Research-grade exposure data for strategists and policy institutions. Financial planning tools for CFOs and talent leaders. One platform. Two products.
The only continuously updated, occupation-level AI exposure dataset integrating multi-model scoring with BLS employment data and live employer AI adoption signals.
Transform AI exposure data into revenue impact, profit impact, and payroll-weighted dollar figures. Built for CFOs, Heads of TA, and recruiters pricing AI-era roles.
No affiliation with any AI laboratory, university, or government agency. Subscriber-funded. Published monthly on the 9th. No AI lab can independently score how exposed occupations are to its own models — that's why independence matters.
Published May 12, 2026 · Three data layers · 923 occupations
AWEI™ gives you the independent standard. AMWI™ turns it into dollars. Individual plans from $59. Enterprise data licensing available.
Multi-model AI exposure scores for all 923 U.S. O*NET occupations, integrated with BLS employment data and live employer AI adoption signals. Published monthly. No AI lab affiliation.
Independence statement. The AI Work Exposure Index is independently developed by GlobalProsResearch. No affiliation with Anthropic, OpenAI, Google, or any academic institution. No AI lab can independently assess how exposed occupations are to its own models. That's why independence matters.
Proprietary multi-model scoring of 19,000 O*NET tasks across four dimensions: performance variance, tacit knowledge intensity, data abundance, and algorithmic gap. Scores averaged across multiple leading AI models.
Bureau of Labor Statistics OES data mapped to 928 O*NET SOC codes. Proprietary nowcast adjustment updates annual BLS figures to current quarter. Transforms exposure percentages into absolute worker counts.
Monthly AI skill mention rates and salary premium data from a licensed database covering ~95% of U.S. job postings. Real employer intent — 12–18 months ahead of BLS wage survey data.
Full 923-occupation dataset, monthly PDF, dashboards, API access on Professional and Enterprise plans. Academic and policy institution pricing available.
AMWI™ adds Revenue & Profit Impact Scores, a Workforce Cost Modeler, and ~2,680 TA recommendations built on the same AWEI™ foundation. AWEI™ subscribers get 20% off.
Revenue Impact Scores, Profit Impact Scores, and payroll-weighted financial planning figures — built for CFOs, Heads of TA, and senior recruiters.
What AMWI™ adds over AWEI™ alone: Revenue Impact Score (0–100 index of AI-driven output opportunity or displacement risk), Profit Impact Score (0–100 index of AI-driven efficiency and margin potential), Workforce Cost Modeler (headcount × avg salary × AI impact % = payroll exposure), and ~2,680 TA and recruiter recommendations across all occupations.
AWEI™ subscribers receive 20% off all AMWI™ plans. Enterprise pricing available for large employers, consulting firms, and financial services data feeds.
Enter your workforce by occupation and headcount. AMWI™ calculates Revenue Impact, Profit Impact, and payroll-weighted dollar exposure — the figures your CFO needs.
What is our financial exposure to AI-driven workforce change? What does it cost us to get ahead of it versus react? AMWI™ answers in dollar terms — the only language that belongs in a board presentation.
CFOs currently have no standardized framework to quantify AI-driven workforce cost exposure. They know AI is restructuring labor costs but lack the tools to translate that knowledge into planning figures that can sit alongside capex, headcount budgets, and margin forecasts.
AMWI™'s Workforce Cost Modeler outputs payroll-at-opportunity and payroll-at-risk figures by occupation. It gives you the first credible financial planning tool built specifically for this problem.
A 0–100 index of AI-driven output opportunity or displacement risk per occupation. Positive = augmentation upside. Used to calculate payroll-weighted dollar exposure in the Workforce Cost Modeler.
A 0–100 index of AI-driven efficiency and margin potential. Incorporates AI exposure score and salary premium signal. Higher = stronger case for AI investment in that role.
Headcount × average occupation salary × AI Impact %. A directional estimate of the payroll at opportunity or at risk — the figure boards want to see.
Forward projections by occupation group and individual role. How your payroll exposure changes as AI capabilities and employer adoption accelerate. Updated monthly.
Why AMWI™ is in your board presentation, not just your HR platform. A CFO who uses AMWI™ has just made GlobalProsResearch part of the company's official financial planning record — built on an independent, multi-model, BLS-integrated research foundation that can be cited, audited, and defended.
