Legal Secretaries and Administrative Assistants
43-6012.00Median wage $55,570/yr156,280 employed (US)Rank #33 of 923 scored · top 4% by substitution
Perform secretarial duties using legal terminology, procedures, and documents. Prepare legal papers and correspondence, such as summonses, complaints, motions, and subpoenas. May also assist with legal research.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution63
Exposure65
Augmentation77
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
14 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
36%
Run 1.0.0-draft.1 · computed 2026-08-05 · rater panel: claude-sonnet-5, claude-haiku-4-5-20251001 · intervals span rater disagreement.
Why this score
The five weighted dimensions of the composite, averaged across this occupation's tasks (importance-weighted, panel mean). Exact weights and formulas: /api/v1/methodology.
Task automatabilityw 35%66
panel mean rating 3.7/5 → substitution pressure 66/100
Technical feasibility todayw 20%62
panel mean rating 3.5/5 → substitution pressure 62/100
Cost vs. human wagew 15%71
panel mean rating 3.8/5 → substitution pressure 71/100
Adoption barriersw 20%inverted — strong barriers lower the score58
panel mean rating 2.7/5 (barrier strength) → substitution pressure 58/100
Sector adoption velocityw 10%54
panel mean rating 3.2/5 → substitution pressure 54/100
Task breakdown (14 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.
Make photocopies of correspondence, documents, and other printed matter.
97CI 95–100 · exposure 100 · augmentation 25 · importance 4.2/5 · click for rater detail
Make photocopies of correspondence, documents, and other printed matter.
97| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 5/5 | Document automation and print management solutions are ubiquitously deployed in legal, financial, and professional services sectors; most mid-to-large law firms have eliminated manual photocopying tasks over the past two decades. |
| Sector adoption velocity | claude-sonnet-5 | 5/5 | Copying/scanning automation has been fully adopted across all office environments, including legal, for a very long time. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI and automation offer minimal productivity assistance here since the task itself is entirely routine mechanical copying; once automated, human augmentation becomes largely irrelevant. |
| Augmentation potential | claude-sonnet-5 | 2/5 | While copiers assist the human, this is a rote mechanical function rather than a cognitive task where AI meaningfully augments judgment or productivity beyond basic automation. |
| Task automatability | claude-haiku-4-5-20251001 | 5/5 | Modern multifunction printers and document management systems can automatically scan, sort, and copy documents at scale with minimal human intervention, easily exceeding 50% time savings compared to manual photocopying. |
| Task automatability | claude-sonnet-5 | 5/5 | Making photocopies is a simple mechanical task fully performed by copier/scanner hardware today with no AI reasoning required; it already meets or exceeds the time-saving bar via automated office equipment. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | While there are minor organizational friction points (equipment procurement, training, print management policies) and occasional client preferences for human handling of sensitive documents, no legal or licensing requirement mandates human-performed photocopying. |
| Adoption barriers | claude-sonnet-5 | 1/5 | No licensing, liability, or regulatory barriers exist for using copiers; this has been standard automated practice for decades. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | The per-page cost of automated copying via multifunction devices or cloud-based document services is a fraction of a cent, orders of magnitude cheaper than paying a human at typical administrative wages to perform this task. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Machine copying costs pennies per page versus the loaded cost of a human secretary manually operating a machine, making automation dramatically cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 5/5 | Automated photocopying, scanning, and document management solutions are mature, deployed products in law offices and corporate settings worldwide, reliably handling high volumes of document reproduction daily. |
| Technical feasibility today | claude-sonnet-5 | 5/5 | Photocopiers, scanners, and networked MFDs reliably perform this task at scale in virtually every office today, requiring no AI innovation. |
Make travel arrangements for attorneys.
87CI 76–97 · exposure 87 · augmentation 75 · importance 3.8/5 · click for rater detail
Make travel arrangements for attorneys.
87| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Professional services and larger law firms are piloting and adopting travel automation through corporate platforms, but adoption remains uneven. Small firms and solo practitioners lag; full-scale agent-driven displacement is not yet widespread in production, placing adoption in the 'middling' range with pilots common but not universal. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Legal and professional services firms have widely adopted digital travel booking and expense platforms, though some smaller firms still rely on manual arrangement by assistants. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI assistants significantly boost human productivity: auto-suggesting flights based on calendar, flagging price drops, pre-filling forms, and integrating preferences into templates. The secretary remains in control of final decisions and customization, creating a clear augmentation scenario that multiplies their task throughput. |
| Augmentation potential | claude-sonnet-5 | 4/5 | Even where a human secretary remains involved, AI tools significantly speed up itinerary building, price comparison, and calendar syncing, improving productivity substantially. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Current AI systems can handle most of this task end-to-end: parsing attorney preferences, searching flight/hotel options, booking accommodations, and managing itineraries through APIs. Manual oversight of final confirmations remains needed, but time savings of 50%+ are achievable with established tools like travel booking APIs and AI agents. |
| Task automatability | claude-sonnet-5 | 5/5 | Booking travel is a well-structured, rules-based task that AI travel assistants and integrated corporate booking tools already handle end-to-end with minimal human input, easily clearing the 50% time-saving bar. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Travel arrangements carry minimal regulatory or licensing barriers; no law requires a human to book travel. Some law firms may prefer human touch for complex/high-stakes trips, and attorney preference for human interaction remains a soft barrier, but these are organizational frictions rather than hard legal constraints. |
| Adoption barriers | claude-sonnet-5 | 1/5 | No licensing, liability, or regulatory requirement mandates human involvement in booking travel; it's a purely administrative function. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI travel booking (inference + API calls + light integration) costs pennies to single-digit dollars per arrangement, whereas a legal secretary's loaded hourly cost ($40–$70/hr) yields $10–$35 per booking at typical time investment. AI is at least 10× cheaper, all-in. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Automated booking tools cost a fraction of a cent to a few dollars per transaction versus a paralegal or secretary's loaded hourly wage for the same task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Multiple deployed products (Expedia API integrations, corporate travel platforms with AI features, calendar-sync booking assistants) reliably perform routine travel arrangements in production environments. Edge cases (complex international travel, special accommodations) may require human intervention, but basic-to-moderate complexity bookings work reliably at scale. |
| Technical feasibility today | claude-sonnet-5 | 5/5 | Corporate travel management platforms and AI-powered booking agents (e.g., TripActions/Navan, Concur with AI assistants) are deployed at scale across law firms and other enterprises today. |
Draft and type office memos.
86CI 79–92 · exposure 87 · augmentation 100 · importance 3.4/5 · click for rater detail
Draft and type office memos.
86| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Legal and professional services sectors are among the fastest adopters of AI assistants for document drafting; memo generation is a high-volume, routine task now widely automated in corporate legal departments, law firms, and administrative functions. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Legal and professional services are among the faster-adopting sectors for AI drafting tools, with many firms already integrating AI writing assistants into document workflows. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI transforms productivity by allowing legal secretaries and administrative staff to draft memos in seconds rather than minutes, with human review and customization. Humans remain in control while AI dramatically reduces composition and typing time. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI drastically speeds up drafting and formatting of memos while the secretary/assistant remains in the loop to verify accuracy, personalize content, and finalize distribution. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Current AI systems can draft office memos end-to-end with minimal human intervention, and modern language models reliably generate clear, formatted memos that meet professional standards. Typing and formatting are fully automated, easily achieving >50% time savings while maintaining equal or better quality than human-drafted memos. |
| Task automatability | claude-sonnet-5 | 5/5 | Drafting and typing office memos is a well-structured text generation task that current LLMs handle at or above human quality with major time savings, requiring only brief human review of tone and facts. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | While some organizations prefer human review and sign-off (adding light oversight friction), there is no legal requirement that a human must draft or authorize office memos. Adoption is primarily limited by organizational inertia and preference for human review, not regulatory or liability barriers. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement governs drafting internal memos, though firms may have confidentiality/data-handling policies and preferences for attorney-reviewed final content that create some friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI inference cost for memo generation is typically $0.01–$0.10 per task, versus a legal secretary's loaded wage cost of $30–$60 per hour for a 15–30 minute memo task, making AI at least 100–1000x cheaper per task. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Generating a draft memo via an LLM costs fractions of a cent versus billable secretarial or paralegal time, an order-of-magnitude cost advantage even after factoring in review. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 5/5 | Multiple deployed products (GPT-4, Claude, Copilot, specialized legal tech platforms) demonstrably generate professional memos in production environments. These systems are widely used today in law firms, corporate legal departments, and administrative settings with reliable output. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Mature products (Word/Google Docs AI, Copilot, ChatGPT) are already deployed at scale in legal and office settings for drafting routine correspondence and memos, though firm-specific formatting and tone still need light editing. |
Schedule and make appointments.
82CI 76–87 · exposure 83 · augmentation 88 · importance 4.0/5 · click for rater detail
Schedule and make appointments.
82| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Legal firms and professional services have already adopted calendar AI and automated scheduling at significant scale. Surveys show 40–60% of knowledge-work organizations now use some form of AI-assisted or fully automated scheduling, with adoption continuing to accelerate in digitized sectors. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Legal services are a professional services sector with growing AI tool adoption, but legal secretarial functions specifically lag behind adoption in more digitized front-office functions like scheduling in tech or finance. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI scheduling assistants powerfully augment human secretaries by eliminating back-and-forth confirmation emails, auto-detecting conflicts, and suggesting optimal slots, freeing them to focus on relationship management and complex administrative tasks. This is one of the most productive AI-human partnerships in administrative work. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI scheduling assistants already substantially augment legal secretaries by handling routine calendar coordination, freeing time for higher-judgment tasks like conflict checks and prioritization. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Scheduling and appointment-making is highly structured and rule-based. Current AI systems can autonomously handle calendar management, find available slots, send confirmations, and handle simple rescheduling with 60–70% time savings. Human oversight remains needed for complex conflicts or nuanced availability negotiations, preventing a full 5 rating. |
| Task automatability | claude-sonnet-5 | 5/5 | Scheduling and appointment-making is a well-structured, rule-based task involving calendar checking, availability matching, and confirmation messaging, which off-the-shelf AI scheduling tools handle end-to-end with substantial time savings. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few regulatory or liability barriers exist for appointment scheduling itself; it is not a licensed task. Minor friction arises from preference for human confirmation with clients and organizational reluctance to fully remove the human from initial contact, but these are soft, not hard barriers. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement exists for scheduling tasks, though law firms may have some institutional preference for personal touch with clients and conflict-of-interest checks that add minor friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | API-based scheduling AI costs pennies per appointment after minimal integration setup, while a legal secretary's loaded cost (salary + benefits + overhead) for the same task runs $25–50+ per appointment. AI is easily an order of magnitude cheaper. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Automated scheduling tools cost a small fraction of a human assistant's loaded wage for the same volume of appointment coordination, especially at scale. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Mature products (calendar integrations, scheduling assistants, meeting coordination platforms) reliably perform appointment scheduling in production at scale. Examples include AI-powered calendar features in Outlook/Gmail and dedicated scheduling tools, though they occasionally struggle with edge cases or multi-party complex constraints. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Mature scheduling assistants (e.g., calendar AI agents, virtual assistants integrated with Outlook/Google Calendar) are deployed in production across many organizations, though legal-specific scheduling with conflicts, court dates, and client sensitivities still sometimes requires human oversight. |
Prepare and distribute invoices to bill clients or pay account expenses.
71CI 67–75 · exposure 75 · augmentation 75 · importance 4.3/5 · click for rater detail
Prepare and distribute invoices to bill clients or pay account expenses.
71| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Legal services and professional services firms are actively adopting billing automation and RPA tools; this is a common productivity initiative in digitizing back-office operations. Survey data shows steady adoption of invoice automation in mid-to-large firms, though smaller practices lag. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Legal services show moderate digitization with growing adoption of practice management software, but many smaller firms still rely on manual or semi-manual billing processes. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI assists by automating data population, formatting, and routing, which amplifies a secretary's productivity when reviewing and approving invoices. The human remains in control of accuracy and client billing decisions, so augmentation is meaningful but secondary to the automation potential. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI-enabled billing tools substantially speed up invoice drafting, expense tracking, and distribution while secretaries retain oversight for accuracy and client communication. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Invoice preparation and distribution are highly structured, rule-based tasks involving data entry, template application, and routing. Current AI systems can reliably extract billing data, populate invoice templates, calculate charges, and send documents via email or systems—achieving significant time savings (>50%) with minimal human oversight once configured. |
| Task automatability | claude-sonnet-5 | 4/5 | Invoice preparation and distribution from templates and billing data is highly structured and already handled end-to-end by billing software with AI-assisted data entry and routing, saving significant time versus manual preparation. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | While invoice generation itself has no legal licensing requirement, law firms often require human review for accuracy and client relationship management concerns, and some regulatory contexts (e.g., trust account billing compliance) impose oversight mandates. This creates moderate friction but does not block automation. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement for invoice preparation itself, though firms may want human review for billing accuracy, client sensitivity, and trust account compliance in legal billing. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI invoice generation costs (including API calls, RPA licensing, and oversight) are substantially lower than the loaded hourly rate of a legal secretary performing manual invoice entry, formatting, and distribution. The cost advantage is typically in the 3–10x range depending on volume and system maturity. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated billing software costs a small fraction of a secretary's hourly wage per invoice cycle, though some human oversight for accuracy and client-specific billing rules remains necessary. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed accounting software, RPA platforms, and AI-powered invoice generation tools already perform this task in production across many law firms and professional services. Products like bill-generation modules in legal practice management systems and accounting automation tools demonstrate reliable performance, though occasional oversight for complex matters is still standard practice. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Legal practice management and billing systems (e.g., Clio, PCLaw) with automated invoice generation and distribution are widely deployed in production at law firms today. |
Review legal publications and perform database searches to identify laws and court decisions relevant to pending cases.
68CI 61–75 · exposure 67 · augmentation 100 · importance 3.3/5 · click for rater detail
Review legal publications and perform database searches to identify laws and court decisions relevant to pending cases.
68| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Major law firms and corporate legal departments have rapidly adopted AI-assisted legal research tools (ROSS, Westlaw, LexisNexis AI modules) since 2020–2023. Adoption is fastest in information-intensive, high-margin practice areas and well-resourced firms. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Legal services is a professional/knowledge sector rapidly adopting AI research tools, with major legal database providers embedding generative AI features into mainstream products already. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI legal research assistants substantially boost paralegal and secretary productivity by instantly surfacing relevant cases, statutes, and secondary sources, allowing humans to focus on strategic relevance assessment and synthesis. This transforms research velocity while keeping human judgment central. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI substantially speeds up and broadens legal research by surfacing relevant cases and statutes, letting the human focus on verification and application to the specific case. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can perform keyword searches, citation tracking, and basic legal document retrieval via systems like LexisNexis integration or legal research APIs, cutting research time substantially. However, judgment about *relevance* to a specific case's nuances and the assessment of which findings truly matter requires human legal expertise, limiting end-to-end automation to roughly half the task. |
| Task automatability | claude-sonnet-5 | 4/5 | AI legal research tools can search case law and statutes, summarize relevant holdings, and identify pertinent precedents far faster than manual review, meeting the time-saving threshold for most of this task's core work. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | No hard legal requirement prevents AI use for legal research; however, attorney accountability for case strategy and opposing counsel's discovery of methodology create organizational friction. Law firms exercise caution in delegating research decisions to reduce liability exposure and malpractice risk. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement restricts who can search databases, though firms often want attorney or trained staff oversight to verify accuracy of case citations given liability for erroneous filings. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-powered legal research (subscription + inference cost) is significantly cheaper than hours of human paralegal/secretary time spent on manual searches and publication review. A single subscription covers many queries; human labor scales linearly. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Subscription-based AI legal research tools cost a fraction of a paralegal's or legal secretary's hourly billing rate for equivalent search coverage, though not quite an order of magnitude when factoring in review time. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Legal research products (ROSS Intelligence, LexisNexis+, Westlaw with AI) demonstrably perform database searches and publication reviews in production law firms. These systems reliably surface relevant cases and statutes at scale, though final relevance judgment remains human-supervised. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Deployed products like Westlaw Edge, Lexis+ AI, and CoCounsel are used in production at law firms today for exactly this kind of legal research, though results still require verification due to hallucination risk. |
Receive and place telephone calls.
64CI 51–76 · exposure 62 · augmentation 75 · importance 4.1/5 · click for rater detail
Receive and place telephone calls.
64| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Professional services (law, healthcare, finance) are in the pilot and early production phase for AI receptionists; adoption is growing but not yet deep, with many mid-size law firms still relying on humans or simple IVR systems rather than advanced agents. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Legal services is a traditionally conservative, relationship-driven sector with slower AI adoption for client-facing communication compared to tech or finance sectors. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI assistants augment secretaries by auto-logging calls, suggesting quick replies, flagging priority messages, and handling routine routing, freeing humans to focus on complex communications and client relationship tasks. Productivity uplift is substantial while the human remains in the loop. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI call transcription, scheduling assistants, and triage tools can meaningfully help secretaries manage call volume and follow-up tasks even while humans remain the primary point of contact. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Current AI phone systems (e.g., voice agents, call routing automation) can handle call placement and reception end-to-end, including message taking, call screening, and basic routing, delivering >50% time savings for routine calls. Complex calls requiring nuanced judgment still benefit from human involvement, but the majority of incoming/outgoing call handling is automatable. |
| Task automatability | claude-sonnet-5 | 3/5 | AI voice agents and call-routing systems can handle scheduling, basic inquiries, and message-taking, but legal contexts often require nuanced judgment, confidentiality handling, and client relationship management that limit full automation. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | No legal authorization bars AI from receiving and placing calls; clients may prefer human contact, and some law firms want staff screening sensitive calls, but these are soft preferences rather than hard regulatory or liability barriers. Call compliance logging and occasional escalation to humans create modest friction. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing requirement to answer phones, but client confidentiality, attorney-client privilege sensitivities, and firm/client preference for human contact create moderate friction against full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI voice agents cost pennies per call (inference + integration overhead amortized), while a legal secretary's fully loaded wage for call handling runs $15–25/hour; AI is 1–2 orders of magnitude cheaper per task equivalent, especially for high-call-volume practices. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | AI-powered call handling (IVR, voice bots, virtual assistants) is substantially cheaper per call than a human secretary's loaded wage, especially for routine call volume. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed products like AI receptionists (e.g., Megan, Amtrak's chatbot, various IVR systems) and voice agents reliably handle call placement and reception in production at scale, though error rates on accent/dialect diversity and complex requests remain material. Legal office call routing is straightforward enough that existing systems perform adequately. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | AI phone assistants and virtual receptionist products exist and are deployed in some legal and business settings, but reliability for complex or sensitive legal calls remains limited and often requires human backup. |
Prepare, proofread, or process legal documents, such as summonses, subpoenas, complaints, appeals, motions, or pretrial agreements.
