**Introduction** The main barriers to global artificial intelligence (AI) participation for developing countries arise from weaknesses in digital and energy infrastructure, limited access to computing capacity, underdeveloped data ecosystems, shortages of skilled labor and institutional capability, constrained financing and scale-up conditions, and limited influence over international rules that increasingly shape AI-enabled trade and services. These constraints confine many developing economies to downstream use of AI technologies, with limited scope for domestic capability building in AI development, adaptation, and value creation[1][2]. **Key barriers to global AI participation** **1.** **Infrastructure and power constraints** AI systems are data- and compute-intensive, requiring reliable broadband connectivity, stable electricity, and adequate digital infrastructure. In many developing countries, persistent gaps in network coverage, bandwidth quality, and power reliability increase operating costs and limit the feasibility of deploying AI at scale. Where these foundational conditions are weak, firms and public institutions face higher investment risk and lower expected returns from AI adoption. This constrains experimentation, slows uptake in productive sectors such as logistics, manufacturing, and public administration, and weakens the formation of domestic AI ecosystems[1]. **2.** **Limited access to computing capacity** Beyond basic infrastructure, effective participation in AI increasingly depends on access to cloud services, data centers, and advanced computing resources. These markets are highly concentrated, with a small number of providers controlling critical layers of AI infrastructure. For developing countries, this concentration raises costs and reduces competition, limiting the ability of domestic firms to train, fine-tune, or deploy AI models. As a result, AI use is often oriented toward imported services rather than locally developed capabilities, limiting opportunities for domestic capability building[3]. **3.** **Weak data ecosystems and limited local relevance** AI performance depends on the availability of large volumes of high-quality, context-specific data. In many developing economies, lower levels of digitization across firms and government services constrain data generation, while weaknesses in administrative systems reduce data usability. These constraints are particularly binding for local languages, small markets, and sector-specific applications such as agriculture, health, and small-firm finance. Weak data ecosystems make it difficult to adapt AI systems to domestic conditions, limiting productivity gains and increasing reliance on externally developed models[1][4]. **4.** **Skills and institutional capacity gaps** AI participation requires not only technical skills, but also managerial, regulatory, and organizational capacity. Many developing countries face shortages of advanced digital skills, weak research and training infrastructure, and limited public-sector capability to procure, regulate, and supervise AI systems. These gaps reduce the effectiveness of AI adoption and increase governance risks, particularly in areas such as public services, customs, and financial supervision. Differences in preparedness across infrastructure, human capital, and institutional capacity strongly influence which countries are able to capture AI-related productivity gains[2][5]. **5.** **Financing constraints and limited scale-up capacity** AI capability formation is capital-intensive and closely linked to the depth of domestic innovation and financing ecosystems. Developing countries often face higher costs of capital, limited availability of risk financing, and less developed innovation systems, which constrain firm entry and scale-up. Investment in digital and AI-related activities remains highly concentrated at the global level. This concentration limits opportunities for firms in developing economies to expand, achieve scale, and integrate into global value chains[6]. **6.** **Limited influence over rules and standards** Participation in the AI economy is increasingly shaped by international standards, regulatory frameworks, and trade rules governing data flows, digital services, and technology use. Many developing countries have limited capacity to influence the design of these frameworks and therefore adjust primarily to externally determined requirements. As AI-related provisions are incorporated into trade agreements, public procurement rules, and regulatory cooperation arrangements, compliance costs increase and policy flexibility for domestic capability building is reduced. These requirements can also affect market access for firms that lack the scale or resources to meet complex or evolving regulatory standards[7]. **Conclusion** The barriers to global AI participation for developing countries are constraints in infrastructure and power, access to computing capacity, data availability and governance, skills and institutional capacity, financing and scale-up conditions, and influence over international rules interact and reinforce one another. Without coordinated progress across these areas, AI adoption is likely to remain uneven and to reinforce existing differences in productivity, competitiveness, and participation in global digital trade.
