Insights
What we know about AI and jobs
The headline findings from the main research on AI and work, followed by patterns from this atlas’s own data on exposure and real-world AI use.
Exposed jobs worldwideMODEL
757.8M
21% of jobs in 183 countries
Women more exposedMODEL
170 of 183
countries
Exposed but less preparedMODEL
16
above-median exposure, below-median IMF preparedness
AI conversations about workANTHROPIC
44%
Claude.ai, April–May 2026
Research findings
Each figure is quoted from the source as published. Different studies measure different things: exposure (could AI do these tasks?), adoption (is it being used?) and outcomes (are jobs or wages changing?). Exposure is not the same as job loss.
1 in 4
One in four workers worldwide is in an occupation with some exposure to generative AI.
About 20.5% of jobs sit in moderate exposure gradients and 3.3% in the highest.
34% vs 11%
GenAI exposure covers 34% of employment in high-income countries but 11% in low-income countries.
9.6% vs 3.5%
In high-income countries, 9.6% of women's jobs are in the highest-exposure category, against 3.5% of men's.
Globally the figures are 4.7% for women and 2.4% for men. Clerical work, where women are over-represented, is the most exposed.
~40%
Almost 40% of global employment is exposed to AI; about 60% in advanced economies, 40% in emerging markets and 26% in low-income countries.
In advanced economies, roughly half of exposed jobs may benefit from AI and half may be negatively affected.
−19%
Employment of 22–25-year-olds in the most AI-exposed occupations is 19% below where it would be had it kept pace with less-exposed peers.
Experienced workers in the same occupations show no such gap; the shortfall comes mainly from reduced hiring. The authors describe these as descriptive, not causal, findings.
Stanford Digital Economy Lab, Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of AI (revised) · Aug 2026
No clear break
Across the wider US labour market there is still no discernible disruption from AI; measures of AI usage show no link to employment or unemployment.
The occupational mix is changing only about 1 percentage point faster than during the early internet era.
Yale Budget Lab & Brookings, Tracking the Impact of AI on the Labor Market · Updated Oct 2026
120,136
US employers cited AI in 120,136 announced job cuts between January and September 2026, about 21% of all cuts and the leading stated reason this year.
That compares with 54,836 in all of 2025. These are reasons employers give in announced plans, not confirmed layoffs.
Challenger, Gray & Christmas, September 2026 Challenger Report · Oct 2026
+78M
Employers expect 170 million jobs to be created and 92 million displaced by 2030, a net gain of 78 million.
Creation plus displacement equals 22% of today’s formal jobs. 86% of employers expect AI and information processing to transform their business, and 40% expect to reduce staff where AI can automate tasks.
World Economic Forum, Future of Jobs Report 2025 · Jan 2025
39%
Workers can expect 39% of their existing skill sets to be transformed or become outdated between 2025 and 2030.
World Economic Forum, Future of Jobs Report 2025 · Jan 2025
62%
Jobs requiring AI skills carry a 62% wage premium; such jobs grew 69% against 9% for all jobs.
In the US, entry-level roles most exposed to AI grew 35% since 2019 while other entry-level roles fell 10%.
PwC, 2026 Global AI Jobs Barometer · Jun 2026
3×
Industries most exposed to AI saw three times higher growth in revenue per employee (27%) than the least exposed (9%).
PwC, The Fearless Future: 2025 Global AI Jobs Barometer · Jun 2025
49%
About 49% of jobs have seen at least a quarter of their tasks performed using Claude.
This measures where AI is being used, not job losses.
Anthropic, Anthropic Economic Index report: Learning curves · Mar 2026
1 in 10
About one in ten job postings in advanced economies now requires at least one new skill, roughly twice the rate in emerging markets.
In highly exposed, low-complementarity occupations, employment is 3.6% lower after five years in regions with stronger demand for AI skills.
300M
Generative AI could expose the equivalent of 300 million full-time jobs worldwide to automation, and raise global GDP by 7% over a decade.
Goldman Sachs Research, Generative AI could raise global GDP by 7% · Apr 2023
Up to 30%
By 2030, activities accounting for up to 30% of hours worked in the US could be automated, a trend accelerated by generative AI.
McKinsey Global Institute, Generative AI and the future of work in America · Jul 2023
27%
Occupations at the highest risk of automation make up about 27% of employment across OECD countries.
This covers all automation technologies, including AI-enhanced robotics, so the most at-risk jobs are low- and middle-skilled.
OECD, OECD Employment Outlook 2023 · Jul 2023
88%
88% of organisations now report using AI, according to the 2026 AI Index. The 2025 edition put the figure at 78%, up from 55% a year earlier.
Stanford HAI, AI Index Report 2026 (and 2025) · Apr 2026
Where AI use is most intense
Regions with the highest AI usage
Each country’s usage index × the region’s share of national use relative to its population share. Regions over 250,000 people. ANTHROPIC
Most exposed regions
Share of jobs exposed to generative AI. Capital and big-city regions lead because they concentrate finance, ICT and professional work. MODEL
Occupations where AI is already doing the work
Highest observed exposure
How much of each US occupation’s task content shows up in real Claude usage (0–1). 54% of 756 occupations score zero. ANTHROPIC
How the world uses AI
Claude.ai conversations, April–May 2026. ANTHROPIC
What conversations are for
Work 44%Personal 39%Coursework 16%
How people work with AI
Augmentation (human stays in the loop) 51%Automation (AI completes the task) 49%
Most common requests
Exposed, but less prepared
These countries have above-median AI exposure but below-median IMF AI preparedness: their workers face substantial exposure with weaker digital infrastructure, skills systems or regulation to help them adapt.