AI Job Risk Country Ranking 2026: All 206 Countries Compared

How the ranking is calculated

Each country's score is a weighted average of AI exposure scores across ISCO-08 major occupation groups, weighted by the share of employed workers in each group. If a country has 30% of its workforce in professional occupations (which score 6.5/10 on AI exposure) and 20% in clerical occupations (which score 8.5/10), these shares determine how much each group contributes to the country's overall score.

Employment data comes from ILO ILOSTAT (Creative Commons CC BY 4.0), reference year 2025 for most countries, using ISCO-08 major occupation group classifications. AI exposure scores per occupation group are research-based estimates informed by Frey-Osborne (Oxford, 2017), OECD Future of Work studies, and IMF GenAI research (2024). They measure the proportion of each occupation's core tasks that current AI can perform or significantly augment - not the probability of job loss.

The top 20 highest AI job risk countries

The top 20 is almost entirely composed of European countries, with Singapore, Australia, and New Zealand representing Asia-Pacific and Oceania. The pattern is consistent: these are economies where the majority of workers are in professional, technical, clerical, and management roles - the occupation groups that score highest on AI exposure.

Luxembourg's 5.87/10 is the global peak. It has one of the highest concentrations of financial services workers per capita in the world, with the banking sector (more than 150 international banks headquartered or with major operations in Luxembourg) employing a disproportionate share of the workforce in roles that are directly targeted by current AI tools: compliance analysts, fund administrators, financial reporting specialists, and legal professionals servicing the fund industry.

The bottom 20: lowest AI job risk countries

The bottom of the ranking is dominated by Sub-Saharan African and South Asian economies where the majority of workers are in agriculture (ISCO group 6, which scores 3.0-3.5/10 on AI exposure) and elementary occupations (ISCO group 9, which scores 2.0/10). Lower AI exposure scores at the country level do not mean these workers are safe from all economic disruption - they face significant robotics risk and general economic instability - but they are currently less exposed to AI specifically.

Tanzania at 2.85/10 reflects a workforce where over 60% of employed workers are in agricultural occupations according to ILO ILOSTAT 2025 data. Elementary occupations account for another 15-20%. The combined weight of these low-scoring groups pulls the national average below 3.0/10. The same pattern holds across most low-income African economies.

What drives the gap between countries

The primary driver of a country's AI job risk score is its occupation structure - how workers are distributed across ISCO-08 major groups. This is almost entirely a function of economic development. As economies develop, workers shift from agriculture and elementary occupations into manufacturing, then into services and professional roles. AI exposure increases at each step of this transition.

A secondary driver is within-sector concentration. Luxembourg scores higher than Germany (5.87 vs 5.30) not because its workers have fundamentally different roles, but because its financial sector concentration skews the professional workforce heavily toward the highest-exposure financial and legal roles. Ireland (5.75) benefits from a similar dynamic with its large pharmaceutical and technology multinational sector employing a disproportionate share of high-exposure knowledge workers.

Regional patterns in the ranking

Western Europe and the English-speaking developed world (US, UK, Canada, Australia, New Zealand) cluster in the 5.0-5.9 range. East Asian developed economies (South Korea, Japan, Singapore, Hong Kong) are in the 4.9-5.7 range. Eastern European economies cluster in the 4.2-5.0 range, reflecting their more mixed occupation structures. Latin American economies cluster in the 3.8-4.5 range. South Asian economies (India, Pakistan, Bangladesh, Nepal) are in the 3.5-4.2 range. Sub-Saharan Africa is in the 2.8-3.6 range.

The Middle East presents an interesting outlier pattern. Saudi Arabia (4.85/10) and the UAE (4.92/10) score higher than their income peers in other regions because their active labour forces are heavily weighted toward professional and clerical work, with lower-wage migrant workers often in administrative support roles that are particularly AI-exposed.