As artificial intelligence (AI) is shaping the global economy and society as a general-purpose technology, concerns are growing that disparities in countries’ capacity to adopt AI may further widen. While AI holds significant potential to enhance productivity, foster economic growth, and expand trade, its benefits could be increasingly concentrated in a limited number of leading countries and firms due to the unequal distribution of technology, capital, talent, and data. In particular, many developing countries face structural constraints in adopting AI as a result of insufficient digital infrastructure, limited human capital, and underdeveloped institutional frameworks. Over the medium to long term, these constraints risk exacerbating global growth gaps and socioeconomic inequality.

Against this backdrop, this study focuses on countries’ AI adoption capacity and analyzes disparities in AI readiness across income levels, with the aim of identifying cooperation strategies tailored to the characteristics of different country groups. It also examines the agendas, policies, and initiatives introduced by leading countries−such as the United States, China, the European Union, Japan, and Singapore−as well as by multilateral cooperation frameworks to support capacity building for AI in developing countries. Through this analysis, the study closely reviews international trends and derives policy implications for Korea’s role in this evolving landscape.

To explore customized cooperation strategies by country group, Chapter 2 employs the IMF’s AI Preparedness Index (AIPI) to assess national AI adoption capacity and analyzes the relationship between AI readiness and SDGs achievement indicators by income group. Significant disparities in AI preparedness exist across countries. Most low-income countries face urgent needs in infrastructure development and basic human capital formation. By contrast, developing countries that have achieved a certain level of infrastructure and human capital have reached a level where technology and innovation cooperation is feasible, and they have also made considerable progress in regulatory frameworks. These cross-country gaps are closely linked to differences in both the level and quality of progress toward the Sustainable Development Goals (SDGs), underscoring the need for differentiated cooperation strategies tailored to country-specific conditions.

Chapter 3 examines the policies and programs of major AI-leading countries−including the United States, China, the European Union, Japan, and Singapore−toward developing countries, analyzing government-led initiatives, public-private partnerships, and multilateral cooperation efforts. International cooperation in AI by leading countries reflects a combination of market-expansion objectives and broader goals related to geopolitical and economic stability. In the context of strategic competition between the United States and China, the role of countries that share similar positions with Korea has become increasingly important. Under these circumstances, it is desirable for Korea to support the adoption of AI based on shared values within the international community and to promote economic cooperation that facilitates its diffusion.

Chapter 4 reviews discussions on AI-related cooperation and initiatives for developing countries within major multilateral frameworks, including the G7, G20, OECD, United Nations, ITU, multilateral development banks (MDBs), and the WTO. These institutions recognize limited data access, inadequate digital infrastructure, and shortages of skilled human resources in developing countries as core challenges. At the same time, they are strengthening efforts to link development cooperation with AI ethics, safety, and standards. Such multilateral efforts complement bilateral cooperation while serving as critical platforms for coordination and the formation of global AI governance.

Based on the foregoing analysis, Chapter 5 presents Korea’s AI cooperation policy as follows. First, cooperation should be customized based on country-specific gaps, with the content and form of cooperation differentiated according to income levels and AI readiness. Second, cooperation should focus on priority areas in which Korea has comparative strengths. Korea should prioritize its areas of strength and develop cooperation projects aligned with the partner country’s level of development, while projects should be designed to contribute simultaneously to partner countries’ SDG achievement. At the same time, cooperation grounded in shared values−such as AI safety, data security, and personal data protection−should be pursued. Third, both bilateral and multilateral cooperation frameworks should be considered in parallel. While AI diffusion serves bilateral interests, it also requires participation in and contributions to multilateral frameworks, given considerations of objectives and scale. Accordingly, Korea’s active role in multilateral cooperation mechanisms is essential, and collaboration with like-minded partners will be particularly important in3 this process.