BlackRock: Why AI Investment will Evolve in 2026

Artificial intelligence remains central to investment portfolios, but the momentum that drove tech giants throughout 2025 is beginning to spread across a wider ecosystem.

Financial institutions are beginning to rethink where AI growth offers the best return – and the spotlight is now falling on energy providers, grid infrastructure and long-term power planning.

The shift in strategy is outlined in BlackRock’s latest Investment Directions report, which involved interviewing more than 700 of its EMEA clients.

Investors cool on big tech in 2026

AI captured a flood of investment in 2025 as companies like Meta, Microsoft, Google and OpenAI ramped up development of language models (LLMs), cloud infrastructure and foundation model platforms.

But while demand for AI systems continues, investment confidence is beginning to rebalance.

"It's increasingly important to risk-manage megacap and AI exposure while also capturing differentiated upside opportunities," says Ibrahim Kanan, BlackRock's Head of Core US Equity.

Just a fifth of those polled by BlackRock say large US technology firms remain a compelling investment for the coming year. By contrast, 54% of respondents consider energy providers to be a stronger opportunity and 37% say energy infrastructure offers a better outlook than the big tech sector.

The data suggests investors are reacting to how AI systems are being built and scaled, with the cost of developing infrastructure remaining high. Consulting leader Gartner estimates that capital expenditure on AI hit US$1.5tn in 2025 – a worrying figure for investors concerned about concentration risk and uncertain returns, particularly when higher interest rates make borrowing more expensive.

There is also growing scrutiny of data centre profitability. Power bills keep rising and with that comes further pressure to identify sustainable, long-term value. BlackRock’s clients are looking beyond platforms and model providers to the utilities and infrastructure firms that supply the backbone of AI deployment.

Energy demand takes centre stage

As AI systems scale up, they bring increased attention to power consumption. Data centres operate around the clock and require continuous cooling and server performance, making electricity providers and grid operators key players in the AI investment story.

BlackRock asserts that this is far from a passing trend. Its research suggests AI has become a “structural component of global power consumption” rather than a short-term spike.

Energy providers are, therefore, in a favourable position. These firms not only provide a path to reduce the carbon footprint of AI growth but also offer a hedge against price volatility and long-term emissions risk.

Large technology companies and infrastructure developers are signing clean energy deals to secure future supplies and to manage emissions targets. These agreements stretch years ahead and underline how energy access becomes a core part of AI planning.

Investors see the same opportunity. As data centre demand expands, it puts pressure on national grids, transmission systems and storage capacity. Building new capacity now takes precedence over backing another large-language model.

Infrastructure builds underpin AI expansion

The focus for 2026 has already shifted to physical infrastructure.

Without investment in the grid, AI’s growth risks hitting a ceiling. Energy storage systems, transmission upgrades and modernised distribution networks are seen as vital to maintain expansion. In many areas, delays in permitting and planning slow the rollout of data centre infrastructure.

BlackRock, Microsoft and NVIDIA have jointly committed to a US$100bn programme to boost both AI data centres and surrounding energy networks, indicating industry recognition of the need to fix bottlenecks before power supply becomes a limiting factor.

However, this phase also brings short-term costs. Building new data centres and transmission networks involves high-emissions materials including steel and cement, as well as energy-intensive manufacturing. That raises further questions about the timing gap between upfront investment and eventual productivity gains.

Despite the concerns, interest in AI remains high. Just 7% of BlackRock’s respondents believe AI is a bubble, indicating that investors continue to believe in its long-term promise.

What's changing is how that promise is funded. Whereas early-stage AI development attracted money towards platform builders and model designers, 2026 marks a pivot towards companies that supply and sustain the ecosystem AI requires to function at scale. Grid stability, clean energy and private infrastructure projects are attracting the capital that once went into software.

Private markets are expected to take on a bigger role, funding efficiency upgrades and grid improvements that fall outside traditional equity investments. 

Ultimately, BlackRock’s research indicates a shift in understanding, rather than a loss of faith in AI. Investors are approaching it as a wide-reaching investment theme that extends to the machines, cables and kilowatts powering digital transformation.