployment in production for this kind of oversight work. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by pre-screening contractor documents, flagging regulatory discrepancies, and organizing inspection checklists, raising inspector efficiency; however, the human must retain final judgment on equipment condition and compliance. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help inspectors by organizing contract terms, flagging discrepancies in records, generating checklists, and drafting reports, improving efficiency around the physical inspection task. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Inspection tasks require physical presence and judgment about equipment condition against specifications, which current AI cannot perform end-to-end. Document review and compliance checking could be partially automated, but the visual inspection and sign-off components require human presence and remain non-automatable. |
| Task automatability | claude-sonnet-5 | 2/5 | This requires physical, on-site inspection of equipment/materials at contractor facilities, which current AI cannot perform end-to-end; only documentation review or scheduling portions are automatable. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Government inspections have strong legal and regulatory barriers: a licensed/authorized government inspector typically must personally conduct and sign off on compliance verification, and liability for misuse of government property rests on that official's judgment. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Government contract compliance and property accountability often require an authorized inspector to certify findings, with legal/regulatory accountability that cannot currently be delegated to AI. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The specialized nature of government compliance inspections, plus required human oversight and liability concerns, means integrated AI solutions would still require significant human supervision, keeping total cost comparable to or exceeding direct human inspection. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI could cheaply assist with paperwork and record cross-referencing, but the core physical inspection still requires a human inspector, so overall cost savings versus the human-performed task are limited. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with document analysis and regulatory cross-referencing, no deployed system reliably performs the full inspection-and-compliance verification task. Visual inspection systems exist but lack the judgment and legal standing required for government compliance sign-off. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs physical property inspection and compliance verification autonomously; AI is at most a documentation or record-keeping aid in this workflow. |
Investigate applications for special licenses or permits.
21CI 18–25 · exposure 25 · augmentation 50 · importance 3.6/5 · click for rater detail
Investigate applications for special licenses or permits.
21| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Government agencies adopt AI slowly due to procurement complexity, legal risk aversion, and public accountability requirements; permit investigation remains largely manual in public sector workflows with minimal measured displacement. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Government sectors are typically slow adopters of AI due to procurement, compliance, and legacy IT constraints, resulting in limited production deployment. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist inspectors by pre-screening applications, flagging missing documents, and extracting key data, improving their review efficiency, though the investigative and decision-making tasks remain human-centered. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist by organizing application data, flagging inconsistencies, and speeding preliminary review, improving investigator efficiency without replacing judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Current AI can assist with data extraction and document review from applications, but investigating compliance and making discretionary permit decisions requires contextual judgment, site visits, and legal interpretation that AI cannot reliably handle end-to-end today. |
| Task automatability | claude-sonnet-5 | 2/5 | Parts of the application review (document completeness checks, cross-referencing databases) could be automated, but substantive investigation involving site visits, judgment calls, and legal interpretation still requires human involvement. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Government authority to issue licenses and permits is legally vested in certified inspectors and officials; regulatory frameworks and licensing laws typically require a licensed human to conduct investigations and sign off on decisions, creating hard barriers to automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Licensing decisions often carry legal and regulatory requirements mandating human authority sign-off, creating strong barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI tools for document processing may reduce some labor cost, but the overhead of integration, verification of results, and legal liability means total cost per permit processed remains comparable to or higher than human-only review. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI can cheaply process paperwork, but the investigative and judgment-heavy portions still require paid human investigators, keeping overall costs comparable to human-only processes. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No production AI system reliably performs permit investigation autonomously; products exist for document classification and flagging inconsistencies, but the core investigative and approval decision remains human-dependent in deployed government workflows. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some government agencies use software for intake and flagging, but no deployed product performs the full investigative process reliably; most systems only assist with data organization. |
Collect, identify, evaluate, or preserve case evidence.
18CI 15–20 · exposure 20 · augmentation 50 · importance 3.9/5 · click for rater detail
Collect, identify, evaluate, or preserve case evidence.
18| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Government agencies adopt evidence-handling technology slowly due to legal constraints, training requirements, and organizational conservatism. Pilots of AI assistance exist, but substantive autonomous deployment of evidence collection and preservation remains limited. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Government inspection and investigation functions are historically slow to digitize and adopt AI, with most current use limited to case management software rather than evidentiary automation. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist investigators by automating image documentation, cataloging, and pattern detection in evidence databases, raising efficiency in those components. However, the core judgment and custody-preservation work remains fundamentally human-driven. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help organize, tag, and search digital evidence or draft investigative reports, improving efficiency in analysis while humans retain control over collection and evaluation. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with image analysis and documentation of evidence, the task fundamentally requires trained investigators to handle chain-of-custody protocols, make complex forensic judgments, and physically collect evidence in ways that preserve legal validity. AI cannot perform the full task end-to-end with 50% time savings at equal quality. |
| Task automatability | claude-sonnet-5 | 2/5 | Physical evidence collection, chain-of-custody handling, and on-site identification require human presence, judgment, and legal accountability that current AI cannot perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Strong legal and regulatory barriers exist: chain-of-custody requirements, forensic standards, and admissibility rules mandate that licensed investigators perform or directly supervise evidence collection. Liability for improper preservation is severe and cannot be delegated to automated systems. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Evidence handling is governed by strict legal chain-of-custody, admissibility rules, and licensing/authorization requirements that mandate qualified human personnel for collection and preservation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI tooling (imaging, analysis software) requires significant integration and human oversight by trained investigators. The loaded cost of compliance, legal review, and the irreplaceability of proper human-conducted evidence handling means AI remains more expensive or at best comparable. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Human investigators remain necessary for physical collection and legal defensibility, so AI can only offset small analytical sub-tasks, limiting overall cost savings relative to the full task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Some AI-powered tools exist for evidence image analysis and document digitization in narrow contexts, but no mature product reliably performs the full evidence collection, identification, and preservation workflow autonomously. Error rates and legal liability concerns limit real-world deployment. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product autonomously collects or preserves legal/investigative evidence; at most AI assists with document review or data organization in research or narrow pilot contexts. |
Submit samples of products to government laboratories for testing, as required.
7CI 0–14 · exposure 8 · augmentation 25 · importance 3.8/5 · click for rater detail
Submit samples of products to government laboratories for testing, as required.
7| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Government agencies operate in low-automation, high-compliance sectors where procedural requirements and audit trails make substituting human submission with automated systems extremely slow to adopt, with minimal current deployment. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Government inspection and regulatory compliance functions involving physical evidence handling are a low-digitization, slow-adopting segment of public sector work. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could provide minor assistance through automated documentation, scheduling, or tracking of sample submissions, but the core task of physically submitting samples to laboratories fundamentally remains human-dependent. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can help with administrative aspects like generating submission forms, tracking sample status, or organizing documentation, but offers minimal help with the physical inspection and sampling process itself. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task involves physical logistics and chain-of-custody procedures that require human judgment about sample selection, proper handling, and regulatory compliance. Current AI systems cannot physically collect, package, or transport samples to laboratories. |
| Task automatability | claude-sonnet-5 | 2/5 | The physical act of collecting, packaging, labeling, and shipping product samples to a lab is a hands-on logistics task that AI cannot perform; AI could only assist with paperwork like chain-of-custody documentation or tracking submissions.dmc |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Government regulatory frameworks typically mandate documented chain-of-custody and verified submission by authorized personnel, creating legal requirements that a human inspector or authorized representative must physically sign off on or execute. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Government inspectors are typically authorized officials whose sample collection and chain-of-custody documentation carry legal weight for enforcement actions, requiring accountable human authorization. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The task requires physical presence and handling; any automation would require robotics integration far more expensive than paying a government employee to perform this submission work. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Since the core physical actions cannot be automated, AI provides no cost savings for the task as a whole; a human inspector's time and travel remain necessary. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI system can autonomously perform the physical submission of samples to laboratories, including proper documentation, chain-of-custody maintenance, and verification of receipt by authorized personnel. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product performs physical sample collection and submission; this remains a manual, field-based task requiring human handling of physical goods. |
Monitor investigations of suspected offenders to ensure that they are conducted in accordance with constitutional requirements.
0CI 0–0 · exposure 0 · augmentation 50 · click for rater detail
Monitor investigations of suspected offenders to ensure that they are conducted in accordance with constitutional requirements.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Government agencies move slowly on automation, especially for compliance-critical functions. This task involves constitutional protections where any error carries legal liability, so adoption of AI-driven monitoring remains minimal in practice. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Government investigative oversight is a highly regulated, low-digitization public sector function with minimal AI agent deployment for compliance monitoring roles. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI could assist by flagging procedural inconsistencies, summarizing case files, or highlighting high-risk investigation patterns, allowing human monitors to focus oversight more effectively. However, the core judgment function must remain with the human investigator. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help summarize case files, flag procedural inconsistencies, or search for precedent, providing moderate assistance while the human retains full oversight responsibility. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Monitoring investigations for constitutional compliance requires real-time judgment about legal standards, evidentiary chains, and rights protections in context-dependent scenarios. Current AI systems cannot reliably assess constitutional compliance or replace the human oversight function that is central to this task. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires legal judgment, oversight authority, and accountability for constitutional compliance that AI cannot exercise or be held responsible for; no end-to-end automation is plausible today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Constitutional oversight of investigations is a legally mandated duty that must be performed by authorized government personnel. Government accountability, liability for rights violations, and regulatory requirements around investigative conduct create hard barriers to automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | This is a government function requiring authorized personnel with legal accountability, sign-off authority, and constitutional/due-process obligations that legally require human judgment and liability. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The task requires specialized legal and investigative knowledge held by experienced government personnel. AI cost savings would be minimal given the need for expert human oversight to validate any AI assessment, making the all-in cost comparable to or higher than direct human performance. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this oversight function, so cost comparison favors the human role entirely; any AI cost would be additive, not substitutive. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product performs constitutional compliance monitoring of investigations as an autonomous system. While AI can flag certain patterns or assist with document review, actually monitoring investigations to ensure constitutional adherence requires legal judgment and accountability that remains research-stage. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs supervisory legal/constitutional oversight of investigations; this is a research-stage or nonexistent capability in production settings. |
Testify in court or at administrative proceedings concerning investigation findings.