AMWI™ tells you which roles to hire for aggressively now, which to redesign before posting, and which to stop backfilling — with occupation-level job description guidance updated monthly.
The AI Readiness classification gives you a clear decision framework: AI Disrupted occupations (≥60% exposure) require immediate role redesign or AI-augmented skill requirements. AI Augmented occupations (30–59%) need monitoring and a transition plan. AI Resilient occupations (<30%) can be hired traditionally near-term.
AMWI™'s TA-specific recommendation corpus covers all 923 occupations — how to update job descriptions, reweight hiring criteria, restructure roles, and price for AI skills in the current market.
AMWI™'s recruiter-specific recommendations — one per occupation, updated when scores change — tell you how to structure job descriptions for AI-era roles, what skills to require, and what the AI salary premium signals about your immediate market opportunity.
What employers are currently paying for AI-capable workers vs non-AI postings in the same role. May 2026 mean: +31%. Updated monthly. A leading indicator 12–18 months ahead of BLS wage data.
What percentage of job postings in each occupation mention AI skills, AI tools, or AI-adjacent responsibilities. Software developers: 34% (up 15pp YoY). Legal: 28% (fastest accelerating).
Role-specific views: CFO financial planning, TA strategy, recruiter operations, and consulting & workforce advisory.
Bubble = BLS employment size. High-right = highest disruption and highest AI wage premium.
The only occupation-level dataset combining AI exposure with revenue impact and profit impact scores. Built for CFOs, talent acquisition teams, and recruiters who need to make workforce decisions based on financial consequence and work-style traits, not just theoretical risk.
Exposure scores tell you what AI can do. AI Measured Workforce Intelligence tells you what it means financially , and what to do about it.
For each occupation: how much does AI change this role’s contribution to top-line revenue? Does AI augment output (Revenue Opportunity) or reduce the role’s revenue-generating function (High-Exposure Watch)?
For each occupation: how much does AI change the role’s contribution to margin and efficiency? Which occupations should you invest in AI training for the highest profit uplift per employee?
For each occupation: a specific talent action , Increase hiring, Continue & train, Maintain & monitor, or Restructure. Based on the combination of AI exposure, revenue impact, and profit impact.
Every occupation classified monthly as AI Disrupted, AI Augmented, or AI Resilient. For each occupation, AMWI™ identifies which of the 21 O*NET work styles predict success in AI-impacted roles , including analytical thinking, adaptability, initiative, and attention to detail , and provides recruiters and TA leaders with specific guidance on exactly how to use this in the hiring process.
456 occupations. AI improves both revenue and profit contribution. Increase or prioritise hiring and add AI training.
189 occupations. High AI exposure AND strong financial upside. Immediate action: reskill and redesign workflows.
209 occupations. High AI exposure with mixed financial signals. Monitor closely; do not expand hiring without review.
69 occupations. Low AI signal. Maintain current staffing; no urgent action required.
Model the financial impact of AI on your workforce cost structure. Which occupations drive revenue and how does AI change that? Where is the profit improvement opportunity? What is the ROI of AI training versus backfilling?
Know which roles to prioritise based on AI-adjusted financial contribution. For each occupation: which of the 21 O*NET work styles predict success, how to structure interviews around AI-adjacency, which skills to add to job descriptions, and what salary premium to budget for AI-fluent candidates.
For each role you’re filling: is this a growth hire or a replacement? What AI skills should you require? What salary premium should you budget? What should the JD say about AI? The dataset answers all of these.
Enter your occupations and headcount. AMWI™ scores each role by Revenue Impact and Profit Impact , and gives you a Recommended Action for every one. No credit card required.
Three representative occupations from the May 2026 dataset. Revenue and Profit scores, AI Impact %, and full Recommended Actions are available to subscribers across all 923 occupations.
| Occupation | AI Band | Flag | Revenue Score | Profit Score | Recommended Action |
|---|---|---|---|---|---|
| Software Developers | Disrupted | Hi-Exp AI Opportunity | High (subscribers only) | High (subscribers only) | Increase hiring. Full occupation-specific guidance for subscribers. |
| Customer Service Representatives | Disrupted | Hi-Exp Watch | Medium (subscribers only) | High (subscribers only) | Restructure role. Full occupation-specific guidance for subscribers. |
| Helpers, Carpenters | Resilient | Monitor | High (subscribers only) | Medium (subscribers only) | Maintain staffing. Full occupation-specific guidance for subscribers. |
Full dataset: Revenue Final Score, Profit Final Score, Revenue AI Impact %, Profit AI Impact %, and occupation-specific recommended actions for all 923 O*NET occupations. Updated monthly.
