Is AI making diversification harder?

Traditional diversification routes are increasingly exposed to the same drivers that dominate global equity markets. Infrastructure may be one of the few remaining sources of genuinely differentiated equity exposure. Here’s how.

The diversification mirage

Most investors would probably describe their portfolios as diversified. They own global equities, emerging markets and perhaps small caps. On paper, that appears to provide exposure to different regions, sectors and return opportunities. Yet beneath the surface, many of these allocations are increasingly influenced by the same underlying driver: AI.

This matters because diversification has traditionally been about spreading risk across different assets and sources of return. But what if seemingly different investments are being driven by the same theme?

The hidden concentration problem

Technology has been the defining driver of global equity returns since the Global Financial Crisis. The rise of AI has accelerated that trend, increasing the influence of a single theme across multiple asset classes.

Successive innovation cycles have created extraordinary wealth as technology's influence on global markets has expanded. The sector's weight within MSCI ACWI Index has risen from around 10% to 31%, while technology now contributes roughly half of overall index risk.

Importantly, the challenge is not technology itself. Nor is it AI. The challenge is the increasing dependence of global equity portfolios on a single return driver. When a growing share of returns is influenced by the same theme, diversification becomes harder to achieve, even if portfolios appear diversified on the surface.

The rise of AI has amplified this trend. While technology contains many different industries and business models, AI has become one of the dominant drivers of market performance in recent years. 

Are investors less diversified than they think?

We examined the relationship between AI and returns across a range of asset classes. Using a popular AI exchange-traded fund (ETF) as a proxy for the theme, we found high correlations across several major equity markets. Global equities exhibited a correlation of 0.79, emerging markets 0.69 and global small caps 0.75. These figures suggest that AI has become an increasingly important influence across multiple parts of a typical equity portfolio [1].

So, the question is: if global equities, emerging markets and small caps are all increasingly influenced by the same theme, are portfolios as diversified as investors believe? Traditional diversification assets such as cash, bonds and listed property have exhibited much lower correlations to AI. However, investors have historically faced a trade-off when using these assets, often sacrificing return potential in exchange for diversification benefits. 

A different way to access the AI build-out

Enter listed infrastructure. It provides exposure to many of the same structural trends underpinning AI adoption, but through a different part of the value chain. Broader equity markets are often driven by developments in AI technology and investor sentiment. Infrastructure companies, by contrast, are typically supported by long-term demand for the essential assets that make digital growth possible.

In other words, infrastructure can benefit from the AI build-out without relying on the same return drivers that dominate wider equity markets.

One example is NextEra Energy. This US-based company is emerging as a key infrastructure partner to the AI build-out, supported by a large pipeline of electricity demand linked to hyperscalers and data centres. As AI adoption increases, demand for power, networks and other enabling infrastructure is growing alongside it.

This highlights an important distinction. Investors do not necessarily need to avoid AI to achieve diversification. Instead, they may need to consider where within the AI ecosystem they are allocating capital.

Final thoughts…

Diversification should mean owning asset that behave differently, not simply owning more of the same.

As technology and AI become increasingly important drivers of market returns, investors may need to think more carefully about the degree to which different parts of their portfolios are exposed to the same underlying theme.

The challenge is not finding more investments but finding genuinely different sources of return. Infrastructure offers exposure to many of the same long-term growth trends reshaping the global economy, but through a different set of economic drivers. For investors seeking diversification without abandoning attractive long-term return potential, that distinction may become increasingly valuable.

[1] Source: Morningstar Direct, two-year data to end of August 2026. Annualised returns in $ and correlation to Morningstar Global AI Select NR USD. Global Equities: MSCI ACWI, Global Small Cap: MSCI ACWI Small Cap, EM Equities: MSCI EM, Cash: Secured Overnight Financing Rate, Global Bonds: Bloomberg Global Aggregate, Listed Property: FTSE EPRA Nareit Developed, Aberdeen Global Infrastructure Strategy

Graham McCraw is senior investment specialist, equities at Aberdeen.