Iran AI Job Risk 2026: Which of 24 Million Workers Are Most Exposed?

The most AI-exposed occupations in Iran

Clerical support workers score 8.5/10 across 843,800 workers (3.51%). This is a relatively small clerical share by regional standards - Saudi Arabia's clerical-adjacent administrative apparatus, Turkey's large public sector bureaucracy, and Egypt's civil service all have higher clerical concentrations. Iran's clerical share reflects a mixed economy where state enterprises employ many workers in skilled blue-collar and technical roles rather than administrative roles. Nonetheless, the 843,800 workers in clerical roles face acute AI exposure: data entry, government form processing, banking back-office, and insurance administration are all LLM-automatable tasks that Iranian employers - once AI tools become accessible - will deploy.

Professionals are the most significant exposure group by size: 3,013,500 workers (12.52%) at 6.5/10. Iran's professional class is notably large for a country at this income level. The 1979 Islamic Revolution prioritised engineering, medicine, and science education; Sharif University of Technology, the University of Tehran, and a network of technical universities have produced graduates at a rate that has outpaced private-sector job creation for decades. This mismatch - qualified professionals in roles that underuse their training - leaves many Iranian professionals effectively in semi-clerical functions that AI will displace. Technicians score 5.5/10 across 1,430,600 workers (5.94%); this group includes oil and gas field technicians at NIOC, telecom workers at Hamrah-e Avval (MCI), and industrial facility operators.

Sanctions, the state economy, and AI deployment delay

Iran's AI disruption timeline is structurally different from every other country in this analysis. US sanctions under the Iran Freedom and Counter-Proliferation Act, combined with EU sanctions and secondary sanctions on third-country entities, restrict Iranian access to: Nvidia H100/A100 AI training chips, AMD Instinct GPUs, major cloud AI platforms (AWS SageMaker, Google Vertex AI, Azure OpenAI), and primary LLM APIs (OpenAI, Anthropic, Google). Iranian companies and research institutions access AI tools through VPNs, resellers in Turkey and UAE, and domestically developed alternatives - but at substantial friction and cost.

The Islamic Republic has responded to this constraint with domestic AI development: the National AI Document (2021), the establishment of the AI Centre within the ICT Ministry, and programmes at Sharif University and IUST (Iran University of Science and Technology). Iran publishes a substantial volume of AI research papers - ranked in the top 15 globally by volume in some 2023 surveys - but publication volume and deployable product are different things. Iran's AI hardware constraint is the binding factor: without access to cutting-edge GPUs, Iranian companies cannot train competitive large models, and without domestic LLM deployment, clerical and professional AI automation stalls.

The result is a risk velocity score of 4.9/10 - disruption delayed 7-12 years. This is not a permanent reprieve: if sanctions lift (as they partially did under the 2015 JCPOA), Iranian AI deployment could accelerate rapidly given the existing human capital base. Workers in exposed occupations should treat the sanctions-delay period as a transition window, not a structural protection.

The safest jobs from AI in Iran

Craft and related trades workers are the largest occupation group at 18.99% (4,572,400 workers) scoring 2.5/10. Iran has a large craft economy: construction workers building the rapid urbanisation of Tehran, Isfahan, Mashhad and secondary cities; oil and gas installation and maintenance workers in Khuzestan and the Persian Gulf; and a significant traditional artisan sector (Persian carpet weavers, tile craftsmen, metalworkers) that is partly export-oriented and partly domestic. The physical, site-specific nature of trades work makes it AI-resistant.

Plant and machine operators score 3.0/10 across 3,448,400 workers (14.32%) - a substantial share reflecting Iran's industrial base: automotive manufacturing (Iran Khodro, SAIPA), petrochemical complexes (Pars Special Economic Energy Zone), steel mills (Mobarakeh Steel), and textiles. These roles involve operating industrial machinery that is already partially automated in their Iranian context; further robotics automation faces the same hardware import restrictions as AI deployment. Skilled agricultural workers at 10.53% (2,534,900) score 3.0/10; Iran has diverse agricultural zones - wheat in Khuzestan, pistachios in Kerman, saffron in Khorasan - and a significant rural workforce that will be the last to be affected by AI.

What this means for workers

For Iranian workers in exposed occupations - clerical, professional, technical - the sanctions period creates a window that does not exist elsewhere: 7-12 years in which global AI capabilities will be visible (Iranians have access to AI news and research) but deployment in Iranian workplaces will be slower than in Turkey, Egypt, or the Gulf states. This window is best used for skill development toward AI-complementary roles: data science, AI operations, prompt engineering, and technical review work that human oversight of AI systems will require.

The professional sector is particularly well-placed to benefit from this window. Iranian engineers and IT professionals already have significant experience with workarounds and constrained-environment problem solving - a skill set that transfers well to managing AI systems under resource constraints. The OECD notes that Iran's annual emigration of skilled workers (the "brain drain" phenomenon, particularly to Canada, USA, Germany) has accelerated since 2022; workers planning to emigrate to economies where AI is already deployed should prioritise AI skills as a differentiator in those destination labour markets.

The sovereign buffer of 9.0/10 does not mean workers can rely on state support. Iran's hyperinflation - the IMF estimated 42.5% CPI inflation in 2024 - means real wages and real state benefits have eroded sharply. State employment provides nominal stability but not real income protection. The long-term structural argument for Iran's low AI risk score (3.71/10) rests on the occupational distribution, not on political protections.

Related analyses