Tracking the structural transformation of global labor markets driven by AI, automation, and economic shifts. Data aggregated from IMF, WEF, Anthropic, OpenAI, BLS, WARN Act filings, and 80+ additional sources.
40%
Global jobs exposed iIMF Staff Discussion Note, Jan 2024 View source
92M
Roles displaced by 2030 iWEF Future of Jobs Report 2025 View source
170M
New roles created iWEF Future of Jobs Report 2025 View source
$4.4T
Annual productivity gain iMcKinsey Global Institute, 2023-2025 View source
Which jobs are growing, which are shrinking, and what's brand new
The job market isn't just losing roles — it's reshaping them. Some jobs are disappearing fast, but others are growing even faster. Here's where the movement is happening.
▲ Hiring More People
+ Farmworkers +6%
BLS projects steady growth through 2034. Physical, outdoor work remains AI-resistant. Avg salary: $32K–$38K.
+ Delivery Drivers Growing
E-commerce surge drives demand. Autonomous delivery still years from scale. Avg salary: $38K–$48K.
Source: BLS Occupational Outlook, Glassdoor, LinkedIn Salary Data 2025
Source: WEF Future of Jobs Report 2025 • BLS Employment Projections 2024-2034
05
Industry Impact
How different industries are being reshaped
Every industry will eventually benefit from AI — but getting there hurts first. Companies typically see a dip in productivity before the gains kick in. Some industries recover faster than others.
Technology & ICT
15–20%Recovers in 1–2 years
AI is writing code and running research faster than humans can
245,900+ layoffs since 2024
Microsoft ~19.5KDell ~25KIntel ~24KGoogle ~4KAmazon ~45K
AI Attribution: ~20-25% AI-Direct • ~25% AI-Adjacent • ~50% Business
Entry-level programmers and QA testers face the most disruption, while senior architects and AI specialists are in higher demand than ever. 783 tech companies announced cuts in 2025 alone, up from 551 in 2024.
AI Attribution: ~30% AI-Direct • ~35% AI-Adjacent • ~35% Business
Citigroup cited "AI-enabled systems for middle-office and operational functions." JPMorgan targeting 40-50% AI productivity gains in operations. Bank tellers, loan processors, and claims clerks most exposed. Fintech engineering and risk management roles booming.
Source: Challenger Report, SEC filings, Bloomberg
Professional Services
10–14%Recovers in 2–4 years
Legal research, consulting reports, and document drafting are being automated
25,000+ BPO/services cuts in 2025
Accenture 11KTCS 12KTeleperformanceConcentrix
AI Attribution: ~40% AI-Direct • ~35% AI-Adjacent • ~25% Business
Accenture cut employees "deemed unable to be retrained to work with AI agents." TCS cited "need to adapt to AI and automation trends." Teleperformance stock fell 60%+ on AI displacement fears. Paralegals, junior consultants, and research analysts face the biggest shift.
Source: Company press releases, Reuters, Bloomberg
Manufacturing
8–12%Recovers in 4–5 years
Robots and AI predict when machines will break before they do
20M jobs at risk by 2030
UPS ~60KAuto sector ~10K
AI Attribution: ~25% AI-Adjacent • ~60% Business/Automation • ~15% Mixed
Longest recovery cycle — physical retooling takes years, not months. But chronic shortage of technicians who can install and maintain automated systems. Industrial maintenance and robotics operations are high-demand trades. 439,000 construction workers needed in 2025.
Source: Oxford Economics, BLS, McKinsey Global Institute
Healthcare
5–8%Recovers in 5+ years
AI helps doctors diagnose faster, but patients still need human care
Safest major industry
AI Attribution: ~5% AI-Direct • ~15% AI-Adjacent • ~80% Protected
AI handles admin and diagnostics, but nursing (+6% growth), therapy, and patient care are expanding. Medical records and billing clerks face the most disruption. Aging population drives sustained demand — nursing professionals among WEF's fastest-growing roles globally.
Source: BLS Occupational Outlook, WEF Future of Jobs 2025
Tap any card for more detail • Source: MIT Sloan, Penn Wharton, WEF
06
Who's Most Affected
Not everyone is affected equally
Job displacement doesn't hit everyone the same way. Young workers, women, people without college degrees, and middle-income office workers are facing the sharpest impact. Here's who's most at risk and why.
2x
Women's automation exposure vs men
-58%
Entry-level tech hiring (early 2025)
4x
Grad degree holders in exposed roles
47%
Higher earnings in most-exposed quartile
Young Workers (22–25)
-16%
Fewer entry-level jobs available compared to experienced workers in the same fields.
January 2026 saw the lowest January hiring on record; YTD announcements fell 56% YoY
40% of young graduates actively pursuing trade careers — viewed as more sustainable in AI economy
AI replicates entry-level technical skills; companies hiring fewer juniors and expecting existing staff to use AI instead
Where displacement is hitting hardest around the world
Wealthier countries face more AI exposure because they have more office and knowledge workers. But they're also better prepared to adapt. Poorer countries face less direct exposure but lack the infrastructure to benefit from AI's upsides.
RegionJobs ExposedHow ReadyLevel
USA / Advanced Economies60%
High
Connectivity: Universal • Compute: World leader • Context: English-first LLMs • Competency: High but uneven
Priority: Workforce reskilling and regulatory oversight. WARN Act filings surging — 3,250+ in 2025 affecting 178K+ workers. Strong safety nets but widening inequality between AI-adopters and displaced workers.
Europe (EU)50–55%
High
Connectivity: Strong • Compute: Growing • Context: Multi-language challenge • Competency: Finland, Ireland, Denmark lead
EU AI Act (Aug 2026) — most comprehensive regulatory framework globally. Classifies workplace AI as "high-risk." Heavy investment in lifelong learning and agile education systems.