59CI 49–70 · exposure 62 · augmentation 100 · importance 4.3/5 · click for rater detail
Prepare, proofread, or process legal documents, such as summonses, subpoenas, complaints, appeals, motions, or pretrial agreements.
59| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Legal tech adoption is rapid in large firms and corporate legal departments, with document automation platforms seeing widespread deployment over the past 3–5 years. Smaller practices and solo practitioners lag, but the overall velocity in the profession is high compared to many other sectors. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Legal services is a professional sector with growing AI tool adoption (document automation, e-discovery), but full production deployment for document preparation remains uneven across firms of varying size and sophistication. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI-assisted document drafting, automated formatting, and intelligent proofreading dramatically improve secretary and paralegal productivity. The human remains in the loop for judgment calls, and AI transforms throughput on routine document preparation while keeping quality control with an attorney. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI dramatically speeds up drafting, formatting, and proofreading of legal documents while legal secretaries/paralegals remain responsible for final accuracy and jurisdictional compliance, making this a strong augmentation use case. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Current AI can draft legal documents from templates, perform high-quality proofreading, format and organize documents, and catch many errors reliably. However, tasks requiring nuanced legal judgment about document completeness, case-specific strategy, or jurisdiction-specific filing rules still benefit from human oversight, preventing full end-to-end automation at the ≥50% time-savings threshold in most real-world scenarios. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can draft and proofread standard legal document templates (summonses, subpoenas, motions) quickly, but jurisdiction-specific formatting, accuracy verification, and filing procedures still require human review, limiting full end-to-end automation. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Attorneys retain ethical and malpractice liability for filed documents; most bar associations require attorney sign-off. This creates a legal and professional barrier to full automation without human review, though it does not prevent AI from handling 70–80% of the mechanical work under attorney supervision. |
| Adoption barriers | claude-sonnet-5 | 3/5 | While secretaries aren't licensed attorneys, legal filings require accuracy and often attorney sign-off, and court rules mandate specific procedures, creating moderate liability and procedural barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-powered document processing costs (per-document inference plus platform fees) are substantially lower than the loaded wage of a legal secretary ($50–70k/year). A few cents per document processed versus $25–35/hour labor creates a strong cost advantage of 5–10x or more for routine tasks. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | AI-assisted drafting and proofreading tools cost a fraction of a legal secretary's hourly wage for repetitive template-based document generation, though oversight costs reduce the savings somewhat. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Multiple deployed products (e.g., LexisNexis, Westlaw AI assistants, specialized legal tech platforms) demonstrably perform document drafting, review, and proofreading in production environments. Error rates on mechanical tasks like formatting and citation checking are low, though nuanced legal analysis still requires human review in most law firms. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Legal drafting assistants and document automation tools (e.g., document assembly software integrated with AI) are deployed in law firms today, but they handle narrow document types reliably while complex or unusual filings still need attorney/paralegal correction. |
Complete various forms, such as accident reports, trial and courtroom requests, and applications for clients.
53CI 39–67 · exposure 58 · augmentation 88 · importance 4.1/5 · click for rater detail
Complete various forms, such as accident reports, trial and courtroom requests, and applications for clients.
53| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Legal services have traditionally been slow AI adopters, with significant cultural and regulatory resistance to automation of document preparation. While some mid-size and large firms pilot form automation, production deployment remains limited and cautious, typical of low-digitization, high-liability sectors. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Legal services are adopting AI/document automation steadily but unevenly; larger firms use these tools while many smaller practices still rely on manual processes. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can substantially assist a legal secretary by auto-populating form fields from intake data, flagging missing information, and organizing document templates, allowing the secretary to focus on review, accuracy checks, and complex edits. This augmentation meaningfully raises productivity while preserving human control. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI-assisted form population, data extraction, and templated drafting substantially speed up this task while the secretary retains oversight and final accuracy checks. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can extract information from source documents and populate many standard forms with moderate accuracy, but most legal forms require human review for accuracy and liability reasons. Setup and validation overhead limit time savings to roughly 40–60%, depending on form complexity and data quality. |
| Task automatability | claude-sonnet-5 | 4/5 | Form completion is highly structured and template-driven, and current LLMs and document-automation tools can extract client data and populate standardized legal forms with substantial time savings, though final review is still needed. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strong barriers exist: courts and regulatory bodies often require human-signed or attorney-verified submissions; professional liability and malpractice exposure create asymmetric error costs; bar association ethics rules and state law restrict who may prepare certain legal documents. These create both legal and organizational friction. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement mandates a human complete these forms, though attorneys typically must review and sign off on court filings, creating light oversight friction rather than a hard barrier. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI inference costs are low, but integration with legal practice management systems, human review overhead, and liability guardrails add significant operational expense. The all-in cost likely remains comparable to or exceeds hiring a legal secretary for routine form work. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated document assembly and AI form-filling tools cost a fraction of a paralegal's or secretary's hourly wage for repetitive form completion tasks. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Document processing tools and form-filling systems exist in production (e.g., contract automation, intake form population), but they often have material error rates on legal forms and typically require human oversight before submission. Few firms deploy end-to-end form completion without attorney or paralegal review. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Legal practice management software (Clio, document assembly tools like HotDocs) already automates much form-filling, but accuracy on jurisdiction-specific or nuanced forms still requires human verification, limiting fully autonomous deployment. |
Assist attorneys in collecting information such as employment, medical, and other records.
