Compliance Officers
13-1041.00Median wage $80,730/yr417,070 employed (US)Rank #170 of 923 scored · top 18% by substitution
Examine, evaluate, and investigate eligibility for or conformity with laws and regulations governing contract compliance of licenses and permits, and perform other compliance and enforcement inspection and analysis activities not classified elsewhere.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution40
Exposure40
Augmentation67
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
16 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
6%
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%40
panel mean rating 2.6/5 → substitution pressure 40/100
Technical feasibility todayw 20%42
panel mean rating 2.7/5 → substitution pressure 42/100
Cost vs. human wagew 15%48
panel mean rating 2.9/5 → substitution pressure 48/100
Adoption barriersw 20%inverted — strong barriers lower the score34
panel mean rating 3.7/5 (barrier strength) → substitution pressure 34/100
Sector adoption velocityw 10%41
panel mean rating 2.7/5 → substitution pressure 41/100
Task breakdown (16 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.
Collect fees for licenses.
81CI 79–84 · exposure 80 · augmentation 50 · importance 4.5/5 · click for rater detail
Collect fees for licenses.
81| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | State and federal licensing agencies, professional boards, and regulatory bodies have been rapidly adopting automated payment and license management systems over the past decade. Widespread digitization of licensing workflows shows high adoption in the public-sector compliance space. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Online and automated payment collection is already widely adopted across government agencies and licensing bodies, following broader fast digitization trends in transactional finance processes. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI systems can assist compliance officers by automating routine invoicing and payment tracking, flagging delinquent accounts, and generating reports, freeing staff to focus on exceptions and customer service. The assistance is meaningful but not transformative since the core task is primarily clerical. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI and automated systems can streamline invoicing, reminders, and reconciliation around fee collection, though the compliance officer's oversight role remains for exceptions and disputes. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | AI systems can handle most of the fee collection workflow end-to-end: identifying licensees, sending invoices, processing payments, and recording transactions. However, dispute resolution and manual follow-up on delinquent accounts may require human judgment, preventing a full 5-rating. |
| Task automatability | claude-sonnet-5 | 4/5 | Fee collection for licenses is a routine transactional task involving payment processing, invoicing, and record-keeping that off-the-shelf automated payment systems already handle with significant time savings. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | While some jurisdictions have legal requirements for human sign-off on regulatory decisions, fee collection itself is a clerical and administrative task with minimal legal barriers to automation. Most licensing authorities can substitute AI-driven systems without authorization hurdles. |
| Adoption barriers | claude-sonnet-5 | 2/5 | Some regulatory record-keeping and audit requirements exist around government fee collection, but the actual payment collection mechanism itself is not restricted to licensed humans. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Automated fee collection via integrated systems costs far less than manual invoicing, payment reconciliation, and record-keeping by compliance staff. The cost per collected fee is orders of magnitude lower than the loaded wage of a compliance officer. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Automated payment processing costs a small transaction fee compared to the loaded cost of a human manually collecting and recording payments, an order of magnitude cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Mature payment-processing and license-management systems are deployed in regulatory agencies and professional licensing bodies today, automating fee collection with minimal human intervention. Some edge cases and exceptions still require staff oversight, keeping it below 5. |
| Technical feasibility today | claude-sonnet-5 | 5/5 | Mature online payment portals, e-commerce checkout systems, and government licensing platforms already collect fees reliably at scale in production today. |
Prepare reports of activities, evaluations, recommendations, or decisions.
59CI 50–67 · exposure 62 · augmentation 75 · importance 4.3/5 · click for rater detail
Prepare reports of activities, evaluations, recommendations, or decisions.
59| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Financial services and larger enterprises are piloting AI-assisted compliance reporting, but adoption remains uneven; many compliance teams still rely on manual drafting. Regulatory uncertainty and liability concerns slow deployment compared to less-regulated sectors. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Financial services and professional compliance functions are piloting AI drafting and summarization tools, but full production deployment for regulatory reporting remains uneven. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI substantially assists compliance officers by drafting boilerplate sections, organizing findings, flagging patterns in data, and suggesting recommendation language. This augmentation meaningfully accelerates the writing phase while the officer retains control over substantive decisions and final approval. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI substantially speeds up drafting, summarizing findings, and organizing evaluations, letting compliance officers focus on judgment calls and final recommendations. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Current AI systems can draft compliance reports end-to-end from structured data and evaluation summaries, significantly reducing the time for composition, formatting, and organization of findings. While final review and human judgment on recommendations remain important, LLMs routinely achieve >50% time savings on report generation tasks in compliance contexts. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can draft structured compliance reports from source data and notes, but synthesizing evaluations and recommendations requires human judgment on regulatory context, so only part of the workflow meets the 50% time-saving bar today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Compliance officers' recommendations often carry legal and regulatory weight, requiring human accountability and sign-off; many jurisdictions and organizations mandate human certification of compliance findings. This creates moderate friction against full automation, though AI-assisted drafting faces few barriers. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Compliance reports often require sign-off by qualified officers and carry liability exposure if inaccurate, creating moderate but not absolute barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | The inference cost for generating a compliance report via current AI is substantially lower than the fully-loaded cost of a compliance officer's time spent drafting, reviewing structure, and formatting—likely 2–10× cheaper depending on report complexity and human wage. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI drafting reduces drafting time but licensed compliance staff must still review, verify, and finalize reports, keeping oversight costs significant relative to savings. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Mature products (enterprise LLM platforms, compliance-focused software with AI modules) reliably generate first-draft and near-final reports from input data. Some organizations have deployed these in production for routine compliance reporting, though complex legal recommendations still often require expert review. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Generative AI tools and document automation products are used to draft compliance reports in production, but they still require substantial human review and editing for accuracy and regulatory nuance. |
Prepare correspondence to inform concerned parties of licensing decisions or appeals processes.