China / East Asia40–50%
High
Connectivity: Strong • Compute: Chip constraints but growing • Context: Localized LLMs • Competency: High
Infrastructure deployment and localized LLMs are strategic priorities. Singapore and South Korea investing heavily. Japan/Korea aging demographics amplify automation pressure.
Southeast Asia (ASEAN)30–40%
Mixed
Connectivity: Uneven • Compute: Limited • Context: Language gaps • Competency: Mixed
Hosts over half of global AI users. Projected 18% GDP uplift by 2030 and digital economy doubling to $2T. Vietnam first to pursue formal AI legislation (March 2026). ASEAN AI Governance Guide endorsed.
South Asia25–35%
Low-Mixed
Connectivity: Gaps • Compute: Emerging • Context: Language diversity • Competency: Low
Digital gender divide: women 40% less likely to own a smartphone. BPO sector heavily exposed — TCS and Infosys cuts signal AI displacement in IT services. Foundational digital literacy is the bottleneck.
Latin America / Africa20–30%
Low
Connectivity: Major gaps • Compute: Minimal • Context: Underserved languages • Competency: Low
Brazil and Mexico: high demand but constrained supply — need STEM education and skilled immigration. Lowest direct AI exposure but also least positioned to capture AI's economic benefits. Foundational connectivity and literacy are prerequisites.
US WARN Act Filings (our data)
We track official layoff filings from 34 US states. Since January 2025, there have been 3,250+ filings affecting 178,000+ workers. See the full tracker above.
Source: IMF, UNDP, World Bank, Stanford warn-scraper (34 states)
08
The Productivity Paradox
Why things get worse before they get better
When companies adopt AI, productivity actually drops at first. Workers need training, systems need updating, and old processes break before new ones are ready. But after the rough patch, growth accelerates. Economists call this the "J-curve" — it dips before it rises.
-1.33%
Initial productivity dip
+7%
Long-term GDP boost
+3.7%
Permanent GDP increase by 2075
Sector Recovery Timeline
SectorDipRecoveryPhase
Technology6–12 mo1–2 yrRecovering
Finance12–18 mo2–3 yrIn Dip
Services12–24 mo2–4 yrIn Dip
Manufacturing2–3 yr4–5 yrEarly Dip
Healthcare3–5 yr5+ yrPre-Dip
"AI-Washing" — Are companies blaming AI for normal layoffs?
tap to expand
The progression tells the story: AI was cited in just 0.6% of US job cuts in 2024, rising to 4.5% in 2025, and ~8% in early 2026. But "market and economic conditions" still drove 4x more cuts (245,000) than AI (54,836) in 2025.
Klarna reversal: Cut 700 agents for AI chatbot (2023-24), then CEO admitted "we went too far" — AI couldn't handle complex interactions. Now rehiring humans in a flexible model.
Block example: Stock rose 22% after AI-attributed layoffs. Bloomberg/Oxford Economics flagged the company as "bloated for so long" — suspected AI-washing for investor optics.
Yale Budget Lab (Feb 2026): Rate of occupational change has not increased enough to signal massive AI displacement. Unemployment duration for AI-exposed jobs remained unchanged. No macro-level evidence of AI labor disruption — yet.
Long-Term Outlook — What economists actually project
tap to expand
Penn Wharton Budget Model: AI creates a permanent +3.7% increase in GDP by 2075. Annual productivity growth peaks in early 2030s, then fades to a permanent +0.04pp boost.
McKinsey Global Institute: 400–800 million workers worldwide will need to switch occupations by 2030. The concept of "superagency" — humans and AI achieving what neither could alone — requires new career pathways and decision rights.
IMF (60/40 split): 60% of advanced-economy jobs are exposed to AI, but roughly half of those will be augmented rather than replaced. Net effect depends on policy: reskilling investment, safety nets, and transition support determine whether AI creates or destroys.
Source: Penn Wharton Budget Model 2025, McKinsey "Superagency" Report, IMF World Economic Outlook
Source: MIT Sloan, Penn Wharton, Challenger, Yale Budget Lab
MOCKUP — Section 09
09
Action Center
What you can do right now
You don't have to figure this out alone. These are real programs, real certifications, and real resources — many of them free. Pick the path that fits your situation.
Reskill Now
Only 4.1% of 1.4B workers needing reskilling have completed AI training
DOL AI Literacy Framework, experiential learning, Google/IBM certificates. See our IT Certifications Guide for career-boosting paths.
How we built this • 80+ sources • every number is traceable
Every data point on this page comes from a verifiable source. We don't guess, estimate, or make things up. Risk scores are composites from multiple independent studies. Layoff events are verified against original press reporting. WARN filings come directly from state labor departments.
How Risk Scores Work
We compare what multiple researchers say about each job, then show the range. Tap to learn more.
Risk scores are ranges from multiple independent sources (Anthropic, OpenAI, Goldman Sachs, WEF, EDsmart, Frey/Osborne, BLS). When sources disagree, we show the range (e.g., "67-95%") rather than picking one number.
Task-level exposure comes from the OpenAI "GPTs are GPTs" study which scored 19,265 tasks across 1,016 occupations. We cross-reference this with O*NET's task database for each role.
AI attribution on layoffs uses a 4-level system: AI-Direct (company said so), AI-Adjacent (restructuring in AI context), Business Cycle (traditional reasons), Mixed (both factors). Attributed by LLM classification, flagged as "unreviewed" until human-verified.
WARN filings are legally mandated government documents pulled from 34 state labor departments via the Stanford warn-scraper. These are not estimates — they're legal records.
