Serbia's 2.85 million workers score a weighted average AI exposure of 4.69/10 - the lowest in this batch of Southeast European economies, reflecting a labour market still anchored in manufacturing, construction, and agriculture. But Belgrade's rapid emergence as a regional technology hub is changing that mix fast. Serbia is an EU candidate country (since 2012), and EU candidate status accelerates tech adoption timelines in a way that separates Serbia from non-candidate Western Balkan peers. The Belgrade IT sector employed over 80,000 workers in 2025 (per SORS), growing at approximately 8% annually since 2020.
Serbia's employment data comes from ILO ILOSTAT 2025, collected by SORS - Republicki zavod za statistiku (Statistical Office of the Republic of Serbia) - using ILO-harmonised Labour Force Survey methodology and ISCO-08 occupation classification. Serbia is not an EU member, so Eurostat data is not available; SORS data is used by international bodies including the ILO and World Bank as the authoritative source for Serbian labour market statistics.
| Occupation Group | AI Score | Workers (est.) | Share (est.) |
|---|---|---|---|
| Clerical support workers (ISCO 4) | 8.5/10 | ~229,000 | ~8.0% |
| Professionals (ISCO 2) | 7.0/10 | ~538,000 | ~18.9% |
| Technicians and associate professionals (ISCO 3) | 6.2/10 | ~350,000 | ~12.3% |
| Managers (ISCO 1) | 5.1/10 | ~114,000 | ~4.0% |
| Service and sales workers (ISCO 5) | 3.7/10 | ~434,000 | ~15.2% |
| Skilled agricultural workers (ISCO 6) | 3.5/10 | ~281,000 | ~9.9% |
| Plant and machine operators (ISCO 8) | 2.9/10 | ~348,000 | ~12.2% |
| Craft and related trades (ISCO 7) | 2.7/10 | ~328,000 | ~11.5% |
| Armed forces (ISCO 0) | 1.9/10 | ~20,000 | ~0.7% |
| Elementary occupations (ISCO 9) | 1.6/10 | ~205,000 | ~7.2% |
Within clerical support (ISCO 4), general and keyboard clerks (ISCO 41) score 9.0/10 - the single highest score in Serbia's occupation structure. Customer service clerks (ISCO 42) score 8.5/10 and numerical and material recording clerks (ISCO 43) score 8.5/10. These groups are concentrated in Belgrade's financial sector, government administration, and the growing shared services industry. Serbia has attracted regional back-office operations from firms including NCR Atleos, Nordeus (acquired by Take-Two), and HTEC Group, whose operations employ hundreds of clerks and business support workers in Belgrade and Novi Sad.
The professional group (ISCO 2) at nearly 19% of the workforce is the second largest single group in Serbia and scores 7.0/10. ICT professionals (ISCO 25) score 8.5/10 and represent a growing share within professionals - Serbia's IT sector is well above the Western Balkans average. Health professionals (ISCO 22) score 5.0/10 - a large sub-group given Serbia's extensive public health system. Teaching professionals (ISCO 23) score 6.5/10.
Serbia in 2026 is two labour markets running in parallel. Belgrade - home to approximately 25% of Serbia's total population - functions as a Central European-tier technology hub. The Serbian IT sector generated approximately EUR 3.3 billion in export revenue in 2024 (per Serbia's Chamber of Commerce and Industry), driven by software development, IT services outsourcing, and the operations of global tech firms that have established Serbian engineering teams. Novi Sad, Serbia's second city, hosts a significant share of IT exports and a growing startup scene.
Outside Belgrade and Novi Sad, Serbia remains a primarily manufacturing and agricultural economy. The Vojvodina region is heavily agricultural. Serbia's automotive manufacturing sector is anchored by Fiat/Stellantis in Kragujevac, employing thousands of assembly workers who score in the 2.5-3.0/10 range. The construction sector employs a large craft and trades workforce (ISCO 71, 72, 73) that collectively scores below 3.0/10. Agriculture at approximately 10% of the formal workforce - and higher in the informal economy - scores 3.5/10.
The dual economy structure is a real feature of Serbia's labour market despite having the second-fastest growing IT sector in the region. Sizeable shares of plant and machine operators (estimated 12% of formal employment), agricultural workers (10%), craft workers (12%), and elementary workers (7%) sit well below the clerical and professional peaks and moderate the overall average. Serbia's exposure profile reflects an occupation mix that still contains substantial physical labour alongside its fast-growing knowledge sector.