58CI 49–67 · exposure 62 · augmentation 88 · importance 4.1/5 · click for rater detail
Prepare correspondence to inform concerned parties of licensing decisions or appeals processes.
58| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Financial and professional services sectors are moderately adopting AI drafting tools for compliance documentation, but deployment remains cautious and primarily augmentative. Regulatory conservatism and liability concerns limit rapid full-automation adoption, though pilots are increasing. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Compliance and legal-adjacent functions are adopting AI drafting tools steadily, but many regulated entities remain cautious about automating official notices. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI significantly assists compliance officers by generating initial drafts, ensuring regulatory language consistency, and reducing time spent on routine correspondence formatting and boilerplate elements. Officers review and personalize each letter, substantially raising throughput while maintaining human accountability and judgment. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI is highly effective at drafting first versions of decision letters and appeal notices, letting compliance officers focus on review, accuracy checks, and final approval. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can draft the bulk of routine correspondence (boilerplate licensing notifications, appeals process explanations) with significant setup, but requires human review for accuracy, legal sufficiency, and tone. Approximately 50% time savings is achievable with template-based generation and fact insertion, though the legal and regulatory stakes demand oversight. |
| Task automatability | claude-sonnet-5 | 4/5 | Drafting standardized correspondence conveying licensing decisions or appeal procedures is a well-structured writing task that current LLMs handle well when given case facts and templates, with human review for final send. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Correspondence on licensing decisions must often be signed by a licensed official or comply with strict regulatory formats; in many jurisdictions, the legal authority to issue decisions cannot be delegated to a non-human entity. Professional liability and potential regulatory non-compliance create strong organizational friction against full automation without human sign-off. |
| Adoption barriers | claude-sonnet-5 | 3/5 | While the officer's determination itself may require licensed judgment, formal correspondence often needs an authorized signature and adherence to regulatory notice requirements, creating moderate procedural friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI inference cost for generating correspondence templates is very low (cents per document), while compliance officer labor for this task costs tens to hundreds of dollars per letter. Even accounting for oversight and revision, AI-assisted generation is substantially cheaper than full human authorship. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | AI drafting drastically cuts time versus manual letter writing, and even with human review the all-in cost is substantially lower than a compliance officer authoring each letter from scratch. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Deployed products (LLMs, legal document automation platforms) can generate initial drafts reliably, but material error rates exist in regulatory citation accuracy and jurisdiction-specific requirements. Production use in compliance departments is emerging but typically requires substantial human review before sending to concerned parties. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Document generation tools and compliance software already produce templated notification letters and decision correspondence in production settings, though customization for edge cases still needs human oversight. |
Advise licensees or other individuals or groups concerning licensing, permit, or passport regulations.
57CI 31–82 · exposure 58 · augmentation 88 · importance 4.4/5 · click for rater detail
Advise licensees or other individuals or groups concerning licensing, permit, or passport regulations.
57| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Financial services, legal tech, and government agencies are rapidly deploying AI-assisted compliance tools for routine regulatory inquiries and document review; production adoption is measurable in information-intensive sectors. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Compliance functions in finance and regulated industries are adopting AI copilots at a moderate pace, with pilots common but full production reliance still limited. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI excels at augmenting compliance officers by instantly retrieving relevant regulations, drafting initial guidance, and flagging edge cases, allowing humans to focus on judgment-intensive or novel regulatory scenarios while maintaining accountability. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can quickly surface relevant regulations, draft explanatory responses, and flag compliance issues, meaningfully speeding up the officer's research and advisory process while the human retains final judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 5/5 | Current AI systems with retrieval-augmented generation can identify relevant licensing/permit regulations, extract key requirements, and generate compliant guidance with significant time savings. Rule-based compliance domains where regulations are documented and stable are among the highest-automation-potential tasks. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can retrieve and summarize regulations and draft advisory language, but tailored advice involving edge cases, jurisdictional nuance, and liability implications still requires human judgment, limiting full end-to-end automation. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Regulatory agencies and organizations typically require a licensed/accountable human to sign off on formal compliance advice, and liability exposure for incorrect guidance creates organizational friction against full automation, though AI-assisted advisory is widely adopted. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Many licensing and passport determinations require accountable, often credentialed compliance staff and carry regulatory and legal liability, creating strong barriers to full AI substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI inference cost for generating regulatory guidance is orders of magnitude cheaper than professional staff time; a single system can handle thousands of routine inquiries annually with minimal oversight overhead. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI can cheaply generate draft guidance, but the need for human verification and oversight to avoid costly compliance errors keeps blended costs roughly comparable to human-only advisory work. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed LLM-based compliance tools and document analysis systems operate in production for regulatory guidance in finance, healthcare, and legal sectors; however, liability concerns and the need for human review on high-stakes advice limit full autonomous deployment. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Chatbots and compliance-assistant tools exist for regulatory Q&A, but production deployments giving authoritative licensing/permit/passport advice with low error rates are narrow and not broadly trusted for final answers. |
Evaluate applications, records, or documents to gather information about eligibility or liability issues.
