Russia vs Ukraine: AI Job Risk Before and During War

Two economies, one AI risk profile - then everything changed

The data tells an unusual story. In 2021, Russia and Ukraine had occupational structures so similar that their weighted average AI exposure scores differed by only 0.01 - Russia at 4.62, Ukraine at 4.63 (computed by WorldJobsData from ILO ILOSTAT data). Both economies peaked at 8.5/10 for clerical workers. Both had similar distributions of professionals, technicians, service workers, and trades workers as shares of total employment. Both were Soviet successor states with comparable industrial legacies, educational systems, and labour market patterns.

Then the full-scale invasion began in February 2022. The comparison that was analytically clean in 2021 became impossible to make on current data within months. Ukraine has no reliable post-2022 national labour force survey. Russia's economy continued functioning under sanctions, adapting its labour market in wartime conditions. The two countries now operate in fundamentally different economic realities - but the 2021 data gives us the clearest available baseline for understanding what Ukraine's AI exposure profile was before the disruption, and what Russia's looks like in 2025.

Russia workforce AI exposure - 2025 data

Russia's ILO ILOSTAT 2025 data covers 73,457,000 workers across all major ISCO-08 occupation groups. The table below shows every group with AI exposure scores, worker counts, and median annual wages in USD (ILO ILOSTAT wage series, converted at prevailing exchange rates).

Russia's largest occupation group is Professionals at 28.2% of all workers - 20.7 million people scoring 6.5/10. This single group is larger than Ukraine's entire pre-war workforce. Service and sales workers number 11.2 million (15.3%), technicians 10.9 million (14.9%), trades workers 9.2 million (12.5%), and plant operators 8.9 million (12.1%). The clerical group at 2.7 million is the smallest major category despite carrying the highest AI exposure score of 8.5/10.

Ukraine workforce AI exposure - 2021 pre-war baseline

Ukraine's 2021 dataset does not include a separate Skilled Agricultural workers category or Elementary occupations category at the ISCO-1 level in the available ILO ILOSTAT series. The total covered across reported categories is approximately 14,979,000 workers. Ukraine's pre-war occupational structure closely mirrors Russia's in proportional terms - Professionals are the second-largest group at 22.3%, behind service and sales at 20.3%, with technicians at 14.6% and trades workers at 15.1%.

Ukraine's pre-war wages are notably lower than Russia's even within the same year. Ukrainian managers earned a median of $3,566 per year versus Russia's $18,291. Ukrainian professionals earned $2,600 per year versus $10,358 in Russia. These differences reflect Ukraine's significantly lower GDP per capita - $5,866 (World Bank, 2025 current figure) versus Russia's $17,547 - and lower productivity levels in the pre-war economy.

The sanction paradox: Russia's deployment gap

Russia's risk velocity score of 10.0 - classified as "Disruption imminent (1-3 years)" - reflects the theoretical task susceptibility of the Russian workforce as of 2025. But this number warrants context. Western sanctions imposed between 2022 and 2024 have materially restricted Russia's access to advanced AI hardware, frontier model APIs, and enterprise software from US and European vendors. The gap between theoretical AI exposure and actual deployment is wider in Russia than in comparable economies.

Russian employers cannot freely access GPT-4 class models via commercial APIs. Nvidia GPU exports are restricted under US export controls. Cloud computing capacity for large-scale AI training and inference is constrained. Russia has pursued domestic AI development through state-directed programs, and partnerships with non-sanctioning countries partially fill the gap - but the practical capability available to Russian employers in 2026 is not equivalent to what employers in Germany, South Korea, or the United States can deploy. The 10.0 risk velocity score was calibrated on global technology access assumptions, not on the specific constraints the Russian economy faces.

Russia's official unemployment rate of approximately 2.4% (2024-2025) also reflects a tight labour market shaped by military mobilisation and professional emigration since 2022. When workers are scarce, employers have less economic incentive to automate. This further dampens the near-term deployment pressure despite the high theoretical exposure score.

Ukraine's IT sector exodus - a structural shift in AI exposure

Ukraine had a significant IT sector before the war - estimated at 200,000 to 285,000 active IT professionals (Ukrainian IT Association, 2021 estimates), generating over $6 billion in annual export revenue. This sector sits within the Professionals category (ISCO major group 2, score 6.5/10) - the highest-exposure group in Ukraine's dataset after clerical workers.

