Russia vs Turkey: AI Job Risk in Rival NATO Frenemies

NATO member, Russian weapons buyer, Ukraine mediator: Turkey's unusual position

Turkey's geopolitical position in 2026 is unique among major economies. It is a NATO member since 1952, yet purchased Russia's S-400 missile defence system in 2019, triggering its removal from the F-35 joint strike fighter programme. It controls the Bosporus and Dardanelles straits under the 1936 Montreux Convention, regulating warship passage to and from the Black Sea - a lever of enormous strategic importance since Russia's 2022 invasion of Ukraine. Turkey has hosted multiple rounds of Ukraine peace negotiations in Istanbul, positioning itself as a neutral mediator between NATO allies and Russia simultaneously.

For AI deployment context, Turkey's dual position matters. Turkey has full access to Western AI technology stacks as a NATO member and OECD country. Russian employers face US and EU export controls restricting access to the same tools. The AI exposure comparison between the two countries (Russia 4.62/10, Turkey 4.11/10, ILO ILOSTAT 2025, computed by WorldJobsData) represents comparable theoretical scores - but Turkey's workers are being displaced by technology that is commercially available today, while Russia's high theoretical exposure faces real constraints on tool access.

Turkey workforce AI exposure - 2025 data

Turkey's ILO ILOSTAT 2025 data covers 32,407,000 workers. The Turkish Statistical Institute (TurkStat / Turkiye Istatistik Kurumu) provides national labour force survey data to the ILO. The table below shows Turkey's top-exposure and high-employment occupation groups.

Turkey's largest single occupation group is market-oriented skilled agricultural workers at 3,389,800 workers (10.5%) scoring 3.5/10. This group is the primary reason Turkey's weighted average (4.11/10) sits 0.51 points below Russia's (4.62/10). Turkey's second-largest group by share is sales workers at 2,884,500 (8.9%) scoring 5.0/10. Teaching professionals at 1,452,900 (4.5%) score 6.5/10. Numerical and material recording clerks at 1,019,700 (3.1%) score 8.5/10 - one of the highest individual group headcounts at that score level in Turkey's dataset.

Russia workforce AI exposure - 2025 data

Russia's ILO ILOSTAT 2025 data covers 73,457,000 workers. Rosstat (the Federal State Statistics Service) provides national data to the ILO.

Russia's dominant group is Professionals at 20.7 million (28.2%) scoring 6.5/10. Russia's agricultural workers at 1.5 million (2.1%) score 3.0/10 - a smaller share than Turkey's 10.5% and at a lower score, reflecting Russia's more mechanised and commercially organised agricultural sector. Russia's service and sales workers at 11.2 million (15.3%) score 3.5/10 - the same score as Turkey's agricultural workers, which is notable. In Russia's data, this group is the main downward weight on the average from the low end of the distribution; in Turkey's data, the agricultural group plays that role.

Why Turkey scores 0.51 points below Russia despite similar GDP per capita

The 0.51-point gap between Russia (4.62) and Turkey (4.11) is primarily explained by Turkey's agricultural employment structure. Turkey's market-oriented skilled agricultural workers at 3,389,800 (10.5% of employment) scoring 3.5/10 represent one of the largest agricultural occupation shares among major OECD economies. Russia's skilled agricultural workers at 1,526,300 (2.1%) score 3.0/10 - a smaller share and at a lower score. Turkey's agricultural workers score 0.5 points higher than Russia's (3.5 vs 3.0) but their much larger employment share provides substantially more downward drag on Turkey's weighted average.

A secondary factor is Turkey's large sales workers group. Sales workers at 2,884,500 (8.9%) score 5.0/10 - a mid-range score. Russia's equivalent (service and sales workers at 15.3%) score 3.5/10. Turkey's sales workers are scored higher per worker, but Russia's service and sales group is proportionally larger and drags Russia's average down from the mid-distribution. The net effect favours Russia's higher overall average despite Turkey having a larger professional class as a share of a smaller total workforce.

The safest jobs from AI in Turkey

Turkey's lowest AI exposure occupations are concentrated in physical, elementary, and outdoor roles - consistent with the global pattern but at larger absolute scale than smaller economies.

Turkey's cleaners and helpers at 1,232,200 workers (3.8%) form the largest single low-exposure group, scoring 1.5/10. Agricultural, forestry and fishery labourers at 1,002,600 (3.1%) also score 1.5/10. Together Turkey's elementary and physical workers represent a substantial portion of the workforce that current AI capabilities cannot displace in near-term timelines. Food preparation assistants at 535,700 (1.7%) and refuse workers at 370,700 (1.1%) round out the low-exposure labour pool. These groups are stable against AI displacement regardless of the geopolitical relationship between Ankara and Moscow.

