The artificial intelligence boom that once dominated global investment strategies may be entering a new phase. According to new research from BlackRock, investors are reallocating capital away from major technology companies and toward the energy infrastructure required to power AI-driven data centres.

BlackRock’s latest Investment Directions report reveals that energy providers and grid operators are now being prioritised over the big technology firms that previously captured most investor attention.

The US-based asset manager surveyed 732 Europe, Middle East, and Africa (EMEA) clients and found that only 20% consider large American technology companies to be compelling investment opportunities today.

More than half of respondents identified data centre energy as an attractive investment, while 37% said they preferred energy infrastructure over big tech as an opportunity.

This shift reflects growing concern over the capital-intensive nature of AI development. Gartner estimates that approximately US$1.5 trillion has recently been invested in AI, much of it concentrated among a small group of companies.

Rising borrowing costs have further increased scrutiny around the uncertain returns tied to the massive capital expenditures required to build AI infrastructure.

“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 US Equity.

Power Consumption Becomes a Structural Concern

The energy demands of AI systems have become a central consideration for investors. Data centres operate around the clock, consuming enormous amounts of electricity to run servers and cooling systems.

As a result, energy providers and grid operators are transitioning from supporting roles to critical enablers within the AI investment ecosystem.

BlackRock expects AI-driven electricity demand to become a permanent feature of global power consumption rather than a temporary surge.

Renewable energy providers may be particularly well positioned as data centre operators seek to reduce carbon footprints while securing stable, long-term energy supplies. Many companies are now entering long-term clean energy contracts to manage both price volatility and emissions risk.

Infrastructure Upgrades Become an Investment Priority

Beyond power generation, physical infrastructure is emerging as a critical investment focus. Grid upgrades, transmission networks, and energy storage systems are essential to support sustained AI growth.

BlackRock, Microsoft, and NVIDIA have announced plans to invest US$100 billion in AI data centres and power infrastructure. However, regulatory and permitting challenges for new facilities and grid connections are slowing progress in several regions.

Without accelerated expansion of transmission capacity, rising AI demand could strain electricity supplies and delay deployment timelines.

While technology companies such as NVIDIA and Microsoft remain part of BlackRock’s investment portfolios, the survey suggests investors are increasingly focused on sustainable returns and risk mitigation rather than exposure to heavy capital spending.

The construction phase of AI infrastructure also introduces near-term emissions from steel, cement, and energy-intensive manufacturing, creating a timing gap between investment and productivity gains.

Long-Term Confidence in AI Remains Strong

Despite the shift in capital allocation, confidence in AI’s long-term potential remains intact. Only 7% of survey respondents believe the recent rise of AI represents a market bubble.

BlackRock’s findings suggest portfolio diversification is driving investment toward energy and infrastructure assets. Private markets are expected to play a growing role in funding grid upgrades and efficiency improvements that fall outside public equities.

As AI investment matures, the focus is expanding beyond software and model developers to the broader ecosystem that enables digital growth.

In the next phase of AI investing, the infrastructure that powers intelligence may prove just as valuable as the intelligence itself.