55CI 54–56 · exposure 50 · augmentation 75 · importance 4.4/5 · click for rater detail
Evaluate applications, records, or documents to gather information about eligibility or liability issues.
55| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Financial services, legal, and insurance sectors are rapidly adopting document AI and compliance automation tools; large organizations already deploy contract and application review systems, indicating strong production adoption momentum in regulated, information-dense industries. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Financial services, insurance, and legal compliance functions are actively piloting and increasingly deploying AI-assisted document review, but full-scale production replacement of judgment-based eligibility/liability assessment remains uneven across sectors. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI excels at document summarization, fact extraction, and flagging anomalies that boost a compliance officer's review speed and coverage; these systems are actively used in production to augment human evaluators, substantially raising their throughput while they retain judgment responsibility. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI substantially speeds up the gathering, summarization, and initial flagging of relevant information from applications and records, letting compliance officers focus their judgment on flagged edge cases and final determinations. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can extract and flag relevant facts from documents and records at significant speed, but evaluating eligibility and liability requires contextual judgment, weighing competing regulatory interpretations, and understanding edge cases that vary by jurisdiction and policy intent—tasks where human oversight remains essential for reliable output. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can extract, cross-reference, and flag information across applications and documents efficiently, but nuanced eligibility/liability determinations often require judgment calls and contextual interpretation that current systems handle unevenly, limiting full end-to-end automation to roughly half the workflow. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Compliance roles often involve regulatory responsibility and liability signoff that may formally or practically require a qualified human, though document review itself (the task stated) can be substantially delegated; organizational and professional norms create friction but not absolute legal barriers in most jurisdictions. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Many compliance decisions require sign-off from a designated compliance officer for regulatory and liability reasons, and errors carry legal/financial consequences, but the underlying data-gathering and initial screening work has fewer inherent licensing restrictions. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-powered document processing and triage can reduce manual review time substantially, with inference costs well below loaded human labor for initial screening and information gathering, though final evaluation still requires expert oversight. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated document extraction and rules-based flagging is dramatically cheaper per document than manual review once integrated, though compliance-grade oversight, exception handling, and audit trails add non-trivial cost overhead. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Document classification and extraction tools exist in production (contract analysis, form parsing), but reliable end-to-end evaluation of eligibility or liability across varied document types and complex regulatory contexts remains inconsistent; deployments typically require human review of flagged issues. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Document review and eligibility-checking products (e.g., in insurance underwriting, KYC/AML compliance) are deployed in production, but they still exhibit meaningful error rates and typically operate within narrow, well-defined rule sets rather than the full breadth of compliance evaluation. |
Keep informed regarding pending industry changes, trends, or best practices.
50CI 41–59 · exposure 42 · augmentation 88 · click for rater detail
Keep informed regarding pending industry changes, trends, or best practices.
50| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Compliance and financial-services firms have adopted AI-powered regulatory monitoring tools and trend alerts, but deployment remains uneven; many organizations still rely on traditional manual scanning and industry newsletters rather than integrated AI systems. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Compliance and legal/professional services sectors are adopting AI monitoring and research tools at a moderate pace, with pilots and partial deployment common but full automation of this continuous task still rare. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI systems substantially enhance compliance officers' productivity by filtering noise, categorizing updates by relevance, and surfacing emerging patterns across thousands of documents, allowing officers to focus analysis and judgment rather than raw information gathering. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI dramatically improves an officer's ability to scan, filter, and summarize vast regulatory and industry information streams, significantly boosting productivity while the officer retains interpretive responsibility. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI can identify and summarize published industry trends and regulatory changes through document scanning and news aggregation, but this task requires contextual judgment about relevance to specific organizational risk profiles and interpretation of subtle shifts in regulatory intent—tasks that currently need human oversight. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can aggregate, summarize, and flag regulatory news and trend reports at scale, but synthesizing relevance and implications for a specific organization still requires human judgment, so only partial time savings are realized end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | While no strict licensing barrier prevents AI from monitoring trends, organizations face pressure to have a human compliance officer formally accountable for staying abreast of changes, and liability concerns around missed critical updates create friction against full automation. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No legal requirement mandates a human personally track industry trends, though compliance officers are often required to demonstrate personal currency with regulations for licensing/certification purposes, creating mild friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | Automated monitoring and AI-powered summarization of regulatory changes and industry trends is substantially cheaper than paying compliance officers to manually scan sources full-time, though some human review time remains necessary. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | Subscription-based AI monitoring tools are cheaper than dedicated research staff hours, but licensing costs plus required human oversight keep the ratio moderate rather than an order-of-magnitude savings. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Multiple products exist that monitor regulatory updates and industry news (e.g., compliance monitoring platforms, news aggregators with AI summaries), but they require significant human filtering and validation to extract actionable insights specific to a given organization's risk exposure. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Products like news aggregators, regulatory intelligence platforms (e.g., Compliance.ai, Thomson Reuters Regulatory Intelligence) with AI summarization exist and are used in production, but they still require human review and curation to be reliable. |
Warn violators of infractions or penalties.