Economy comparison: Russia vs Turkey

The table below uses World Bank Open Data (CC BY 4.0) and OECD data. Turkey is not included in UNDP HDR 2025 and HDI figures are therefore not available for this comparison.

Russia and Turkey's GDP per capita within $1,052 of each other (Russia $17,547, Turkey $18,599, World Bank 2025) is an unusual economic symmetry for two geopolitical rivals. Turkey's OECD membership provides an OECD median wage benchmark of $47,253 (2024) - a figure that reflects formal sector wages in a mixed economy and is not directly comparable to Russia's GDP per capita. Turkey's unemployment rate of 8.52% is substantially higher than Russia's 2.13%, with Russia's rate reflecting military mobilisation effects that have compressed the civilian labour supply rather than full employment in the economic sense.

Turkey is not included in UNDP Human Development Report 2025. This is noted as a data disclosure. WorldJobsData does not have Turkey HDI data from the UNDP HDR 2025 series and does not substitute estimates for official UNDP figures.

The S-400 paradox: same AI exposure risk velocity, opposite technology access

Both Russia and Turkey score 10.0/10 on risk velocity - the highest tier, classified as "Disruption imminent (1-3 years)." This shared score reflects both countries' occupational structures as assessed by global AI capability benchmarks. But the practical technology access situation is dramatically different.

Turkey, as a NATO member and OECD economy, has full access to Western AI infrastructure: AWS, Azure, Google Cloud, OpenAI APIs, Anthropic APIs, and the full range of enterprise AI software from US and European vendors. Turkish employers deploying AI tools face no regulatory export barriers. Russia's employers face US Bureau of Industry and Security (BIS) export controls restricting access to advanced semiconductors, cloud services, and frontier model APIs. Russia's 2019 S-400 purchase - the transaction that fractured its F-35 partnership - symbolises a broader pattern of Russia pursuing strategic autonomy from Western technology supply chains that now extends to AI infrastructure.

The result: Turkey's 4.11/10 average represents near-term actual AI displacement risk, because the tools are available and being deployed. Russia's 4.62/10 represents higher theoretical exposure but with a deployment lag caused by access constraints. Turkey's workers in the clerical, administrative, and professional categories are facing active disruption now. Russia's equivalent workers face disruption on a timeline determined partly by geopolitics.

Why Turkey's agricultural employment matters for the AI comparison

Turkey's market-oriented skilled agricultural workers at 3,389,800 (10.5% of employment) scoring 3.5/10 deserve specific attention in this comparison. This group - farming, horticulture, fishery, and related market-production workers - is much larger as a share of employment in Turkey than in Russia (2.1% at 3.0/10). Turkey's agricultural sector remains a significant employer partly due to the structure of its labour market: lower wages in agriculture relative to urban services, and slower urbanisation in certain regions, sustain a proportionally larger agricultural workforce compared to EU member states of similar GDP per capita.

The 3.5/10 score for Turkish skilled agricultural workers reflects some exposure to AI - precision agriculture tools, crop disease detection, automated irrigation systems, and market pricing algorithms are all commercially available. But full autonomous displacement of skilled agricultural labour is not imminent at scale. The work requires physical presence in variable outdoor environments, adaptability to weather and disease, and judgement calls that current AI cannot replicate reliably. This group provides a substantial low-exposure buffer in Turkey's weighted average calculation that Russia's smaller agricultural sector does not.

What this means for workers in both countries

For Turkish workers, the most immediately at-risk groups are the 1,019,700 numerical and material recording clerks (3.1% at 8.5/10) and the 599,500 general and keyboard clerks (1.9% at 9.0/10). These administrative and data-entry roles are the clearest targets for near-term AI tool deployment in Turkish firms that have access to Western enterprise AI software. The 948,700 business and administration associate professionals (2.9% at 7.5/10) represent a larger mid-risk group where AI augmentation is already beginning but displacement is less uniform.

Turkey's 8.52% unemployment rate (World Bank) is already elevated relative to its workforce size. AI displacement of clerical and administrative workers could compound unemployment if displaced workers cannot transition into lower-exposure roles or higher-skill roles fast enough. Turkey's demographic structure - a younger median age than Russia - gives it more time to build reskilling capacity, but also means more workers entering the labour market who will need to navigate AI-disrupted professional environments.

For Russian workers, the deployment-constrained version of disruption is still real. Russia's domestic AI industry - Sber AI, Yandex, and various state-directed research institutes - is producing tools that are already deployed in Russian enterprises. Cloud infrastructure partnerships with non-sanctioning countries partially substitute for Western provider access. The 20.7 million professionals at 6.5/10 and 10.9 million technicians at 5.5/10 face a disruption timeline that is slower than Turkey's but not indefinite.