A growing number of US investors are rotating away from high profile artificial intelligence giants and into infrastructure companies positioned to benefit from long term AI capital spending, as volatility hits major technology stocks.
After years of strong gains, shares of leading AI driven firms such as Alphabet and Amazon have come under pressure. Investors are increasingly questioning whether massive investments in advanced AI systems will generate returns sufficient to justify elevated valuations.
Instead of betting directly on consumer facing AI platforms, asset managers are targeting the companies supplying the physical backbone of the AI expansion. This includes semiconductor manufacturers, data centre developers, optical communications firms and utilities responsible for delivering the electricity required to power large scale computing clusters.
Market performance has reinforced this pivot. Infrastructure linked names including Caterpillar, Lumentum and Western Digital have posted double digit gains this year. contrast, the S&P 500 has delivered only marginal returns, while exchange traded funds tracking the largest AI hyperscalers have declined.
The shift has prompted a wave of new financial products focused specifically on AI infrastructure exposure. Asset managers including BlackRock have adjusted portfolio weightings to increase allocations toward chipmakers and other enabling technologies. In several actively managed funds, infrastructure related holdings now account for a significant majority of total assets.
The logic behind the strategy is straightforward. As companies like Meta and Amazon commit billions of dollars to new data centre builds, that capital flows through supply chains spanning memory chips, networking equipment, cooling systems and energy providers. Investors seeking more predictable revenue streams are aiming to capture returns from these underlying investments rather than from consumer AI applications, which face greater competitive and regulatory uncertainty.
Semiconductor producers remain central to the thesis. Firms such as SK Hynix, Micron and Intel are viewed as critical beneficiaries of sustained demand for high performance memory and processing capacity required to train and operate advanced AI models. While Nvidia remains influential within the ecosystem, some infrastructure focused funds have diversified exposure to reduce concentration risk.
Beyond chips, attention is also turning to utilities and grid operators as AI driven data centres increase electricity consumption. Reliable power supply and transmission capacity are emerging as key constraints on expansion, creating potential opportunities in energy infrastructure.
The reallocation reflects broader market dynamics. With headline AI stocks facing scrutiny over earnings visibility, investors appear to be seeking steadier cash flow profiles tied to tangible assets. Rather than abandoning the AI theme, many are repositioning within it, emphasising the industrial and infrastructural foundations supporting the sector’s growth.
