Investors Favour Energy and Infrastructure Over Big Tech for AI Exposure in 2026, BlackRock Says

BlackRock remains constructive on the long-term investment case for artificial intelligence, but says investors heading into 2026 are increasingly looking beyond U.S. megacap technology stocks and rotating into the physical backbone of the AI economy — namely power and infrastructure.

In its latest Investment Directions report released on January 13, 2026, the world’s largest asset manager said the AI theme is evolving from a race to build chips into what it described as a “race for megawatts,” as data centers place rising demands on electricity supply, grid capacity and industrial infrastructure.

A pivot away from megacaps — not away from AI

BlackRock said AI-linked U.S. “hyperscalers” helped drive equity returns in 2025, but investor attention is widening as the cycle matures and capital requirements grow.

The firm noted that planned spending across the AI buildout could reach $5 trillion to $8 trillion through 2030, a figure that has reinforced investor concerns around capital intensity and uncertain near-term returns.

“It’s increasingly important to risk-manage megacap and AI exposure while also capturing differentiated upside opportunities,” said Ibrahim Kanan, BlackRock’s head of core U.S. equity, in the report.

Survey: energy and grid themes now dominate AI positioning

BlackRock said its report included findings from a survey of 732 EMEA-based client companies, showing a visible shift in preferred AI exposure.

According to the survey:

• More than 50% now see energy providers supplying data centers as the most compelling AI opportunity.
• 37% prefer infrastructure exposure, including grid upgrades, copper-wire demand, and data-center cooling systems.
• Only 20% named the largest U.S. tech firms as the most attractive AI play.

The message: investors still want AI upside — but they increasingly want it through what must be built, not only through software narratives.

“Race for megawatts”: electricity becomes the scarce asset

BlackRock argued that a key constraint in the next stage of AI expansion will be power availability.

AI data centers are far more energy-intensive than conventional cloud infrastructure, and the firm highlighted projections suggesting data centers could consume a significantly larger share of U.S. electricity demand by 2030, depending on deployment pace and grid buildout.

That framing supports a broader market trend: AI is no longer only about models and chips. It is increasingly about utilities, power markets, copper, cooling, and construction-scale capex.

Bubble fears remain limited — for now

Despite elevated valuations in parts of the AI complex, BlackRock said confidence in the long-term theme remains strong.

Only 7% of respondents in the survey described the AI theme as a bubble, indicating what BlackRock characterised as a rotation and broadening of opportunity, rather than a collapse in conviction.

Summary

BlackRock says investors seeking AI exposure for 2026 are shifting focus away from U.S. megacap tech and toward energy and infrastructure providers supporting data-center expansion. In a survey of 732 EMEA-based clients, more than half favored power providers as the best AI opportunity, while 37% preferred infrastructure companies. BlackRock said the next phase of AI investment is increasingly driven by physical constraints — electricity, grid upgrades and data-center buildouts.

Frequently asked questions (FAQs) 

Q1: What is BlackRock’s main message on AI investing in 2026?

BlackRock says AI remains a durable theme, but investors are increasingly broadening exposure beyond megacap tech into energy and infrastructure linked to data centers.

Q2: Why are energy stocks gaining attention as “AI plays”?

BlackRock argues electricity supply is emerging as a core constraint, as AI infrastructure expansion requires large amounts of reliable power and grid capacity.

Q3: What survey data is BlackRock citing?

BlackRock cites a survey of 732 EMEA-based client companies, where more than half preferred energy providers and 37% preferred infrastructure, while 20% favored large U.S. tech firms.

Q4: How much spending does BlackRock expect in the AI buildout?

BlackRock estimates AI buildout capital spending intentions could reach $5 trillion to $8 trillion through 2030, reflecting data centers, compute, and supporting infrastructure.

Q5: Does BlackRock think the AI market is in a bubble?

Not broadly. BlackRock notes that only 7% of surveyed respondents consider AI to be a bubble, suggesting most investors still view the trend as structurally supported.