52CI 36–67 · exposure 53 · augmentation 75 · importance 4.1/5 · click for rater detail
Assist attorneys in collecting information such as employment, medical, and other records.
52| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Mid-size and larger law firms are adopting e-discovery and document management AI tools, but adoption remains uneven. Many solo and small practices still rely on manual collection; production deployment is growing but not yet dominant in the sector. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Legal services are adopting AI tools for document review and drafting at a moderate pace, though administrative record-collection workflows lag more transformative legal AI use cases. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI significantly assists attorneys and paralegals by rapidly searching records, flagging relevant documents, and organizing information, freeing humans to focus on legal analysis and judgment. Systems that surface and categorize records while humans retain control substantially boost productivity. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can significantly speed up drafting request letters, tracking status, summarizing incoming records, and flagging missing items, meaningfully boosting the secretary's productivity while they remain responsible for follow-through. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Current AI systems can automate substantial portions of record collection via document retrieval, indexing, automated requests, and data extraction with 50%+ time savings. However, the need for human judgment on sensitive records (medical, employment) and legal privilege verification prevents full end-to-end automation without human oversight. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can help draft requests, track outstanding records, and organize incoming documents, but the actual collection process involves contacting third parties, following up, and handling sensitive authorizations that still require human coordination and judgment. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strong barriers exist: attorney-client privilege, HIPAA/medical record confidentiality, data protection regulations, and bar association ethics rules require human oversight and responsibility. Attorneys must certify collection methods and handle sensitive materials, restricting pure automation and creating legal liability exposure. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement blocks AI assistance, but confidentiality, HIPAA-related medical record handling, and chain-of-custody concerns create moderate friction against full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-driven document retrieval and extraction is substantially cheaper than paralegal time for the same volume of records collected. Once systems are trained on a firm's processes, the marginal cost per record is low, likely 5–10× cheaper than human labor for routine collection tasks. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | While AI can cheaply draft and organize records, the human effort of phone calls, follow-ups, and verifying receipt of third-party records remains largely manual, keeping overall costs comparable to human labor. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Products exist for document retrieval, OCR, and automated data extraction (e.g., legal tech platforms, RPA tools), but they typically require significant setup, validation, and human review due to accuracy and liability concerns. Deployed solutions handle routine record gathering but struggle with edge cases and complex legal context. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some legal practice management and document automation tools assist with record requests and tracking, but no mature product fully manages the end-to-end collection process across employers, hospitals, and other custodians reliably. |
Organize and maintain law libraries, documents, and case files.
49CI 43–55 · exposure 50 · augmentation 75 · importance 4.3/5 · click for rater detail
Organize and maintain law libraries, documents, and case files.
49| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | While large law firms have adopted document management systems, meaningful adoption of AI-driven automation for library and case-file organization remains limited. Most firms rely on legacy systems, manual processes, or human paralegals; automation adoption is slower than in less-regulated sectors due to liability concerns and the heterogeneity of case structures. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Legal services is a professional sector with growing but uneven AI adoption; document management automation is common in larger firms but smaller practices lag, giving middling overall velocity. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI-assisted search, automated tagging, and intelligent summarization of documents can substantially raise legal secretary productivity by reducing time spent on manual filing and retrieval. Current tools like AI-enhanced legal research platforms and document intelligence systems demonstrably assist humans while keeping critical validation in their hands. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-powered document management and search tools significantly speed up filing, tagging, and retrieval tasks, meaningfully boosting productivity while legal secretaries retain oversight of organization schemes. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can automate significant portions of this task—document categorization, file indexing, metadata extraction, and basic organizational schema—but meaningful human judgment is still required for case-sensitive document classification, privileged material handling, and maintaining the nuanced organizational logic that lawyers depend on. A hybrid approach could achieve ~50% time savings, but full end-to-end automation would require constant human correction. |
| Task automatability | claude-sonnet-5 | 3/5 | AI/document management systems can automate indexing, tagging, and retrieval of digital documents, but organizing physical libraries and maintaining consistent filing conventions still requires human setup and judgment for edge cases. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Regulatory requirements around attorney-client privilege, work-product doctrine, and data security create meaningful barriers to full automation. Law firms face liability risks if automated systems misflag privileged materials or lose case files, and many jurisdictions require human responsibility for maintaining case integrity. Client expectations for confidentiality also slow adoption of third-party AI tools. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement mandates a human perform filing/organization, though law firms often prefer human oversight for accuracy and confidentiality reasons, creating moderate but not hard barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 3/5 | The all-in cost of AI-assisted document management systems (licensing, integration, data cleaning, human oversight) is roughly comparable to the loaded cost of a legal secretary performing these tasks, especially when accounting for the human expertise required to set up and maintain the system. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | Software licensing and integration costs for document management systems are moderate, and while cheaper than dedicated staff time for filing, ongoing maintenance and correction by paralegals/secretaries keeps costs roughly comparable rather than dramatically cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Document management systems with basic AI-assisted categorization and search (e.g., contract lifecycle management tools with ML-based tagging) are deployed in law firms, but they typically require significant setup and don't reliably handle complex case-file organization without human oversight. Error rates on sensitive documents and edge cases remain material, limiting production-ready deployment at scale. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Legal document management platforms with AI-assisted categorization and search (e.g., iManage, NetDocuments) are deployed in production, but full end-to-end organization of mixed physical/digital case files still requires human oversight. |
Submit articles and information from searches to attorneys for review and approval for use.
