Institutional investors are no longer bystanders in the artificial intelligence boom; they are helping to drive it. Their vast pools of capital make them natural financiers of AI’s most expensive assets, from hyper-scale data centres to training systems. But these investments carry distinct risks: high upfront costs, rapid obsolescence and growing geopolitical strain across supply chains.
Those risks raise a broader question. How durable is the current AI investment cycle? AI’s boom resembles past waves of industrial investment: vast, transformative, but prone to overreach. Institutional capital is not only funding the build-out of AI infrastructure; it may also magnify the next technology downturn.
This report examines how institutional investors are navigating that tension. Drawing on a global survey of 300 institutional investors, it explores what is driving allocations, how investors would respond to a sharp correction and how portfolios are positioned to capture AI’s longer-term gains.
Key findings of the research:
The AI investment boom is fragile, driven in part by market momentum rather than economic fundamentals.
such as recent stockmarket performance or highly concentrated returns among a handful of dominant firms. In other words, rising prices are helping to justify further investment. Yet the longer-term case is more grounded: about half cite long-term productivity gains as the main motivation, while only 14% expect short-to-medium-term cost savings. The boom rests on a mix of conviction and momentum—an unstable combination.
Investors expect a sharp correction in AI equities—but may not be able to absorb losses of that scale.
Yet fewer than 1% say they could withstand a drawdown of that magnitude. Despite this, more than 80% express confidence in their portfolios, though many qualify this as only “somewhat confident”. This mismatch suggests resilience may be overstated.
Portfolios are already heavily exposed to AI—mostly through equities—creating meaningful concentration risk.
This is higher than that for fixed income or alternative investments. Equity exposure runs across the ecosystem, from infrastructure providers to developers to adopters. Holdings are most concentrated in infrastructure and passive funds, suggesting a preference for indirect bets. Even so, roughly one-third of investors report that AI-linked equities are overweight relative to their strategic allocations, leaving portfolios exposed to swings in AI valuations. Heavy equity exposure across ecosystem actors raises the risk of false diversification: if valuations fall together, diversification may offer less protection than expected.
Institutional investors are preparing for volatility, but many plan to buy rather than retreat if valuations fall.
Governance measures—such as stricter investment-committee oversight and predefined rebalancing triggers—are the most widely adopted strategies. Few plan to reduce exposure. Instead, most expect to increase allocations if valuations fall. Investors are positioning themselves as opportunistic buyers in a downturn, not defensive sellers.
Investors expect AI to generate returns across many sectors, though technology and finance remain the primary beneficiaries.
Technology is widely seen as the sector most likely to deliver the largest AI-driven investment gains. Financial services also ranks highly, as investors anticipate efficiency gains and new AI-enabled financial products. Some regional investors, particularly in Asia-Pacific, expect stronger gains in sectors such as media and logistics. Even so, expectations remain anchored in sectors already closely tied to the technology.
Investors are betting heavily on the US to win the race for AI supremacy.
China trails significantly at 13%, suggesting investors may be underestimating its role. Overall, expectations are skewed towards developed markets: 27% expect the highest returns there, compared with just 13% for emerging markets.
About this research
Exuberance and exposure: Institutional investors and the AI boom is a report by Economist Enterprise, supported by Xtrackers by DWS. The report examines how institutional investors are shaping—and potentially amplifying—the AI investment boom and balancing long-term conviction with short-term fragility as they prepare for a likely market correction.
The report is based on a global survey of 300 institutional investors from North America, Europe and Asia—including pension funds, insurers, sovereign wealth funds, endowments, family offices and government agencies—conducted in February and March 2026.
AI risk mitigation: Beyond active versus passive
With the “Magnificent Seven” making up over a third of the S&P500, heightened concentration is certainly giving cause for thought to investors worried about an AI bubble. How can investors understand their AI exposure and mitigate the risks to their portfolios?
