Bosnia and Herzegovina AI Job Risk 2026: Which of 1.25M Workers Are Most Exposed?

1.25 million workers, ILO ILOSTAT 2025 data

Employment data comes from ILO ILOSTAT (CC BY 4.0), based on the Agency for Statistics of Bosnia and Herzegovina (BHAS), Labour Force Survey 2025. ISCO-08 major group classifications, aggregated from 2-digit ISCO sub-groups published in the BHAS Labour Force Survey microdata. The 2025 data is the most current available for Bosnia and Herzegovina and covers approximately 1,247,700 formally employed workers. BHAS conducts the Labour Force Survey quarterly in accordance with EU methodology standards - Bosnia's EU accession process (candidate status granted December 2022) has driven methodological harmonisation with Eurostat requirements, which improves data comparability with Croatian and Serbian figures.

Bosnia's labour market operates under a structural constraint unique in Europe: the Dayton Peace Agreement of 1995 created a constitutional architecture that divides administrative authority across two entities (the Federation of Bosnia and Herzegovina, itself divided into 10 cantons, and Republika Srpska), the Brcko District, and the state-level government. Each layer employs civil servants, administrators, judges, and support staff. The result is a public sector that is large relative to the economy's size. By most estimates (World Bank, IMF Article IV consultations), the public sector employs 35-40% of formal workers in Bosnia - among the highest ratios in Europe outside of Belarus. This is the structural reason why clerical and professional occupation shares are elevated relative to neighbours like Croatia or Serbia, which have comparable economies but less fragmented governance.

The 2025 data also reflects Bosnia's post-pandemic economic recovery and the acceleration of EU-related structural reforms. Net emigration is a significant factor: Bosnia has one of the highest emigration rates in Europe, with an estimated 150,000-200,000 workers having left primarily for Germany, Austria, and Slovenia since 2013 (when Bosnian citizens gained Schengen visa-free travel). The workers who remain are disproportionately in public sector employment and in occupations with lower mobility - which affects the composition of the 2025 labour force data relative to the pre-2013 baseline.

The most AI-exposed jobs in Bosnia and Herzegovina

Clerical support workers score 8.5/10 on AI exposure and cover 6.71% of employment (approximately 84,000 workers). Within the 2-digit ISCO breakdown, general and keyboard clerks (ISCO 41, 2.86%) carry the sharpest AI risk: data entry operators, administrative assistants processing government forms, and document filing staff across Bosnia's public sector entities are precisely the workers that AI document processing, robotic process automation, and natural language tools target first. Customer service clerks (ISCO 42, 1.48%) in banking, insurance, and telecommunications are the second sub-group - sectors where AI-assisted customer interaction tools are already being deployed by regional telecoms operators including BH Telecom and m:tel.

Professionals at 6.5/10 cover approximately 225,000 workers (18.02%). The largest sub-group, legal, social, and cultural professionals (ISCO 26, 4.91%), is structurally linked to Bosnia's complex legal system - the BiH constitution alone generates a multi-layered legal framework requiring extensive professional legal interpretation and case management. AI legal research and document analysis tools are advancing rapidly in English-language markets and will reach the region as Bosnian legal databases become digitised and machine-readable under the EU accession harmonisation process. Teaching professionals (ISCO 23, 4.30%) - the second-largest sub-group - work across Bosnia's tri-parallel education systems (three separate curricula operating in some cantons of the Federation, along with the RS system), an arrangement that creates demand for curriculum designers, translators, and content producers whose work AI tools increasingly assist.

Technicians and associate professionals score 5.5/10 (approximately 111,000 workers, 8.87%) and managers 5.5/10 (approximately 34,000 workers, 2.73%). The technicians group includes health associate professionals (ISCO 32, 2.56%) - nursing staff and paramedics in the BiH health system - and business and administration associates (ISCO 33, 2.63%) in middle-office finance and HR roles. These are exactly the roles where AI-assisted scheduling, diagnostic triage support, and HR automation tools are having early-stage impact in peer EU accession economies.

