**Introduction** Artificial intelligence (AI)-driven automation is expected to raise productivity and support output growth over time, while also posing challenges for labor market adjustment and income distribution in developing countries. Average exposure to AI-related task automation is lower in low-income economies than in advanced economies. However, high informality, skills constraints, and uneven diffusion of technology can limit the employment gains associated with productivity growth and increase adjustment pressures as automation expands[1][2][3]. **AI impacts on employment and economic growth** **1.** **Implications for employment in developing economies** Across income groups, around 40% of global employment is exposed to AI-related task automation. The share is higher in advanced economies (around 60% of employment), compared with about 40% in emerging market economies and approximately 26% in low-income countries. Lower shares in poorer economies reflect differences in occupational structure rather than an absence of potential impact[1]. For generative AI, 11% of total employment in low-income countries has some degree of exposure, compared with 34% in high-income countries. In developing economies, exposure is concentrated in occupations where AI affects specific tasks rather than replacing entire jobs[2]. These patterns are relevant in India and the Philippines, where IT-enabled services, business process outsourcing, and administrative services employ large numbers of urban workers in occupations with measurable AI exposure. In these sectors, task automation can reduce labor demand growth even when output continues to expand[2][3]. In Vietnam, the adoption of AI-enabled quality inspection and production management systems in electronics and apparel manufacturing has supported productivity and export performance, while employment growth has shifted toward more technical occupations, limiting demand for routine assembly labor[2][4]. **2.** **Productivity gains from AI are unevenly distributed across firms and sectors** AI adoption raises productivity through improvements in efficiency, quality control, and information processing. Widespread adoption of AI technologies can raise output by close to 10% over the medium term. These gains, however, are not evenly distributed across firms or sectors[1]. In developing countries, adoption is more prevalent among larger firms and multinational companies with access to capital, data, and skilled labor. This pattern is evident in Mexico’s automotive and electronics sectors, where AI-enabled production systems have improved productivity and integration into regional value chains. Smaller domestic suppliers face higher barriers to adoption, contributing to output growth that outpaces employment growth in technologically advanced segments of manufacturing[1][3]. **3.** **Constraints on labor reallocation** Labor market adjustment following AI adoption depends on the ability of workers to move between jobs, sectors, and tasks while maintaining income and skills. In economies with high levels of informality, these adjustment channels are weaker. Informal employment is characterized by limited access to training, social protection, and employment services, which reduces workers’ capacity to transition into new formal jobs when tasks are automated[4]. Informal employment accounts for around 58% of global employment, with substantially higher shares in low- and lower-middle-income economies. In this setting, workers affected by automation are more likely to move between informal jobs or experience downward occupational mobility, rather than transition into higher-productivity activities. This limits the extent to which productivity gains from automation translate into broad-based income growth[2][4]. In Bangladesh’s garment sector, efficiency-enhancing technologies linked to digital production management and compliance monitoring have supported competitiveness and export performance. At the same time, limited reskilling capacity and high informality have constrained workers’ ability to move into new formal roles, increasing adjustment costs and slowing the diffusion of productivity gains across the workforce[2][4]. **4.** **Distributional effects vary within developing countries** Within developing countries, exposure to AI-related task automation differs across various segments of the workforce. Urban workers, women, and workers with higher levels of education, particularly in clerical, administrative, and professional occupations face greater exposure. Such occupations account for around 12% of employment in low-income countries and about 15% in lower-middle-income countries[3]. In Malaysia, higher investment in digital infrastructure and workforce skills has supported the adoption of AI alongside upgrading into higher-productivity activities and more stable employment outcomes. Where these complementary factors are weaker, automation is associated with wider wage dispersion and greater differences in productivity across firms[3][5]. **Conclusion** AI-driven automation in developing countries is expected to support productivity growth while placing pressure on labor market adjustment and income distribution. Although average exposure to AI-related task change is lower than in advanced economies, structural characteristics such as high informality, skills constraints, and uneven diffusion of technology can limit the employment gains associated with automation. Economic outcomes depend on whether productivity gains extend beyond leading firms and whether displaced workers are able to transition into higher-productivity activities rather than remain in informal or lower-quality employment.