The question isn't whether to use data for workforce AI decisions. The question is whether your data can actually answer the financial questions your CFO and board are asking.
"What is our exposure to AI workforce disruption and what is it costing us?", AMWI™ is the only product that answers both halves with monthly data and occupation-level precision.
Priced for all company sizes. Cancel anytime. All plans include the full 923-occupation dataset.
For small businesses and SMBs making hiring decisions for a handful of roles. Know which occupations to prioritise for AI training and which roles to hire with AI fluency requirements.
For growing companies with multiple hiring managers and a TA team. Full financial impact data, AI skill demand signals, and the workforce cost modeller.
For Fortune 1000 employers, HR platforms, and talent consulting firms needing full data access, API integration, and board-level reporting.
GlobalProsResearch subscriber? Add AMWI™ to your existing plan at 20% off, use the buttons above or email research@globalprosresearch.org with your account email and we'll apply the discount at checkout.
A neutral, multi-model AI exposure index for researchers, policy makers, HR platforms, and financial services. Measures what AI can theoretically do to each occupation. No company affiliation. Published as a research product.
A financial decision tool for employers. Translates AI exposure into revenue impact and profit impact scores for each occupation, with specific talent actions for CFOs, TA leaders, and recruiters.
Bubble = occupation (size = AI exposure level). Colour = AI Risk/Opportunity flag. High-right = highest combined financial value from AI.
How 923 occupations are classified by their AI financial risk and opportunity profile.
Occupations where AI delivers the greatest margin improvement. Prioritise AI training investment here first.
Average revenue and profit AI impact score across the four flag categories. High-Exposure AI Opportunity occupations show the strongest combined uplift.
Enter your occupations and headcount below. Get weighted Revenue and Profit Final Scores, flag distribution, and Recommended Actions for your workforce composition.
Individual occupations sorted highest to lowest Revenue AI Impact %. Green = positive / Red = negative. Annotation: Revenue Final Score band.
X-axis: AWEI AI exposure %. Y-axis: AMWI Profit AI Impact %. Colour: flag category. Identifies highest-ROI occupations for AI training investment.
All issues since inaugural publication. Three data layers per release.
Subscriber access. Web releases and PDF downloads for all paid subscribers. API access on Professional and Enterprise plans. All releases permanently archived.
Task-level scores, BLS employment, AI posting signals, 3- & 5-year projections, and salary premium data for all occupations.
Role-specific views drawing from three core data layers.
BLS employed workers (thousands) stacked by exposure tier per occupation group. Highly exposed (>50%), moderate (20–50%), low (<20%). May 2026: 47M workers in high-exposure roles.
Occupation counts by AI Readiness tier (Disrupted / Augmented / Resilient) per O*NET major group. May 2026: 312 occupations classified AI Disrupted (34% of 923).
Disrupted occupation counts now vs projected 2029 per group. 84 occupations forecast to reclassify from Augmented → Disrupted within 36 months.
AI posting share % by occupation group: May 2025 vs May 2026. Comp & Math: 19%→34% (+15pp). Legal: 20%→28% (+8pp). Business & Fin: 12%→22% (+10pp).
Dual bar: median advertised salary for AI-mentioning vs non-AI postings by occupation group. May 2026: Legal $156K vs $122K (+28%). Comp & Math $148K vs $110K (+35%). Source: AI Postings database, not BLS wage survey data.
Citation. GlobalProsResearch AI Work Exposure Index, Vol. 1 No. 1, May 2026. Data: BLS OES 2025; Proprietary AI Posting Signals, May 2026. Publisher: GlobalProsResearch. GlobalProsResearch.org
Bubble size = BLS employment. Q1 (top-right) = urgent action: high exposure AND high employer signal.
Which occupation groups are accelerating fastest toward the disruption threshold.
Composite: posting share rate of change × exposure acceleration. Higher = faster-moving.