Since February 2022, a substantial portion of Ukrainian IT workers have relocated to EU countries. Poland, Germany, Portugal, and the Czech Republic have received the largest inflows, with many workers continuing to work for Ukrainian companies remotely while living abroad. This relocation removes some of the highest-AI-exposure workers from the Ukrainian count - paradoxically reducing Ukraine's measured AI exposure within the country while increasing it in the destination countries. The 2021 baseline figure of 3.33 million Professionals overstates how many remain in Ukraine today.

For post-war reconstruction purposes, this matters. If Ukraine's IT workforce returns after the war ends - as some analysts project based on the geographic concentration of displaced workers in nearby EU countries - Ukraine's Professionals group will reconstitute relatively quickly. If relocation becomes permanent, Ukraine will face a structural shortage of its highest-productivity workers, which has implications for both economic recovery and AI adoption capacity.

Economy context: Russia vs Ukraine side by side

The table below uses World Bank Open Data (CC BY 4.0) and UNDP Human Development Report 2025 (2023 data year, licence CC BY 3.0 IGO) for both countries.

* Ukraine unemployment rate shown is 2021 (World Bank), matching the ILO data year. Current wartime unemployment is not reliably tracked. Ukraine workforce figure is 2021 pre-war baseline only.

Ukraine's GDP per capita of $5,866 (World Bank, 2025) is a wartime figure that reflects both the destruction of economic capacity and ongoing external support. Pre-war 2021 GDP per capita was approximately $3,984 (World Bank), so the current figure includes significant reconstruction aid flows. Russia's $17,547 (World Bank, 2025) reflects an economy that has adapted to sanctions and maintained basic functioning, though growth has been constrained. The 3x GDP gap between the two countries was narrower before the war and has widened since.

Safest jobs from AI in both countries

At the low end of the AI exposure scale, both countries show the same pattern visible across most of the world's major economies. Physical, variable-environment work resists current AI capabilities regardless of geography.

Russia has 9.2 million craft and trades workers in 2025 - a group that scores 2.5/10. Ukraine's pre-war trades workforce numbered 2.26 million at the same score. Both countries inherited large industrial and construction sectors from the Soviet period, producing proportionally large physical-sector workforces relative to the size of their clerical and administrative classes. In Ukraine, trades workers also include significant numbers in construction and industrial maintenance - sectors that are both essential for wartime repair and for post-war reconstruction. These are among the least AI-exposed roles in either dataset.

What this means: Russia's disruption timeline and Ukraine's reconstruction uncertainty

For Russia, the practical question is how quickly AI deployment accelerates once sanctions constraints loosen - or are bypassed. Russia's official unemployment rate of approximately 2.4% reflects a labour market under pressure from military mobilisation. Many of the workers who remain in the civilian economy are in roles with moderate to high AI exposure: 20.7 million professionals scoring 6.5/10, 10.9 million technicians at 5.5/10, and 2.7 million clerical workers at 8.5/10 together represent more than 46 million workers at scores of 5.5 or above. Domestic AI development and partnerships with non-sanctioning technology providers will determine how fast that theoretical exposure converts into actual displacement.

For Ukraine, the question is different in kind. AI displacement is not the primary risk facing Ukrainian workers in 2026. The immediate challenges are physical security, economic survival, and workforce reconstitution. Ukraine's HDI of 0.779 (UNDP HDR 2025, rank 87) reflects strong educational attainment - mean years of schooling of 11.4 years (UNDP HDR 2025) - which is an asset for post-war reconstruction and for reskilling into whatever occupational structure emerges from the war. Ukraine's pre-war Professionals group of 3.33 million had the educational and technical capacity to engage with AI-augmented workflows. The question is whether those workers return, and in what numbers.

Post-war reconstruction will reshape Ukraine's occupational structure in ways impossible to predict from 2021 data. Infrastructure reconstruction - roads, bridges, housing, energy systems - will require large numbers of craft and trades workers (2.5/10 AI exposure). Healthcare reconstruction will require medical professionals and technicians (5.5-6.5/10). Administrative reconstruction of government functions will require clerical workers and managers (5.5-8.5/10). Which sectors rebuild fastest, and at what AI exposure level, will determine whether post-war Ukraine has a higher or lower AI risk profile than the 2021 baseline.