0CI 0–0 · exposure 0 · augmentation 38 · importance 3.5/5 · click for rater detail
Testify in court or at administrative proceedings concerning investigation findings.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | This task involves mandatory human participation governed by law and courtroom procedure, making automation conceptually impossible and adoption velocity irrelevant. Government and legal sectors show no movement toward AI testimony. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Courtroom and administrative testimony processes are highly regulated, slow-changing, and show no meaningful movement toward AI-delivered testimony. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could marginally assist by organizing evidence summaries or preparing testimony outlines, but the core act of testifying must remain human-performed, limiting augmentation value to pre-testimony preparation only. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help investigators prepare testimony, organize findings, draft statements, and anticipate cross-examination questions, though the testimony itself remains unassisted in the moment. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Courtroom testimony requires human judgment, credibility assessment, cross-examination responsiveness, and legal standing that current AI cannot provide. This task is inherently human-centered and legally requires a person with firsthand knowledge to testify under oath. |
| Task automatability | claude-sonnet-5 | 1/5 | Testifying requires live, sworn human presence, real-time responses to cross-examination, and personal accountability that current AI cannot replicate or perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Testimony in court or administrative proceedings has absolute legal barriers: only a qualified human witness can testify under oath, be cross-examined, and satisfy rules of evidence. Law and regulation explicitly require human participation and legal standing. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Legal and procedural rules require sworn human testimony, oaths, personal accountability, and cross-examination capacity, creating hard legal barriers to substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | This task cannot be performed by AI at any price, making cost comparison meaningless. The human investigator must appear in person to satisfy legal and evidentiary requirements. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI alternative to compare cost against, since testimony must be given by the human investigator, making AI substitution cost inapplicable or infinitely higher. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI system can testify in court or administrative proceedings; the legal system requires a qualified human witness with personal knowledge and the ability to be sworn and examined. This remains entirely outside the scope of current AI capabilities. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product substitutes for a human witness testifying in court or administrative hearings; this remains firmly a human-only function today. |
Coordinate with or assist law enforcement agencies in matters of mutual concern.
0CI 0–0 · exposure 0 · augmentation 38 · importance 3.3/5 · click for rater detail
Coordinate with or assist law enforcement agencies in matters of mutual concern.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Government agencies operate under strict legal and operational constraints with slow digitization of inter-agency processes. Real adoption of autonomous AI for law enforcement coordination is negligible. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Government inspection and law enforcement liaison functions are slow-adopting, procedurally bound, and low-digitization sectors compared to fast-moving information sectors. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist by organizing case files or summarizing prior coordination history, but the core task of negotiating and committing agency resources requires human authority and judgment that AI cannot meaningfully augment. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with drafting communications, summarizing case files, flagging discrepancies, or organizing information shared with law enforcement, but the coordination and judgment itself remains human-driven. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Coordinating and assisting law enforcement requires interpersonal negotiation, judgment about agency priorities, legal discretion, and real-time decision-making in complex situations. Current AI systems cannot independently conduct these relationship-based, context-dependent negotiations or make binding decisions on behalf of government agencies. |
| Task automatability | claude-sonnet-5 | 1/5 | This task involves interpersonal coordination, judgment calls, and often sensitive information-sharing with law enforcement that requires human relationship management and legal accountability, not something current AI can execute end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Government coordination and law enforcement assistance are heavily regulated; only authorized personnel with specific credentials can legally represent agencies and make commitments. These hard regulatory and authorization barriers prevent substitution. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Coordinating with law enforcement often involves legal authority, chain-of-custody, confidentiality, and jurisdictional requirements that mandate accountable human government officials to act in this role. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI would require extensive human oversight and validation at each coordination step, making the combined cost higher than a property inspector performing the task directly with standard tools. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this coordination role, so the human cost remains the only real option, making the AI-to-human cost comparison moot in favor of humans. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product reliably performs inter-agency coordination or legal assistance to law enforcement today. This requires human judgment, authority, and accountability that production systems do not possess. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product performs interagency liaison or coordination functions; this remains a human relational and procedural activity. |
Related occupations — Business & Financial Operations
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.