Artificial intelligence readiness is often viewed as a reflection of national wealth, digital infrastructure, and economic power, but evidence from 24 overperforming countries shows that policy choices can play a decisive role in shaping a country’s ability to adopt and govern AI.
The global artificial intelligence divide remains significant. High-income countries continue to dominate global rankings for AI preparedness, while many low- and middle-income countries remain far behind. However, a new study challenges the idea that economic strength alone determines AI readiness. It shows that several countries are outperforming what their economic structure would normally predict, proving that deliberate policy action can help nations move ahead despite resource limitations.
The study uses the International Monetary Fund’s 2023 AI Preparedness Index, which measures countries across four major areas: digital infrastructure, human capital, regulation and ethics, and innovation. It compares these scores with each country’s level of economic complexity, which reflects the sophistication of what a country produces, exports, and researches. Countries with more complex economies are generally expected to perform better on AI readiness, but the study identifies nations that score significantly above that expected level.
In total, the research identifies 24 AI overperformers. Among high-income countries, Australia, Denmark, Finland, Hong Kong SAR China, Japan, the Republic of Korea, the Netherlands, New Zealand, Norway, and Singapore stand out as global overperformers. These countries already have advanced economies, but their AI preparedness scores still exceed what their economic complexity would predict.
The study also identifies several upper-middle-income overperformers, including Albania, China, Costa Rica, Indonesia, Kazakhstan, Malaysia, and Ukraine. Among low- and lower-middle-income countries, Ghana, India, Morocco, Rwanda, Sri Lanka, Tunisia, and Viet Nam emerge as important examples of countries that are moving faster than expected in AI readiness.
Rwanda is one of the most striking cases. Despite having one of the lowest per capita incomes in the world, it has established a dedicated Responsible AI Office, introduced privacy protection laws aligned with international standards, and achieved an AI readiness score that places it ahead of many countries with greater resources. Rwanda’s example shows that even resource-constrained countries can make meaningful progress when they prioritize governance, coordination, and long-term digital capacity.
One of the study’s strongest findings is that regulation and ethics are the most consistent drivers of AI overperformance across all income groups. Countries that perform better than expected are not only investing in technology, but also creating institutions, rules, and ethical frameworks that guide responsible AI adoption. This suggests that AI governance is not a luxury reserved for wealthy countries. Instead, it can be one of the most practical and affordable starting points for countries seeking to improve their AI readiness.
For low- and lower-middle-income countries, human capital is another major factor. Building digital skills, preparing the workforce, and aligning education with future technology needs are critical for countries that cannot immediately invest heavily in advanced infrastructure. This makes workforce development a central part of AI preparedness, especially for emerging economies that want to compete in a fast-changing digital environment.
The study also shows that different countries follow different coordination models. Some countries use a state-led model, where government institutions guide national AI strategy and coordinate implementation. Singapore is a strong example of this approach, while Rwanda shows how a similar model can be adapted in a lower-income setting through institutions such as its Responsible AI Office and Centre for the Fourth Industrial Revolution.
Other countries follow a market-responsive model, where the government creates enabling conditions while private sector actors support implementation. Malaysia and Kazakhstan reflect this approach by combining regulation, public infrastructure, and private sector participation. Malaysia’s workforce training levy system is especially important because it shows how governments can encourage private operators to contribute to skills development without relying only on public funding.
Large and complex countries such as India demonstrate a more distributed innovation model. In this approach, multiple ministries, state governments, private actors, and institutions contribute to AI development within broader national frameworks. This model allows experimentation across regions and sectors, but it also requires strong coordination to avoid fragmentation and ensure that national goals remain aligned.
The study makes clear that countries cannot simply copy each other’s policies. Singapore’s success depends on decades of trust between government and industry, while China’s infrastructure progress depends on centralized planning systems that may not exist elsewhere. The real lesson is not to imitate specific policies, but to understand the logic behind them and adapt that logic to local institutions, resources, and political realities.
For policymakers in developing economies, the most useful comparison may not be with global leaders such as Singapore or Switzerland. Instead, it may be more practical to look at structural peers that are performing better under similar constraints. A country with limited resources may learn more from Rwanda, Ghana, Morocco, Sri Lanka, or Viet Nam than from the world’s richest economies.
The evidence from these 24 overperforming countries shows that AI readiness is not fixed by income level alone. Governance, regulation, workforce development, and strategic coordination can help countries exceed expectations and reduce the AI preparedness gap. This is especially important for emerging markets and developing economies that want to participate meaningfully in the global AI economy.
The global AI divide is real, but it is not inevitable. Countries that build responsible governance systems, invest in skills, and coordinate effectively across government, business, and society can improve their readiness even when resources are limited. Rwanda’s experience shows that strong policy choices can create progress where economic indicators alone might suggest limited potential.
AI readiness is therefore not only a question of technology. It is a question of leadership, institutional planning, and policy commitment. The 24 overperforming countries identified in the study demonstrate that nations can shape their AI future by making deliberate choices today.
