Weekinfo

Artificial intelligenceWeek 47

AI and jobs: what the ILO, IMF and Stanford studies actually say

Read in Français Español

Mains d’une employée de bureau tapant sur un clavier près de documents papier
Contents
  1. How many jobs are exposed to AI?
  2. Does exposed mean at risk?
  3. Who is most affected?
  4. What does the latest data show?
  5. What does European law say?
  6. Frequently asked questions
  7. Sources

Will artificial intelligence take your job? The main institutions studying the question, the International Labour Organization (ILO), the International Monetary Fund (IMF) and researchers at Stanford University, agree on one point: many jobs are “exposed,” but exposure does not mean disappearance. The first measurable effects mainly hit young people entering the workforce.

How many jobs are exposed to AI?

Figures vary with the method, because each study defines “exposure” differently. They should not be added together or read as layoff forecasts.

Main published estimates
Source (date)EstimateWhat it measures
ILO (May 2025)25% of global employmentOccupations with some tasks exposed to generative AI
ILO (May 2025)3.3% of global employmentHighest exposure level (gradient 4)
ILO (May 2025)34% in high-income countriesExposure, vs. 11% in low-income countries
IMF (January 2024)Almost 40% of jobsAI exposure; about 60% in advanced economies
Stanford (August 2026)19% gap for ages 22-25Employment in exposed vs. less exposed occupations

Does exposed mean at risk?

Not necessarily. The ILO index, built with the NASK institute from nearly 30,000 occupational tasks, concludes that generative AI is more likely to transform jobs than eliminate them. Some tasks can be automated, but rarely an entire occupation.

The IMF stresses both sides. In advanced economies such as the US and UK, about 60% of jobs may be affected, but some workers could gain from higher productivity. Exposure is estimated at 40% in emerging markets and 26% in low-income countries.

Who is most affected?

  • Clerical workers: the ILO finds office and administrative occupations have the highest exposure.
  • Women: in the most exposed category, the ILO counts 4.7% of female employment versus 2.4% of male employment; in high-income countries the gap widens to 9.6% versus 3.5%.
  • Young graduates: Stanford finds employment for 22- to 25-year-olds in exposed occupations is 19% below where it would be had it kept pace with less exposed peers.
  • Middle-skill roles: the IMF notes that jobs such as routine office work are being squeezed.

What does the latest data show?

The “Canaries in the Coal Mine” study by Erik Brynjolfsson, Bharat Chandar and Ruyu Chen, revised on August 12, 2026, uses payroll data from ADP covering millions of US workers through June 2026. It finds no evidence of economy-wide job displacement. But the gap for young workers has widened steadily since the first version in August 2025.

The mechanism: companies are hiring fewer entry-level workers rather than laying people off. The decline is concentrated where AI substitutes for tasks; where it complements workers, employment is flat or rising, especially for experienced staff. The authors describe these as early indicators, not causal estimates.

In January 2026, the IMF added another angle: about one in ten job postings in advanced economies requires at least one new skill, and such postings pay about 3% more in the UK and US. In regions with strong demand for AI skills, employment in AI-vulnerable occupations is 3.6% lower after five years.

What does European law say?

The EU AI Act lists AI used for hiring and managing workers, such as CV screening, as “high-risk.” Those obligations have been pushed back to December 2, 2027 (see our AI Act guide). Emotion recognition in the workplace, however, has been banned in the EU since February 2025. On AI assistants arriving at work, read our office AI agents piece.

Frequently asked questions

Will AI wipe out 40% of jobs?

No, that is not what the IMF says. The figure refers to jobs exposed to AI, some of which could instead become more productive.

Have job losses already been measured?

Stanford’s study of US data through June 2026 finds no economy-wide displacement, but fewer hires of 22- to 25-year-olds in exposed occupations.

Why are women more exposed?

According to the ILO, because they are more often in clerical jobs, which are the most exposed to generative AI.

Do these studies apply to my country?

The ILO and IMF publish global or regional averages; the Stanford study covers the United States. Trends apply broadly to advanced economies, but there is no national figure here.

Sources

Read next

Sections

Language