EU candidate status is an important accelerator. Serbia applied for EU membership in 2009, received candidate status in 2012, and opened accession negotiations in 2014. This alignment with EU regulatory frameworks - including the EU AI Act, GDPR, and labour market directives - means Serbian employers adopt AI-enabled enterprise software on the same procurement cycle as EU members, often through the same multinational vendors. Workers in Belgrade's finance and professional services sector face AI substitution risk on a similar timeline to Prague or Warsaw, not on a later Balkan timetable.
Serbia's traditional economy - manufacturing, construction, agriculture, and physical services - protects approximately 40% of the workforce from near-term AI disruption.
| Occupation Group | AI Score | Workers (est.) | Share (est.) |
|---|---|---|---|
| Elementary occupations (ISCO 9) | 1.6/10 | ~205,000 | ~7.2% |
| Building trades workers (ISCO 71) | 2.0/10 | ~81,000 | ~2.8% |
| Vehicle drivers and operators (ISCO 83) | 2.5/10 | ~150,000 | ~5.3% |
| Craft and related trades (ISCO 7 avg) | 2.7/10 | ~328,000 | ~11.5% |
| Skilled agricultural workers (ISCO 6) | 3.5/10 | ~281,000 | ~9.9% |
Craft and trades workers (ISCO 7) at an estimated 11.5% of formal employment is a substantial major group alongside service and sales workers and professionals. The group includes construction workers, metalworkers, auto assemblers, and precision mechanics employed at Stellantis Kragujevac, Gorenje Serbia, and automotive parts manufacturers. These roles require physical dexterity, spatial reasoning in unstructured environments, and quality judgment that current industrial robotics cannot cost-effectively replicate in Serbia's factory footprint.
Agricultural workers (ISCO 6) at approximately 10% of formal employment - and a substantially higher share of total employment including informal and subsistence farming - score 3.5/10. The Vojvodina region produces over 60% of Serbia's agricultural output (Serbian Chamber of Agriculture, 2024). Seasonal harvest work, animal husbandry, and smallholder farming remain labour-intensive at Serbia's agricultural scale and economic conditions. Vehicle drivers (ISCO 83) at 2.5/10 reflect Serbia's position as a transit country between EU markets and Turkey, sustaining demand for truck drivers along the E75 and E70 corridors.
Serbia's 4.69/10 average is the lowest in this batch of Southeast European and Central European economies - below Bulgaria (4.71/10), Greece (4.88/10), Hungary (4.82/10), Czechia (4.84/10), and well below Croatia (4.93/10). This lower score is a structural artifact of the occupation mix, not evidence that Serbian workers face less AI risk in their specific roles. A Serbian general clerk faces the exact same 9.0/10 risk as a Czech general clerk - the difference is that clerks are a smaller share of Serbia's total workforce than they are in more service-oriented economies.
If you work in Belgrade's IT, finance, or professional services sector, your exposure profile mirrors Central European peer cities, not a lagging Balkan economy. AI tools for software development, financial reporting, customer service automation, and business process outsourcing are actively being deployed by the multinationals with Serbian operations. The 2 to 3 year adoption lag versus Western Europe that applied in 2020 has compressed to near-zero for firms that operate across EU markets and use global SaaS platforms.
Recovery resilience of 6.2/10 is the lowest in this batch - primarily because Serbia is not yet an EU member and lacks full access to EU Cohesion Funds, Just Transition Funds, and European Social Fund retraining programs. Domestic retraining infrastructure exists (the National Employment Service runs vocational training programs in 27 cities) but is less resourced than EU member state equivalents. If you are considering career decisions in a high-exposure role, the transition support available in Serbia if your role is automated is thinner than in Romania, Hungary, or Czechia. That asymmetry is worth factoring into planning horizons of 5 to 10 years.
Interactive breakdown of every occupation group, sortable by AI exposure score and worker count.
View Serbia DataYour reaction helps us understand what to cover next.
Thanks for your reaction!
Get new country analyses, data updates, and AI labour market insights. No spam - one email when something worth reading drops.
AI exposure, robotics, offshoring and work-from-home scores are model-estimated: an AI language model (Claude Opus 4.7) scored each ISCO-08 occupation group on 28 May 2026, guided by published research. They are informed judgements, not measurements, and not forecasts of job losses. "Risk velocity" and its time windows come from a digital-infrastructure index (broadband and secure servers), not from occupations or measured AI adoption. Read the full methodology and how the scores were checked.
Spotted an error or know a better source? Report it and we will check it against the source data.