AI posting share % by occupation group, May 2025–May 2026. May 2026 endpoint: Comp & Math 34% / Legal 28% / Arts & Media 24% / Business & Fin 22% / Management 19%.
Median advertised salary: AI-mentioning vs non-AI postings (AI Postings database). Line overlay: premium %. Per-occupation compensation from Lightcast.
Enter occupations and headcount to see weighted AI exposure %, Readiness donut (Disrupted / Augmented / Resilient), and 3-yr projection for your specific workforce.
Sortable by risk tier. Fields: Occupation, AI exposure %, AI Readiness, BLS employed, Workers at risk, Exposure velocity, AI posting %, Salary premium, Risk tier.
| Occupation | Exposure % | Readiness | Workers at Risk | AI Post% | Sal. Prem. | Risk Tier |
|---|---|---|---|---|---|---|
| Software Developers | 71.3% | Disrupted | 1.31M | 34% | +37% | High |
| Data Scientists | 80.4% | Disrupted | 0.14M | 41% | +44% | High |
| Financial Analysts | 62.8% | Disrupted | 0.19M | 22% | +28% | High |
| HR Specialists | 52.1% | Augmented | 0.42M | 18% | +24% | Medium |
| Registered Nurses | 28.4% | Augmented | 0.89M | 13% | +18% | Low |
| Full 923-occupation dataset with filtering and CSV export in subscriber dashboard | ||||||
Projected classification shift: how many workers move from Augmented to Disrupted over the 3- and 5-year horizon.
AMWI™ adds Revenue Final scores, Profit Final scores, AI Impact % per occupation, and a Workforce Cost Modeler. Available as an add-on at 20% off for GlobalProsResearch subscribers.
Bubble size = BLS employment. High-right = highest disruption risk and highest AI wage premium.
Gap between what AI can theoretically do and what employers are currently requiring in job postings.
Occupations ranked highest to lowest AI salary premium. Absolute $ gap annotated.
| Sector | Mean Exposure |
Workers at Risk |
Avg AI Posting % |
Avg Salary Premium |
3-yr Proj. | AI Readiness |
Trend |
|---|
AWEI™ · BLS OES 2025 · AI Posting Signals May 2026
Total U.S. workers in occupations >50% AI exposure. Right axis = % of total workforce.
Month-over-month exposure change (pp) by occupation group. Darker = faster acceleration.
AI salary premium % trending upward across all exposed groups , a wage pressure leading indicator.
Each group’s at-risk workers as % of the 162M total workforce. ESG and portfolio risk sizing.
Bubble = workforce size. High-right = highest AI adoption speed and highest wage premium.
How many occupations cross classification thresholds over the projection horizon. Disrupted count grows as Augmented crosses threshold.
API JSON feed published same-day as each monthly release. Fields per occupation: AI exposure % (current & 6-mo prior), delta (pp), BLS employed, workers at risk, AI Readiness, AI posting %, posting 12-mo delta, AI salary premium, 3-yr & 5-yr projections. Available to Professional and Enterprise subscribers.
Sources: AWEI™ · O*NET · BLS OES · AI Postings · Derived · API access details →
Mean exposure across all 923 occupations by release month. Band = 25th,75th percentile range.
Number of occupations in each 10pp exposure bin. Right axis = cumulative % of workforce.
What AI can theoretically do (AWEI™) vs what employers are already requiring (posting signal). Identifies adoption lag.
Mean AI exposure % by ISCO-08 major occupational group with U.S. O*NET equivalent reference bar. Enables ILO and international policy analysis.
Full O*NET task-level dataset: task ID, occupation/SOC, task description, importance/frequency weighting, exposure score (multi-model average), per-model breakdown (NDA), dimension scores. API or CSV. LLM and institutional access only.
| Task ID | Occupation | Task | Importance | Exposure | Dimensions |
|---|---|---|---|---|---|
| T4.A3.a | Software Developers | Write code for applications | 5 / 5 | 84.2% | High perf. variance |
| T3.B1.c | Financial Analysts | Analyse financial statements | 4 / 5 | 78.6% | High data abundance |
| T6.C2.b | Lawyers | Advise clients on legal rights | 5 / 5 | 34.1% | High tacit knowledge |
| Full 19,000-task dataset via API or CSV, LLM & institutional access | |||||
Key headline metrics across all published releases. Full occupation × month time series available via API to licensed institutional partners.
| Release month | Mean exposure % | AI Disrupted (occ.) | AI Augmented (occ.) | AI Resilient (occ.) | Workers at risk (M) | Change vs prior |
|---|
Task-level data. Full 19,000-task scoring dataset with per-dimension scores and per-model breakdown (anonymised in public; identified for licensed partners under NDA) available via API. Contact research@globalprosresearch.org to discuss institutional data access.
