Afghanistan AI Job Risk 2026: Which of 7.43 Million Workers Are Most Exposed?

7.43 million workers, ILO ILOSTAT data

Employment data comes from ILO ILOSTAT (CC BY 4.0), based on CSO (Central Statistics Organisation) Afghanistan's Labour Force Survey (LFS). Classification follows ISCO-08 major group structure across Afghanistan's 7.43 million employed workers. Note: data quality is constrained by the country's ongoing governance and statistical capacity situation; figures should be understood as best available estimates.

Afghanistan's labour market in 2025 is shaped by three decades of conflict, the 2021 political transition, and the withdrawal of most international aid and development organisations. The formal economy that existed in Kabul during 2002-2021 - NGO sector, international contractors, government civil service supported by aid revenues, a small private banking sector - contracted significantly after 2021. Many of the workers who held formal professional roles in that period have emigrated or shifted to informal employment. The 86.14% informality rate (ILO ILOSTAT 2025) and the dominance of agriculture reflect a labour market where formal institution-based employment is rare and heavily concentrated in the capital.

The most AI-exposed jobs in Afghanistan

Clerical support workers score 8.5/10 and cover just 11,000 workers (0.1% of total employment) - the smallest clerical share of any country in this dataset. What clerical work exists in Afghanistan's formal sector is concentrated in government ministries in Kabul, the remaining operational aid organisations (UNHCR, WFP, ICRC maintain significant local staff), and the small private banking sector serving the dollar-based formal economy. These workers perform document processing, record management, scheduling, and financial reporting - the core clerical tasks that AI tools already target in other countries. The difference is scale: 11,000 workers versus 186,000 in Uruguay or 1,062,000 in Venezuela.

Professionals at 6.5/10 cover 371,000 workers (5.0%). Afghanistan's professional class in 2025 includes doctors in the public and private health system, teachers in the government school and university system, engineers in infrastructure and construction, and lawyers in the formal legal system. This cohort is smaller than it was during the 2010s due to emigration but remains substantial in absolute terms relative to Afghanistan's total workforce. Medical professionals and university educators in particular face AI augmentation pressure through the same global diffusion of AI tools that affects their counterparts in neighbouring Pakistan and India.

Technicians score 5.5/10 at 112,000 workers (1.5%). IT support staff in Kabul's banking and telecoms sector, health technicians in hospitals and clinics, and technical supervisors in construction and engineering projects. Managers at 5.5/10 cover 41,000 workers (0.6%) - the smallest management cohort in this dataset, reflecting the shallow formal corporate sector.

The safest jobs from AI in Afghanistan

Skilled agricultural workers score 3.0/10 in Afghanistan and cover 3,646,000 workers - 49.1% of total employment. Wheat cultivation in the northern plains, rice in Helmand and Kunduz, fruit orchards in Kandahar and Nangarhar, livestock herding across the central highlands, and subsistence mixed farming across rural provinces. These are physical, seasonal, weather-dependent, and geographically dispersed tasks that AI cannot currently perform or meaningfully augment at the smallholder level. Agricultural workers are almost entirely informal, operating outside the institutional structures through which AI tools would be procured.

Craft and trades workers at 2.5/10 cover 1,113,000 workers (15.0%). Carpenters, weavers (Afghanistan's carpet industry is a significant employer), metalworkers, tailors, and construction tradespeople performing skilled manual work that is both non-routine and physically demanding. Elementary occupations at 2.0/10 cover 543,000 workers (7.3%) - loading, cleaning, basic labour, and subsistence activities. Service and sales workers at 3.5/10 (1,164,000 workers, 15.7%) include shopkeepers, market traders, transport workers, and informal service providers.

What this means for Afghan workers

For the approximately 535,000 Afghans in formal professional and clerical roles, AI exposure is structurally real. A doctor in a Kabul hospital who has access to AI diagnostic tools faces the same augmentation dynamic as doctors in Pakistan or India. A university professor using AI writing tools, a bank clerk whose document-processing tasks are being automated, an engineer using AI design software - these individuals experience AI adoption at the individual level regardless of whether their institution has a formal AI strategy. The global accessibility of browser-based AI tools means individual formal-sector workers can and do adopt AI tools without institutional procurement decisions.

For the other 6.9 million Afghan workers - in agriculture, informal trades, and elementary occupations - AI is not a near-term employment concern in the conventional sense. The structural challenges facing these workers are access to land, water, credit, markets, physical security, and basic infrastructure. Addressing these challenges does not require AI policy but does require economic development, governance, and investment. Afghanistan's near-term workforce challenge is not AI displacement but rather creating conditions where more workers can access formal employment, education, and economic opportunity.

The data-literacy note worth adding: a 3.12/10 average does not mean Afghanistan's workers are better off than workers in Uruguay (4.22/10). It means they work in different roles. The Uruguayan clerical worker facing AI displacement has formal employment, social protection, and upskilling pathways. The Afghan subsistence farmer is not displaced by AI - but neither does she have the institutional supports that make labour market transitions manageable.