Where global economies sit in the AI stack

Source: Artificial Intelligence Index Report 2025, in USD.

The Top 10 government investment commitments highlight that AI leadership is geographically diverse. The United Kingdom, Germany and India each reflect distinct strategic approaches—from public services integration and industrial automation to digital public infrastructure and computing expansion—while China’s cumulative commitment underscores the scale of its long-term industrial AI ambitions. We will examine these ecosystems, including the United Kingdom’s regulatory-first framework, Germany’s manufacturing integration and India’s digital public stack in greater detail in an upcoming commentary. Here, we focus more directly on where the AI stack intersects with equity market exposure, particularly across Asia and energy-enabled infrastructure markets.

Taiwan sits at a critical chokepoint in the AI compute stack, producing roughly 90% of the world’s most advanced logic chips. The concentration is often described as Taiwan’s “silicon shield” for its strategic global importance.

While Taiwanese manufacturers have expanded production footprints in the United States and other regions, replicating the full level of sophistication found in Taiwan’s semiconductor ecosystem is expected to take many years. Advanced chip manufacturing depends not only on fabrication facilities, but also on dense networks of specialized suppliers, experienced engineering talent, advanced packaging capabilities, and rapid production learning cycles. This ecosystem depth remains difficult to reproduce quickly, reinforcing Taiwan’s competitive advantage as AI-driven demand for high-performance computing accelerates.

South Korea’s memory-chip segment plays a central role in the AI hardware stack, with global leadership in dynamic random-access memory, including high-bandwidth memory, and a top position in NAND flash storage. As AI spending increasingly shifts toward graphics processing units and accelerator-based servers, demand tilts toward higher-value memory configurations, and this shift toward more memory-intensive AI workloads should benefit South Korea’s tech sector. But South Korea also illustrates how AI advantage does not require leading the race to build the largest global models. Instead, domestic firms are focusing on localized and industry-specific AI systems, particularly for non-English languages and regionally tailored enterprise use cases. In late 2025, the country climbed seven places in global AI diffusion rankings, driven by increased workplace integration and digital skills uptake. A new report from the AI Economy Institute identifies South Korea’s two largest sectors, technology and industrials, as embedding AI tools at a faster pace than many global peers.³

South Korea’s AI opportunity is increasingly visible in heavy industry, not just software. In late 2025, a major Korean industrial group touted that AI and big-data systems are being integrated across shipyard design, planning, and production workflows, with the goal of creating fully connected “smart yards” capable of predicting and optimizing output. The country’s major shipbuilder already deploys digital twin replicas of physical shipbuilding facilities and is expanding these systems into end-to-end, AI-enabled production platforms. By 2030, they aim to operate an intelligent autonomous shipyard expected to boost productivity by roughly 30% and reduce construction timelines by a similar magnitude. These initiatives are designed not only to address expected labor constraints in shipbuilding, but also to strengthen Seoul’s competitive position relative to Beijing. For investors, South Korea’s position in the AI stack is defined by deployment over dominance.

Japan: industrial AI and productivity capture

Japan’s position in the AI stack is defined less by capital intensity and more by industrial deployment, execution and policy alignment. That backdrop has become more constructive following Prime Minister Sanae Takaichi’s sweeping electoral victory, which markets have interpreted as a strong mandate to push through structural reforms and long-term growth initiatives. Gauging by strong year-to-date net inflows into Japan-focused exchange-traded funds, early investor response has reflected optimism that clearer leadership and policy continuity could accelerate implementation in areas such as AI, advanced manufacturing and semiconductors.

While Japan does not rank among the largest government spenders on AI in absolute terms, it consistently places among the world’s top AI economies in Stanford’s AI Vibrancy rankings, reflecting strength across research, infrastructure and private-sector adoption. Under Takaichi’s leadership, the government approved Japan’s first national AI plan, committing roughly US$6.3 billion over five years, to strengthen foundational AI capabilities, robotics integration and industrial deployment. The emphasis is not on chasing frontier models, but on embedding AI into sectors that already anchor Japan’s economy.

Importantly, Japan is also pursuing a strategy to recapture semiconductor leadership across critical parts of the value chain. Government-backed efforts, such as Rapidus, are targeting 2-nanometer production later this decade, while Japan retains world-class strength in semiconductor materials, wafers, test equipment and precision manufacturing inputs that are essential to advanced chip fabrication globally. This ecosystem depth reinforces Japan’s strategic relevance within AI hardware supply chains. This positions the nation not only as a manufacturing location, but as a foundational supplier of the tools and materials every advanced fab depends on.

This matters from an equity-market perspective. Japan’s market is heavily geared toward industrials (roughly 28%), financials (about 16%) and consumer discretionary firms (around 14%), with technology close behind (near 14%).⁴ This tilt reflects an economy built on manufacturing depth and capital discipline. Few countries are as comfortable integrating robotics into daily life and industrial production, and that long familiarity with automation shapes how AI is deployed.

Brazil and Saudi Arabia: Energy abundance as an AI infrastructure advantage

Brazil’s AI exposure extends beyond commodities. As we recently noted in our Brazil commentary, the country’s energy mix is among the cleanest globally, with renewables accounting for roughly 90% of electricity generation. This clean energy advantage is increasingly relevant for enterprise-scale AI infrastructure, where power cost and stability directly affect data center economics.

Brazil has also emerged as a leading Latin American AI hub. With more than 7,000 kilometers of coastline, the country has direct access to major submarine cable systems linking the Americas, Europe and Africa—an increasingly important structural advantage as data intensity and cross-border digital traffic expand.

For investors, Brazil’s role in the AI stack is less about frontier research and more about infrastructure conditions: clean power, connectivity, and a growing domestic digital ecosystem that supports regional AI deployment.

Saudi Arabia is pursuing one of the most ambitious state-led AI infrastructure strategies outside the United States and China. Under Vision 2030, Riyadh is leveraging abundant low-cost energy, capital and land to position itself as a regional hub for AI infrastructure and computing capacity.

In 2025 alone, the government announced nearly US$15 billion in AI and digital transformation investments, with significant capital directed toward data centers, cloud buildout and AI commercialization.

For investors, Saudi Arabia’s position in the AI stack lies in its ambition to become a supplier of enabling infrastructure—power, data center capacity and digital connectivity—rather than solely an end user of AI systems.

More broadly, AI should be viewed as a geographically dispersed, layered buildout across semiconductors, power, automation and digital infrastructure. As spending shifts from model development to deployment, the set of beneficiaries broadens—and so does the corresponding investment map.