World Bank pushes 'small AI' to boost developing economies
The World Bank is promoting the use of low-cost AI tools tailored to local needs in developing economies, coinciding with its Global AI and Digitalization Summit in Seoul. The approach focuses on practical applications to support jobs and productivity, while warning that weak connectivity and digital skills may keep applications at the experimental stage.
The World Bank is pushing for the expanded use of low-cost AI tools adapted to local needs in developing economies, coinciding with the scheduled start of its Global AI and Digitalization Summit in Seoul. The summit agenda focuses on practical applications capable of supporting jobs and productivity, rather than measuring progress solely by the size of computer models. The summit is organized in partnership with South Korea's Ministries of Finance and Economy, and Science, ICT and Future Planning. The Bank presents the concept of 'small AI' as an approach to solving specific problems, potentially using simple phones or sensors and limited-capability devices, with services designed to suit available connectivity and data conditions. According to an interview published by The Guardian, the Bank's Knowledge Officer Pascal Donohoe called for adapting AI to the needs of developing economies. The summit agenda clarifies that this approach is not limited to shrinking the model itself; it may include predictive or generative tools, and deliver services via text or voice messages. However, the Bank emphasizes that weak connectivity, digital skills, and the lack of trusted local institutions may keep these applications at the experimental stage, rather than becoming ongoing services. In a separate presentation of the 2026 World Development Report findings, the Bank says developing countries do not need to build general-purpose models at enormous costs to benefit from the technology. But it also warns that importing ready-made tools is not enough.