Over the past year, markets have remained remarkably resilient despite persistent macroeconomic uncertainty. This resilience is underpinned by a combination of global fiscal support, a more accommodative monetary policy, and underlying strength in households and businesses. Deeper than that, however, is the AI investment cycle, one of the largest infrastructure buildouts in modern history that is increasingly contributing to economic growth.

Importantly, the AI buildout is unfolding amid heightened geopolitical tensions, persistent energy security concerns, and an increasing number of countries prioritizing national interests. The AI arms race has therefore evolved beyond a pursuit of tech superiority into one in which innovation, supply chain resilience, and military capabilities increasingly collide.

This has broadened the AI investment opportunity set from an initial handful of U.S. tech companies. Beneath the hyperscaler layer sits a much larger global ecosystem spanning data centers, power, semiconductors, hardware components, and natural resources. For investors, the opportunity now lies in identifying where those global layers are deepest, where bottlenecks are emerging, and which regions are best positioned to benefit as AI infrastructure scales.

The U.S. has been the epicenter of the AI buildout, supported by unmatched private investment and leadership in advanced semiconductor design, the foundation of AI systems. The early phase of the AI cycle largely rewarded U.S. hyperscalers, which initiated the investment wave by committing hundreds of billions of dollars to the infrastructure required for AI adoption.

As the AI cycle matures, however, the opportunity set is expanding to industries that provide the physical inputs for AI infrastructure, including semiconductors, electrical equipment suppliers, data center real estate, utilities, and power producers. Importantly, today’s AI data centers’ power requirements significantly exceed those of earlier periods, with equipment racks consuming 50-100 kW of electricity compared with 5-15 kW in traditional and cloud computing data centers. Growing power requirements, therefore, make access to reliable and abundant power another key determinant of AI leadership. 

However, AI-related electricity demand has outpaced available supply, creating a mismatch between the rate of AI adoption and the ability to bring new power online. While data center developers race to expand capacity to meet demand, projects are encountering delays. More than half of the data centers planned for 2027 completion have not yet started, while lengthy permitting and grid connection timelines, labor shortages, and constrained supplies of critical equipment, such as natural gas turbines, further add to the uncertainty.

These constraints should keep demand durable across energy, battery storage, cooling solutions, and the critical equipment needed to expand and modernize the power grid. For investors, the opportunity is likely to center on companies whose products help address the infrastructure constraints that are now shaping the pace of AI capacity growth. However, as the AI cycle enters a more mature phase, the run-up in AI-related valuations suggests that further gains will increasingly depend on earnings delivery rather than valuation expansion.

Taiwan and South Korea: beneficiaries of hyperscaler spending

Although the U.S. has established itself as the world’s leader in AI development, much of the hardware that is enabling the buildout is produced in Asia. The semiconductor supply chain is globally interconnected, with U.S. firms leading chip design while Taiwan and South Korea dominate advanced chip manufacturing. The surge in AI-related U.S. hyperscaler spending has generated positive spillover effects across these regions’ tech sectors, boosting earnings and equity market performance. With hyperscaler capex expected to remain robust and surpass $1 trillion in 2027, demand for advanced semiconductors is likely to remain an earnings tailwind for key Asian suppliers. 

China: beneficiary of domestic self-sufficiency

In China, reliance on the U.S. and its allies for advanced tech components has long been viewed as a vulnerability. To address this challenge, Chinese policymakers have spent years promoting domestic innovation to strengthen their tech capabilities. More recently, geopolitical tensions and U.S. export controls on advanced semiconductors have accelerated Beijing’s efforts to establish a more self-sufficient tech ecosystem.

Increasingly, Chinese tech firms are adopting domestically designed hardware, and reports suggest that China has made notable progress in developing domestic extreme ultraviolet (EUV) lithography, a system recognized as one of the most complex required to manufacture the most advanced semiconductors. Although domestic systems remain meaningfully behind the most advanced EUV technology and are not yet ready for large-scale commercial deployment, these efforts underscore China's determination to build a more self-sufficient semiconductor ecosystem. This will likely reinforce demand across China’s entire semiconductor value chain—spanning chip design, equipment, memory, processors, and IC tooling.

For investors, Asia offers two distinct AI exposures: globally integrated suppliers in Taiwan and South Korea that remain closely tied to U.S. hyperscaler capex, and China’s domestic substitution cycle, where policy support and supply-chain resilience are driving demand across local semiconductor and AI infrastructure firms.

Throughout much of the AI cycle, Europe has been viewed as a diversifier relative to more tech-heavy markets, such as the U.S., Taiwan, and Korea. The MSCI Europe Index’s relatively low tech weight, however, may understate the region’s critical role as an AI enabler.

At the center of the global AI ecosystem is ASML—the Netherlands-based semiconductor production equipment company and Europe’s largest listed firm—which remains the world’s dominant supplier of EUV lithography machines. While the firm is positioned to benefit from rising demand for increasingly advanced chips, the opportunity set extends well beyond a single company. 

Each EUV machine contains hundreds of thousands of components and relies on an extensive network of global suppliers for production, maintenance, and replacement parts, with roughly a third of suppliers located in Europe, particularly in Germany, France, the UK, and the Netherlands. Rising demand for advanced semis, therefore, generates positive spillovers for European component suppliers. 

Moreover, Europe’s AI opportunity set extends beyond its role in the semiconductor supply chain. U.S. hyperscaler spending is increasingly driving demand for industrial equipment that powers data centers, positioning leading European original equipment manufacturers (OEMs) as key beneficiaries. Beyond AI, European industrial companies are also supported by a pivotal shift in fiscal policy. After years of austerity, increased spending on defense, infrastructure, and energy security is creating broader tailwinds for industrial companies. Europe’s growing emphasis on domestic suppliers, particularly in defense systems, may further reinforce these trends. 

Taken together, Europe has a clearer role in the AI buildout than its benchmark tech weight might suggest. Its opportunity sits in the industrial backbone of AI: semiconductor equipment, precision components, and the data center inputs needed to support global capacity growth.

The AI supply chain ultimately depends on raw materials, including the metals and minerals used in semiconductors, data centers, electrification, and defense systems. China’s dominant role in several of these inputs, including rare earths, remains a global chokepoint, especially as governments look to build more resilient supply chains with trusted trading partners to support their AI, defense, and electrification needs.

That push supports resource-rich economies with deep reserves and established mining capacity, including Canada and Australia. It also brings Latin America further into focus, given the region’s large share of global reserves across Central and South America, including 45% of the world’s lithium, 30% of silver and copper, 23% of rare earths. While the opportunity is meaningful, accessing it is not particularly straightforward for investors. Mining and refining projects across parts of Latin America have historically faced permitting delays, regulatory hurdles, social opposition, and political volatility. And while recent policy shifts toward more pro-mining agendas could improve the outlook, investors will need to be selective, with country-level policy, project execution, and supply-chain reliability likely to matter as much as resource ownership itself.

As AI becomes a competitive arena among major economies, the next phase of equity market returns may become more geographically diverse than the first. Gains are likely to accrue not only to U.S. tech leaders, but also to the regions and industries enabling the AI buildout: Asian semiconductor manufacturers, European industrial suppliers, power and grid infrastructure firms, and resource-rich economies supplying the materials required to scale AI.

In an uncertain macro landscape, diversification remains a first line of defense against volatility, but it is also a way to capture selective exposure to the global infrastructure layers underpinning one of today’s most powerful secular growth trends.