48CI 37–59 · exposure 45 · augmentation 88 · importance 3.9/5 · click for rater detail
Submit articles and information from searches to attorneys for review and approval for use.
48| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Legal services remain relatively conservative on automation; while document automation and e-discovery tools have traction, the adoption of AI for attorney-facing workflow tasks like material curation is still mostly pilot-stage rather than widespread production deployment in most law firms. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Legal services are adopting AI research and drafting tools at a moderate pace, with large firms deploying production tools while many smaller practices remain in pilot or non-adoption stages. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI tools for search automation, document summarization, and categorization can meaningfully accelerate how a secretary compiles and organizes materials for attorney review, raising throughput while the secretary retains gatekeeping and quality-control functions. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI dramatically speeds up gathering, filtering, and summarizing articles and case information for attorney review, making this an area of strong augmentation even where full automation is incomplete. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | Collecting and organizing articles/information for attorney review can be partially automated through web scraping, document retrieval, and summarization—reducing time spent on search and initial curation. However, determining relevance to a specific case and judging what merits attorney review still requires human judgment, so full end-to-end automation falls short of the 50% time-saving bar. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can search, summarize, and compile relevant articles/information into a submission-ready format, but the final routing, contextual judgment about relevance, and formatting for attorney review still often need human oversight, capping full end-to-end automation.The core research-compilation half is highly automatable, but the coordination and quality-control loop is not. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Attorneys have significant professional liability and ethical duties regarding case materials; legal ethics rules typically require attorney judgment on relevance and case strategy, creating a strong preference for human intermediaries (secretaries/paralegals) who understand context and carry institutional accountability in the firm. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement applies to secretarial research-compilation tasks, though law firms exercise caution due to confidentiality and malpractice liability concerns tied to inaccurate legal information reaching attorneys. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 3/5 | AI-powered search and curation tools have dropped in cost, making them roughly comparable to or slightly cheaper than a secretary's time on repetitive search tasks, but the human oversight required for quality control prevents the cost advantage from reaching an order of magnitude. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | AI-assisted legal research tools cost a fraction of a secretary's or paralegal's hourly billed time for compiling similar information, though oversight and subscription costs reduce the savings somewhat. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While information retrieval and basic summarization tools exist, no deployed product reliably performs the full task of intelligent article curation and submission to attorneys; most require significant manual filtering and human decisions on what is case-relevant, keeping this at research or narrow prototype stage rather than production. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Legal research AI tools (e.g., Westlaw AI, Casetext, Lexis+ AI) are deployed in production and can retrieve and summarize legal information, but attorneys still report errors and hallucination risks requiring careful verification before submission. |
Mail, fax, or arrange for delivery of legal correspondence to clients, witnesses, and court officials.
42CI 25–59 · exposure 45 · augmentation 63 · importance 4.3/5 · click for rater detail
Mail, fax, or arrange for delivery of legal correspondence to clients, witnesses, and court officials.
42| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Law firms have been slow to automate correspondence dispatch despite decades of available fax and email systems, preferring human oversight due to high error costs. Adoption remains manual and human-supervised across most sectors. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Legal services are adopting e-filing and practice management software steadily, but many courts and firms still rely on traditional mail/fax, so adoption is moderate rather than fast. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by suggesting recipients, auto-populating addresses from case files, and tracking delivery status, meaningfully improving secretary workflow without removing human control over a legally sensitive task. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-enabled practice management and e-filing tools substantially speed up drafting, addressing, and tracking correspondence, letting the secretary focus on verification and exceptions. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can draft correspondence and identify recipients, the task of physically mailing, faxing, or arranging delivery with correct addresses and formats requires human oversight due to legal stakes. Current systems lack reliable end-to-end execution of the multi-step coordination needed to ensure proper delivery to courts and witnesses. |
| Task automatability | claude-sonnet-5 | 4/5 | The substantive work here—drafting notices, addressing, choosing delivery method, tracking—is largely administrative and can be handled by software (e-filing systems, email, scheduling tools) with minimal human intervention, though physical mailing/faxing still needs some human action. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Legal correspondence delivery is subject to court rules, filing requirements, and proof-of-service obligations that often require human verification or licensed professional sign-off. Liability for misdirected legal documents creates strong organizational and regulatory friction against full automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Court rules often mandate specific service methods and proof-of-service documentation, creating procedural friction, though this is more compliance-driven than requiring licensed human judgment. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Integration costs for reliable mail/fax automation systems, plus necessary human oversight to verify correct recipient addresses and legal compliance, approach or exceed the cost of a secretary performing the task directly. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | Digital delivery via automated systems is cheap, but physical logistics (courier, mail, fax) and the need for verification of proper service keep total cost only modestly lower than a paralegal handling it, not order-of-magnitude cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Some products can automate email dispatch and fax transmission, but reliable, production-grade automation of the full task—including address verification, format compliance, proof of delivery, and court-specific requirements—is not demonstrably deployed at scale in law firms. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | E-filing and document management systems already automate service and delivery tracking in many firms, but faxing, physical mail, and court-specific delivery rules still require manual steps and human oversight in production today. |
Attend legal meetings, such as client interviews, hearings, or depositions, and take notes.
29CI 25–34 · exposure 30 · augmentation 63 · importance 3.4/5 · click for rater detail
Attend legal meetings, such as client interviews, hearings, or depositions, and take notes.
29| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Law firms have been slow to adopt full automation in client-facing roles due to regulatory constraints, liability risk, and client expectations. While transcription tools are used, the core task of attending and taking notes remains heavily human-performed in traditional settings. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Legal services are a lower-digitization, compliance-heavy sector; while transcription tools are increasingly used, actual replacement of human attendance/note-taking remains limited and cautious. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI transcription and note-summarization tools meaningfully assist legal secretaries by capturing full audio and generating drafts, reducing manual transcription burden and allowing focus on organizing and flagging important items—but the human remains essential for judgment and legal context. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI transcription, summarization, and highlighting of key points significantly boost the productivity of the human legal secretary while they remain present and responsible for judgment calls. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can transcribe audio and generate meeting summaries, the task requires real-time presence, selective note-taking judgment (knowing what details matter legally), and interaction with sensitive client information in a live setting. Current AI cannot reliably replace the discretionary judgment and contextual awareness needed in a legal environment at scale. |
| Task automatability | claude-sonnet-5 | 2/5 | AI transcription can capture speech accurately, but the task also requires physical presence, real-time judgment about what to flag, and handling confidential in-person interactions, which current systems cannot fully replace end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Courts, client meetings, and depositions often have strict rules about recording and attendance; many jurisdictions require a human present to ensure confidentiality, proper intake, and attorney-client privilege compliance. Liability concerns around missed details in critical legal proceedings create strong institutional friction. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Depositions and hearings often require a human note-taker or court reporter for legal validity, and attorney-client privilege and confidentiality rules create strong barriers to full AI substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | A legal secretary's loaded cost is modest (perhaps $30–50/hour). Transcription services and AI processing add up, and human review of transcripts is still required, making the total cost comparable to or higher than direct human note-taking. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI transcription/summarization is cheap per hour compared to a paralegal's time, but liability, confidentiality safeguards, and human review needed for depositions offset much of the savings. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Speech-to-text and meeting transcription products exist and are deployed (e.g., court reporting software, meeting recorders), but they typically produce raw transcripts requiring human review and curation. No production system reliably generates lawyer-ready legal notes autonomously from live meetings. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Transcription and note-summarization tools are deployed and used in legal settings, but reliable attendance, contextual note-taking, and integration into legal workflows still require human oversight, especially in hearings/depositions with strict accuracy and privilege concerns. |
Related occupations — Office & Administrative Support
How to read this
A high substitution score does not mean this job disappears — it means a large share of its current tasks face replacement pressure, so the mix of tasks is likely to change. High augmentation alongside substitution typically means the occupation reorganizes around the protected tasks. Wide confidence intervals mean the rater panel disagreed: treat those scores as open questions, not verdicts.
What would change this score
New model capabilities (automatability, feasibility), falling inference costs (cost ratio), regulation and licensing shifts (barriers), and measured sector adoption (velocity) all re-enter at every index release. Each release is recomputed, versioned and kept queryable — scores are claims with a date on them, not permanent labels.