49CI 31–67 · exposure 50 · augmentation 75 · importance 4.4/5 · click for rater detail
Warn violators of infractions or penalties.
49| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Large regulated entities (financial services, healthcare, utilities) have begun deploying automated notice systems, but adoption remains mixed; smaller organizations and public sector agencies still rely heavily on manual processes. Production deployments are increasing but not yet dominant. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Compliance functions are adopting AI for monitoring and drafting at a moderate pace, with pilots common but full-scale autonomous violation notices still rare. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI dramatically assists compliance officers by drafting notices, cross-referencing regulation codes, calculating penalties, and organizing violation data, leaving the officer to review, personalize, and authorize. This augmentation significantly accelerates the warning workflow while the human retains final accountability. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can significantly speed up drafting, flagging violations, and suggesting penalty language, greatly boosting officer productivity while the officer retains final authority. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Automated systems can reliably generate and send violation notices with legal language, reference codes, and penalty tiers based on rule databases. While final human review of high-stakes cases is often required, the core task of warning violators (identifying the infraction, calculating penalties, and composing notices) is highly automatable at >50% time savings. |
| Task automatability | claude-sonnet-5 | 2/5 | Drafting warning notices can be automated, but deciding when a warning is warranted, tailoring tone/legal nuance, and delivering it in context of authority requires human judgment not fully replaceable end-to-end today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Regulatory bodies often require human authorization or sign-off on enforcement actions, and some jurisdictions mandate that a licensed compliance professional review notices before issuance. Organizational inertia and auditability requirements create moderate friction, though the barrier is not absolute. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Warnings often carry legal or regulatory weight and require an authorized officer's judgment and accountability, creating strong institutional and liability barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | Automated warning systems cost a small fraction of human compliance officer time when amortized per violation; inference and integration costs are low, and oversight is minimal for routine infractions. The cost per notice is likely 10–100x lower than manual drafting and dispatch. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI can cheaply draft notice language, but the overall task still requires human review and authority to issue, keeping blended cost roughly comparable to human-only workflows. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Multiple deployed products in compliance, legal tech, and regulatory software already generate violation notices, send automated warnings, and track penalty calculations in production environments. Systems exist across finance, environment, labor, and tax domains, though some require human approval before issuance. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some compliance software can auto-generate notices or flag violations for review, but no mature product independently issues authoritative warnings to violators without human sign-off. |
Administer oral, written, road, or flight tests to license applicants.
41CI 0–82 · exposure 45 · augmentation 50 · importance 4.4/5 · click for rater detail
Administer oral, written, road, or flight tests to license applicants.
41| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Adoption is accelerating in professional licensing, education, and remote testing contexts post-pandemic. Many licensing bodies, driving test administrators, and exam providers have already deployed or are piloting AI-driven systems, though full road/flight automation lags. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Government licensing and testing agencies are slow-moving, highly regulated, and physically embedded, showing minimal AI adoption for this specific function. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI augments human examiners by automating scheduling, question generation, real-time scoring, and feedback, significantly reducing administrative burden. A human examiner can focus on proctoring integrity and edge-case judgment while AI handles routine test delivery and recording. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can help with scheduling, scoring written tests, or generating test questions, but offers little assistance during the actual road or flight evaluation itself. |
| Task automatability | claude-haiku-4-5-20251001 | 5/5 | AI systems can now generate, administer, and score written tests automatically with high reliability, and can conduct oral examinations via conversational AI interfaces. Road and flight tests involve physical observation, but the administrative, scoring, and record-keeping components (the core of 'administer') are fully automatable, meeting the 50% time-saving threshold. |
| Task automatability | claude-sonnet-5 | 1/5 | Administering road or flight tests requires real-time physical observation, judgment of human performance in dynamic environments, and safety oversight that current AI cannot perform end-to-end.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | While test administration itself has limited legal barriers, many jurisdictions require licensed human examiners for road and flight tests, and liability concerns around test validity and fairness create organizational friction. However, written and oral components can be automated with less regulatory resistance. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Licensing tests are legally required to be administered and certified by authorized human examiners under government regulation, creating a hard legal barrier to automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI-driven test administration costs (automated invigilation, scoring, record management) are orders of magnitude cheaper than paying human examiners per applicant, with negligible marginal cost per additional test once systems are deployed. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute for in-vehicle or in-flight test administration, so no meaningful cost comparison exists; human examiners remain the only option. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Mature products exist for automated written test administration, scoring, and grading in educational and licensing contexts. Oral testing via AI chatbots is deployed in some licensing contexts, though road/flight test observation automation remains limited. Most elements of the task have production-grade tooling, though some modalities (proctored vehicle/aircraft testing) still require human oversight. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product administers actual road or flight license tests; written/oral test delivery may use software but the core evaluative task remains human-performed and legally mandated. |
Identify compliance issues that require follow-up or investigation.