The safest jobs in Bosnia and Herzegovina

Elementary occupations score 2.0/10 on AI exposure - the lowest in the ISCO-08 framework - and cover approximately 118,000 workers at 9.45% of employment. This group includes refuse and street cleaners (ISCO 91), agricultural, forestry, and fishery labourers performing simple tasks (ISCO 92), labourers in mining, construction, manufacturing, and transport (ISCO 93), food preparation assistants (ISCO 94), and street and market vendors (ISCO 96). Bosnia's service sector, construction activity, and agro-food processing create sustained demand for elementary workers. The physical, variable, and low-capital character of elementary work means current AI cannot automate these roles on any near-term horizon - and Bosnia's wage levels do not generate the automation economics that might justify robot deployment in these roles even where technology exists.

Craft and related trades workers score 2.5/10 (approximately 229,000 workers, 18.33%). Bosnia's industrial identity is built around its craft and manufacturing sectors. The metals industry - centred on the Zenica steel complex (ArcelorMittal BiH is the largest private employer in the country) and smaller metalworking operations across the Federation - employs metal and machinery workers (ISCO 72). The wood processing industry, particularly in Republika Srpska, employs furniture and woodworking craft workers (ISCO 73). Construction workers (ISCO 71) are active across all entities as post-war reconstruction continued through the 2000s and 2010s and new EU-funded infrastructure projects (roads, wastewater) have kept construction active into the 2020s. Food processing workers (ISCO 75) complete the picture in the agro-industrial sector.

Plant and machine operators score 3.0/10 (approximately 133,000 workers, 10.68%). This is a larger share than most of Bosnia's Balkan neighbours, reflecting the remaining industrial base in metals, chemicals, and food processing. ISCO 83 - drivers and mobile plant operators at 7.74% - is the largest sub-group, encompassing truck drivers, bus drivers, and heavy equipment operators whose physical and variable-environment work is not automatable on current autonomous vehicle timelines in Bosnia's road network conditions.

What this means for Bosnia and Herzegovina workers

Bosnia's 4.31/10 weighted average AI exposure sits in the amber zone - neither the high-risk profile of a casino economy like Macau (5.74) nor the low-risk profile of a predominantly agricultural economy. The 4.31 score is driven upward by the professional and clerical groups linked to the public sector, and held down by the craft, plant operator, and elementary occupation groups linked to the industrial and service base. The key structural risk is not that AI will displace Bosnian workers broadly, but that it will accelerate pressure on the public sector's clerical and professional workforce precisely as EU accession requirements drive digital government reforms.

For the approximately 84,000 clerical workers - government administrators, bank tellers, data entry operators - the AI exposure score of 8.5/10 is the highest in the workforce. The EU accession process is directly relevant here: EU digital government standards require member states to digitise public services and reduce paper-based administrative processes. As Bosnia progresses toward accession (screening process began in 2023), the pressure to automate routine administrative tasks will intensify. Workers in cantonal and entity government offices who spend their days processing paper forms, entering data into legacy systems, and managing document filing are in the highest-displacement-risk positions. The practical protection is not to resist digitisation but to develop the skills to operate and supervise the digital systems that will replace manual processes - a transition that has occurred in Croatia, Slovenia, and other recent EU accession states.

For the 225,000 professionals - teachers, lawyers, healthcare workers, ICT professionals - the 6.5/10 exposure score reflects real but medium-term risk. AI tools in education (automated marking, personalised tutoring, curriculum generation), legal research, and diagnostic support are advancing. Bosnia's EU accession pathway means technology transfer from EU partners will accelerate, and the BiH healthcare system's digital modernisation (partly funded through EU IPA instruments and World Bank health sector projects) will bring AI-assisted diagnostic tools sooner than market forces alone would. Teachers and healthcare workers have professional protections and public sector employment stability that create a buffer - but workers early in their careers should treat AI tool literacy as a professional requirement, not an optional skill. For service, craft, and elementary workers - which together account for nearly half the workforce - near-term AI displacement risk is low, and the more relevant challenges are wages, emigration pressure, and the structural economic reforms required by the EU accession process.