Three independent data layers, integrated monthly. Each with its own sourcing, update cadence, and documentation.
The AI Work Exposure Index applies a proprietary scoring framework grounded in Moravec's Paradox — the principle that AI finds cognitively complex tasks more automatable than tasks requiring physical dexterity or embodied tacit knowledge.
Each of the 19,000 O*NET tasks is scored across four dimensions. The four scores combine using a proprietary weighting function to produce the task-level exposure value.
How consistently can AI perform this task across varied contexts? Low-variance, well-defined tasks score higher.
How much relies on embodied, experiential knowledge difficult to encode? High intensity suppresses the score.
Is sufficient training-relevant data available for AI to have learned this domain at scale?
Is there a known pathway to automating this task, or does it require capabilities current AI architectures lack?
Independence statement. The AI Exposure Index is independently developed by GlobalProsResearch. No affiliation with Anthropic, OpenAI, Google, or any academic institution. Methodology is proprietary. Scoring is neutral by design.
Employment counts from the BLS Occupational Employment and Wage Statistics (OES) program, mapped to 928 O*NET SOC occupation codes. A proprietary nowcast adjustment updates annual BLS data to the current quarter.
Attribution. Employment data sourced from U.S. Bureau of Labor Statistics, OES program. BLS data is public domain. GlobalProsResearch's BLS-to-O*NET SOC mapping methodology is proprietary.
A data license with a proprietary job posting database covering ~95% of U.S. job postings. Each month we extract the percentage of postings mentioning AI skills, AI tools, or AI-related work by O*NET occupation code.
The posting signal is a leading indicator — employer intent before actual workforce changes appear in BLS data. A lag of 12–18 months between employer skill signaling and measurable employment impact is a reasonable planning assumption.
Choose AWEI™ for research and workforce intelligence. Choose AMWI™ for financial planning and operational decisions. AWEI™ subscribers receive 20% off AMWI™.
For independent researchers, journalists, and workforce professionals needing monthly AI exposure data with employment context.
For university departments, think tanks, and research institutions. Full three-layer data with citation kit.
For BLS, DOL, state workforce agencies, and international policy bodies.
AMWI™ subscribers on an active AWEI™ plan receive 20% off. Add Revenue & Profit Impact Scores, the Workforce Cost Modeler, and ~2,680 TA recommendations.
GlobalProsResearch is an independent workforce intelligence firm publishing the AI Work Exposure Index (AWEI™) and the AI Measured Workforce Intelligence platform (AMWI™) — the only continuously updated, occupation-level AI workforce intelligence platform that integrates multi-model scoring with BLS employment data and live employer AI adoption signals.
Founded in West Palm Beach, Florida. No affiliation with any AI laboratory, university, or government agency. Subscriber-funded. Published monthly on the 9th of each month.
The inaugural release, Vol. 1, No. 1, was published May 12, 2026, covering 923 O*NET occupations across 19,000 individual tasks.
Contact: research@globalprosresearch.org · billing@globalprosresearch.org · West Palm Beach, Florida · Mon–Fri 9am–6pm ET
API access, data distribution, LLM partnership, and government procurement. All conversations handled under mutual NDA on request.
Multi-model index + BLS employment + AI posting signals. Per-model score breakdown. Full term sheet on request under NDA.
Product embed rights, full three-layer API, co-branded data products. Custom occupation segments included.
Procurement-compatible terms. Policy redistribution rights. Quarterly briefing and testimony support.
Contact. All enterprise and partnership enquiries: research@globalprosresearch.org · GlobalProsResearch · West Palm Beach, Florida · All conversations handled under mutual NDA on request.
GlobalProsResearch.orgSign in to access your subscriber dashboard
By signing in you agree to our Terms of Service and Privacy Policy
Manage your profile, plan, and billing