37CI 32–41 · exposure 42 · augmentation 75 · click for rater detail
Identify compliance issues that require follow-up or investigation.
37| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Financial services and large enterprises have piloted AI-driven compliance monitoring and are deploying it at scale for transaction screening and policy enforcement. However, adoption remains uneven across smaller firms and less-regulated industries, and most organizations still pair AI tools with mandatory human review rather than relying on automation alone. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Financial services and other regulated industries have adopted AI-based monitoring and RegTech tools at a moderate pace, with pilots and partial production use common but full autonomous identification still rare. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI dramatically augments compliance officers by automating initial document review, flagging anomalies in transaction data, and surfacing patterns that would require hours of manual inspection. The human remains in the loop for judgment calls and formal investigation, but AI transforms the speed and breadth of issue identification, allowing officers to focus on complex or ambiguous cases. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI substantially augments compliance officers by surfacing anomalies, summarizing large data sets, and prioritizing alerts, meaningfully increasing productivity while the officer retains judgment over escalation. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can identify potential compliance issues through pattern recognition and rule-based flagging of documents, transactions, or communications against known regulatory frameworks, achieving meaningful automation of initial triage. However, the nuanced judgment required to assess whether an issue truly requires follow-up or investigation often demands contextual understanding, precedent knowledge, and business-specific factors that current AI systems handle inconsistently, limiting full end-to-end automation below the 50% time-saving bar. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can flag anomalies and pattern-match against rules, but determining which issues truly warrant investigation requires contextual judgment, materiality assessment, and organizational knowledge that current systems handle unreliably at scale. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Compliance work operates under regulatory scrutiny; organizations face liability if automated systems miss violations, and many jurisdictions require a licensed compliance officer to document, sign off, or formally investigate findings. Additionally, regulatory bodies and internal audit functions often expect human judgment in compliance determinations, creating institutional and legal friction against full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Regulatory frameworks (e.g., BSA/AML, SEC, industry-specific compliance regimes) often require designated compliance officers to exercise judgment and sign off on escalations, creating strong liability and licensing-related barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | While inference costs for AI screening are low, the end-to-end cost includes integration with legacy compliance systems, model tuning, and continuous human oversight to validate and investigate flagged issues. This overhead keeps total cost per investigation roughly comparable to or modestly cheaper than a junior analyst performing initial triage, with ongoing licensing expenses. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI screening tools reduce volume of manual review but still require skilled compliance staff to validate alerts and investigate context, so overall cost savings versus a human-only process are moderate rather than order-of-magnitude. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Compliance monitoring products using AI exist in production (e.g., transaction monitoring, policy violation detection), but they suffer from material false-positive rates and narrow scope to specific regulatory domains or data types. Deployed systems require significant tuning and human oversight; they reliably screen for known issues but miss emergent or cross-cutting compliance problems. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Transaction monitoring, AML, and compliance analytics products deployed in production do flag potential issues, but they generate high false-positive rates requiring substantial human triage, so they narrow rather than replace the identification task. |
Provide assistance to internal or external auditors in compliance reviews.
29CI 25–32 · exposure 30 · augmentation 75 · click for rater detail
Provide assistance to internal or external auditors in compliance reviews.
29| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | While large financial institutions pilot AI-assisted compliance tools, most sectors and smaller firms rely on traditional audit teams; broad production deployment remains limited and manual review remains the standard. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Compliance and audit functions in finance and professional services are moderately adopting AI for document review and analytics, though human-mediated audit assistance remains largely manual in production. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI excels at accelerating document preparation, flagging anomalies, and organizing evidence for human reviewers, materially improving auditor productivity in evidence gathering and preliminary analysis phases of compliance reviews. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can significantly speed up locating relevant records, drafting responses, and summarizing compliance data for auditors, meaningfully boosting the compliance officer's productivity while they remain the primary interface. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Current AI can help gather and organize documentation and flag potential issues, but compliance reviews require nuanced judgment of regulatory interpretation, organizational context, and risk prioritization that AI cannot fully automate end-to-end with equal quality. |
| Task automatability | claude-sonnet-5 | 2/5 | This task involves gathering documents, explaining processes, and answering nuanced questions from auditors, requiring contextual judgment and interpersonal interaction that current AI cannot fully replace end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Regulatory bodies (SEC, SOX, GDPR authorities) often require a qualified human auditor to attest to compliance reviews; liability for missed violations rests with the organization and typically demands human professional judgment and sign-off. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Compliance functions often require designated accountable personnel, and auditors typically expect a responsible human point of contact for attestations and sign-offs, creating strong organizational and regulatory friction against full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI document processing is cheaper per unit, but integration costs, continuous tuning for regulatory changes, and required human verification overhead keep total cost-per-review relatively high compared to human auditors. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | While document retrieval and summarization can be cheaply automated, the human judgment, negotiation, and accountability components still require costly compliance staff time, keeping overall cost comparable to human-only work. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Products exist for document review and anomaly detection in compliance workflows, but they operate with material error rates and require significant human oversight; no deployed system reliably performs full compliance review assistance independently. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI tools can help retrieve and summarize records or flag anomalies, but no deployed product autonomously handles the full liaison and explanatory role compliance officers play with auditors. |
Verify that all firm and regulatory policies and procedures have been documented, implemented, and communicated.
