World Bank Warns Balkans Risk Falling Behind on AI

By Suad Seferi · 7 October 2026 at 10:46

Artificial intelligence is spreading faster than economies across Europe and Central Asia are prepared to absorb it, with limited business adoption, shortages in computing capacity and weak organisational readiness threatening to widen the technology gap, according to a new World Bank report. The World Bank's latest Europe and Central Asia Economic Update, published on October 6, includes the Western Balkans within the region it examines and puts artificial intelligence at the centre of its assessment of future productivity and employment. Its message is unusual at a time when much of the public debate focuses on the risks of excessive AI deployment. For developing economies in Europe and Central Asia, the World Bank argues that the greater risk over the coming decade may be too little AI adoption, rather than too much. Around one in five workers across the region already holds a job considered meaningfully exposed to artificial intelligence. Yet fewer than one in ten firms currently use AI in their operations, approximately half the share recorded among companies in the European Union. Even among businesses that have adopted the technology, use is often limited. The report says firms commonly apply AI to routine activities such as answering customer queries, managing internal knowledge and preparing documents. These uses have so far produced little measurable improvement in productivity at the wider economic level. The gap is not only about AI models The World Bank identifies several structural barriers that could matter particularly for smaller developing economies. Computing infrastructure is one of them. Data-centre capacity across the developing economies covered by the report stands at approximately 2.1 megawatts per million people, compared with 31.6 megawatts per million in high-income economies outside the region. Data availability presents another problem. Organisations may hold substantial amounts of information, but much of it remains fragmented, unstructured or difficult to use across systems. Skills are also becoming a constraint. The region has a relatively strong supply of STEM graduates, according to the World Bank, but fewer professionals with the combination of technical, managerial and business skills needed to redesign organisations around new technologies. Demand for workers with AI skills has nearly quadrupled. For the Western Balkans, these findings add another dimension to an AI debate that has often concentrated on regulation, education and access to tools. Using ChatGPT or another generative AI service is not the same as integrating AI into companies, public administration or production systems. That distinction becomes increasingly important as governments across the region develop AI strategies, digital programmes and new infrastructure projects. Jobs exposed does not mean jobs lost The World Bank also cautions against treating AI exposure as a direct measure of job displacement. Workers whose occupations are highly exposed to AI tend to be younger and more highly educated, while women are also disproportionately represented among exposed occupations. But exposure can result in different outcomes. AI may automate parts of a job, assist workers with existing tasks or create demand for different skills. Early labour-market evidence examined by the World Bank points more toward reduced hiring in some areas than widespread displacement of existing workers. It also finds cases where AI is complementing work rather than replacing it. The challenge is therefore not simply protecting existing jobs. Countries will also need workers capable of moving into roles where AI becomes part of everyday work. Adoption becomes an economic question The report arrives as many countries in Europe face shrinking working-age populations and persistent productivity problems. Economic growth across Europe and Central Asia is expected to slow from 2.6% in 2025 to 2.2% in 2026, according to the World Bank. AI could partly compensate by allowing workers and businesses to produce more with existing resources. But that depends on countries having the infrastructure, skills, capital and usable data required to move beyond experimentation. The World Bank recommends expanding affordable access to computing capacity, strengthening digital and managerial skills, improving data systems and helping workers transition toward roles that complement AI. Private companies are expected to drive much of that adoption, but governments also have a role in building the foundations on which it depends. For the Western Balkans, the central question may therefore be shifting. Access to AI tools is becoming widespread. The harder test is whether businesses, institutions and workers can turn that access into meaningful economic use before the gap with more advanced economies grows wider. The World Bank is presenting the findings in a dedicated discussion on October 7 focused on how Europe and Central Asia can close gaps in computing power, data readiness and skills.