26CI 25–28 · exposure 25 · augmentation 63 · click for rater detail
Verify that all firm and regulatory policies and procedures have been documented, implemented, and communicated.
26| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Compliance functions are conservative and risk-averse; while organizations use AI for narrow document-scanning tasks, end-to-end policy verification and sign-off remain manual. Pilot programs are emerging but production displacement is minimal. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Financial services and other regulated sectors are adopting AI-assisted compliance monitoring tools at a moderate pace, though full verification workflows remain largely human-driven pilots. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI assists meaningfully by automating document scanning, cross-referencing policies, and generating gap reports, which raises an officer's productivity in review cycles. However, the final verification judgment and accountability remain with the human. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can significantly speed up document review, gap analysis, and communication tracking, greatly aiding compliance officers even though final verification and sign-off remain human tasks. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Compliance verification requires checking complex, context-dependent organizational documentation and understanding regulatory frameworks across multiple domains. While AI can scan documents and flag inconsistencies, the nuanced judgment of whether policies are properly 'implemented' and 'communicated' in practice demands human oversight, and current systems cannot reliably achieve 50% time savings end-to-end. |
| Task automatability | claude-sonnet-5 | 2/5 | Involves judgment-based verification across documentation, implementation, and communication which requires cross-referencing organizational context and regulatory nuance that current AI cannot fully autonomously validate and certify. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Compliance verification sits within heavily regulated sectors where documented human sign-off, audit trails, and organizational accountability are often legally required; liability for certification errors is material and asymmetric, creating strong institutional friction against full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Regulatory frameworks generally require a designated compliance officer to attest to policy implementation, creating a strong human-accountability barrier that AI cannot legally satisfy. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The all-in cost of AI systems (inference, integration, regulatory audit trails, mandatory human oversight) is currently comparable to or exceeds the loaded cost of a compliance officer performing spot checks or systematic reviews, especially given liability concerns. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI tools can cheaply scan documents but the oversight, liability review, and sign-off still require compensated compliance staff, keeping overall cost comparable to human-led verification. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed product reliably performs complete compliance verification across a firm's full policy landscape at scale. Existing tools assist with documentation auditing and gap detection but require substantial human review, interpretation of regulatory context, and validation that communications were effective. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some compliance software can flag policy gaps or track document versioning, but no deployed product reliably performs the full verification and attestation process end-to-end without significant human review. |
Score tests and observe equipment operation and control to rate ability of applicants.
25CI 25–25 · exposure 25 · augmentation 50 · importance 4.3/5 · click for rater detail
Score tests and observe equipment operation and control to rate ability of applicants.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Compliance and safety-regulated sectors (aviation, nuclear, manufacturing) move cautiously on automation of assessment tasks due to regulatory and liability constraints; adoption remains limited to clerical scoring aids rather than the full observation-and-rating task. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Compliance and licensing/testing sectors tend to be slow-moving, regulation-bound, and cautious about replacing human judgment in certification, so AI adoption for this specific task is limited. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by pre-scoring written components, flagging anomalies in recorded equipment data, and organizing results, moderately raising productivity for the compliance officer while they retain decision authority on applicant ratings. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist by automating scoring of written/computerized test components and flagging anomalies, but the physical observation portion sees limited AI-based augmentation. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Some parts can be automated (scoring standardized written tests), but the core task—observing equipment operation, assessing control ability, and rating applicant competence—requires real-time judgment of physical performance and safety that current AI systems struggle to do reliably without human oversight. |
| Task automatability | claude-sonnet-5 | 2/5 | This task combines objective test scoring (partially automatable) with real-time observation of physical equipment operation and human control quality, which requires situated judgment AI cannot yet fully replicate.$ |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Regulatory frameworks often mandate that licensed compliance officers directly observe and certify applicant competency; liability for failures in safety-critical equipment operation falls on the human overseer, creating strong legal and organizational barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Certification and licensing evaluations often carry legal/regulatory requirements that a qualified human examiner sign off on applicant competency, creating strong liability and authorization barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI can reduce overhead on test scoring, but the substantial human labor needed for in-person observation, performance assessment, and compliance sign-off means overall cost savings remain marginal compared to trained human compliance officers. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Scoring standardized components can be cheap via software, but the observational/physical evaluation portion still requires a human evaluator, keeping blended costs closer to human-level. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While automated test-scoring systems exist for multiple choice components, the observational and rating aspects of applicant performance during live equipment operation remain primarily human-dependent in production compliance settings; no deployed product handles the full task end-to-end reliably. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Automated scoring systems exist for standardized written tests, but no deployed product reliably observes and rates hands-on equipment operation performance in production compliance settings. |
Confer with or interview officials, technical or professional specialists, or applicants to obtain information or to clarify facts relevant to licensing decisions.
25CI 25–25 · exposure 25 · augmentation 63 · importance 4.1/5 · click for rater detail
Confer with or interview officials, technical or professional specialists, or applicants to obtain information or to clarify facts relevant to licensing decisions.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Compliance and regulatory sectors are moderately digitized but cautious; AI adoption in licensing interview workflows remains nascent, with most organizations still relying on human officers despite available document-processing tools. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Government and regulatory compliance functions are typically slow adopters of AI for interview-based fact-finding due to procedural rigidity and accountability concerns. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist by drafting interview templates, flagging missing information, summarizing specialist responses, and highlighting inconsistencies in applicant statements—meaningful augmentation without replacing the human interviewer's judgment. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can significantly assist by preparing interview questions, transcribing and summarizing conversations, flagging inconsistencies, and organizing information for the compliance officer's review. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can extract and summarize information from documents and conduct preliminary fact-checking, the task critically requires real-time interaction, judgment about credibility, and nuanced clarification of complex technical facts—capabilities that current AI lacks for autonomous end-to-end performance at parity with human outcomes. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can assist with drafting interview questions and summarizing responses but conducting nuanced interviews to clarify facts and assess credibility requires human judgment and interpersonal skill that current AI cannot reliably replicate end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Licensing decisions often carry legal and regulatory weight; many jurisdictions require a licensed or authorized officer to conduct interviews and sign off on fact-finding, creating substantial liability and compliance barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Licensing decisions often require accountable human officials to interview applicants and specialists, with legal and procedural requirements for due process and professional judgment, creating strong regulatory and liability barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI-assisted interview preparation and fact summaries are cheap, but the human compliance officer remains essential; the marginal cost savings from AI do not yet offset the loaded human wage for this specific interviewing and judgment task. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Human compliance officers must still conduct these interviews; AI can reduce prep and documentation time but cannot replace the interview itself, so overall cost savings are limited. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed system reliably conducts unscripted interviews or conferences with specialists to obtain and clarify licensing-relevant information at scale; AI can assist with note-taking or preliminary questions, but human judgment and interaction remain essential. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI transcription and chatbot intake tools exist, but no deployed product reliably conducts substantive fact-clarifying interviews with officials or applicants for licensing decisions in production. |
Issue licenses to individuals meeting standards.
24CI 20–29 · exposure 25 · augmentation 63 · importance 4.5/5 · click for rater detail
Issue licenses to individuals meeting standards.
24| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Government and heavily regulated sectors adopting this task move slowly on automation due to legal risk, political oversight, and public accountability concerns. Adoption remains confined to pilots and narrow ancillary use cases rather than production displacement. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Government and regulatory licensing bodies are historically slow adopters of AI due to legal, procurement, and public-accountability constraints, with pilots more common than full deployment. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist by automating document collection, flagging missing criteria, cross-referencing standards, and drafting preliminary compliance summaries, allowing the compliance officer to focus on judgment and exceptions. This assistive capability is actively deployed in some organizations. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can significantly speed up document review, cross-checking credentials against databases, and flagging incomplete applications, meaningfully aiding the human officer's workflow. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Issuing licenses requires verifying complex, context-dependent criteria against regulatory standards and often demands human judgment on edge cases. Current AI can assist with data entry and initial document verification, but cannot reliably make discretionary approval decisions or handle exceptions that represent a meaningful portion of the task. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can check documentation and eligibility criteria against rules, but final license issuance often requires verification of authenticity, judgment on edge cases, and legal accountability that current systems cannot fully replace end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Licensing authority is typically vested in licensed professionals (compliance officers, attorneys, or government agencies) by statute, and final approval often requires human attestation. Regulatory frameworks and liability frameworks create substantial legal and organizational barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Issuing licenses is typically a statutorily authorized government or regulatory function requiring a designated human official's legal authority and signature, creating a hard regulatory barrier. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The cost of AI systems with sufficient oversight, error-checking, and human review integration approaches or exceeds the loaded cost of a compliance officer handling routine license reviews, especially when regulatory liability is factored in. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI can cheaply automate document intake and eligibility screening, but the human sign-off, audit trail, and liability management keep overall costs closer to parity with human processing. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While some regulatory bodies have deployed automated initial screening systems, end-to-end license issuance in production typically requires human review and sign-off. No widely deployed product performs this task autonomously with the reliability required for legal compliance across varied regulatory domains. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some government and professional bodies use software to pre-screen applications, but actual license issuance is still performed or formally authorized by human officers in production systems today. |
Report law or regulation violations to appropriate boards or agencies.
18CI 11–25 · exposure 17 · augmentation 63 · importance 4.2/5 · click for rater detail
Report law or regulation violations to appropriate boards or agencies.
18| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Compliance functions remain heavily regulated and risk-averse; organizations use AI mainly for preliminary data gathering and report drafting, not autonomous submission to regulators. Adoption of AI-driven end-to-end violation reporting is slow and limited to narrow, well-defined violation categories. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Compliance functions in regulated industries are cautious adopters of AI for consequential legal actions, with pilots common for monitoring but actual autonomous reporting rare. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by analyzing voluminous data to surface potential violations, flagging patterns, and drafting initial report language, which helps the officer review and decide which violations to report. However, the augmentation is limited because the officer retains full decision and signature authority. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI significantly aids by scanning transactions/communications for anomalies, summarizing regulations, and drafting report language, greatly increasing officer efficiency while the human retains final judgment and accountability. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Reporting violations to regulatory bodies requires judgment about which violations are material, what agency has jurisdiction, and how to frame findings—discretionary steps that depend on legal interpretation and professional liability. Current AI cannot reliably make these determinations without human legal review, so automation would not achieve 50% time savings at equal quality. |
| Task automatability | claude-sonnet-5 | 2/5 | Deciding when and how to report violations requires professional judgment, legal interpretation, and accountability that current AI cannot reliably exercise end-to-end; AI can help draft or flag issues but not independently perform the reporting act with equal quality. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Regulatory agencies require reports to be signed by a responsible official (often the compliance officer) who certifies accuracy and completeness. Legal liability for false, misleading, or omitted reports falls on the organization and officer, creating a hard barrier against full automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Reporting violations often carries legal liability and requires an authorized, accountable individual (e.g., licensed compliance officer) to certify and submit reports to regulators, a hard institutional and legal barrier. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI tooling (document analysis, form-filling) may reduce drafting time, but the compliance officer must still review, validate, and take responsibility for submissions. The loaded cost of a compliance officer remains the dominant expense, and AI cost savings are modest against that baseline. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Because human review, judgment, and legal accountability remain necessary, AI reduces some drafting/research cost but doesn't replace the oversight cost, keeping totals closer to comparable rather than order-of-magnitude cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can draft boilerplate violation reports and identify agency contact information, no deployed product reliably performs the full task of determining violation materiality, selecting appropriate agencies, and producing legally defensible reports without expert human oversight. Compliance reporting involves legal exposure that organizations do not delegate to AI systems. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some compliance software can flag anomalies or generate draft reports, but no deployed product autonomously determines violations and files formal reports to regulatory bodies without human sign-off. |
Visit establishments to verify that valid licenses or permits are displayed and that licensing standards are being upheld.
5CI 0–10 · exposure 5 · augmentation 38 · importance 4.4/5 · click for rater detail
Visit establishments to verify that valid licenses or permits are displayed and that licensing standards are being upheld.
5| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Compliance and regulatory inspection remains a heavily regulated, human-centric function with legal accountability requirements. Adoption of autonomous AI systems for these tasks is minimal, with only incremental digitization of reporting rather than automation of inspections themselves. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Government and regulatory inspection sectors are typically slow-moving, low-digitization environments with minimal AI-driven field automation. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist by pre-screening establishment records or flagging documents for review, but the core task of on-site verification and judgment remains fundamentally human-driven. Limited augmentation potential given the statutory nature of the role. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with scheduling visits, generating checklists, analyzing photos/documents collected during visits, and drafting reports, improving efficiency around the core physical task. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires on-site physical inspection, visual verification of documents, and judgment about compliance with contextual standards. Current AI cannot reliably conduct in-person visits or make nuanced compliance determinations without human oversight. |
| Task automatability | claude-sonnet-5 | 1/5 | Requires physical presence at an establishment to visually inspect displayed licenses and observe on-site conditions, which current AI cannot perform without robotic embodiment.: |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Regulatory frameworks typically mandate that a licensed compliance or inspections officer personally verify licensing standards, and many jurisdictions have legal requirements that a human official conduct establishment inspections and sign off on compliance. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Many jurisdictions require authorized government inspectors or officers to conduct site visits and certify compliance, creating legal/authorization barriers to non-human substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The task requires physical presence and human judgment about regulatory compliance, making autonomous deployment infeasible. The cost of human compliance officers remains lower than any current AI-based alternative that would still require human validation and on-site presence. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | AI cannot substitute for the physical visit itself, so there is no cost comparison where AI displaces the core task; any AI use would be supplementary, not cost-saving on the visit. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While computer vision could theoretically identify displayed licenses from photos, no deployed system reliably performs end-to-end establishment visits and compliance verification at scale. Products exist only for narrow, structured components (document OCR) but not the full task. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs physical site visits and in-person compliance verification; this remains a human field-inspection task. |
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.