National Wealth Fund: AI boom boosts electricity demand and reshapes investment opportunities

A report by the Thought Leadership division of National Wealth titled “Investing in the Energy Needs of Artificial Intelligence” states that over the past few years, the narrative surrounding AI has centered on software and chips: models capable of writing, coding, and logical reasoning, along with the processors trained to run them. During this phase, electricity was treated as a secondary operational input, receiving little attention from investors.

However, the rapid expansion of AI applications has altered this equation. The primary challenge is no longer limited to chip capabilities but now lies in securing adequate energy supplies and providing the physical infrastructure necessary to support large-scale operations. The report notes that the scale of investment is significant enough to directly impact investment portfolios, with estimates from the International Energy Agency (IEA) for data center capacity indicating cumulative investments of approximately $3.9 trillion by 2030.

These facilities can now consume amounts of electricity comparable to those of heavy industry, and their appetite for power is growing at a rate far exceeding the pace of expansion of the electrical grids that supply them. For long-term investors, this is no longer merely a technology story; it has become an infrastructure and energy story, carrying its own distinct opportunities and risks.

Actual Demand

The report adds that data centers consumed an estimated 485 terawatt-hours (TWh) in 2025, representing approximately 1.5% of global electricity consumption, according to the IEA. More important than the current share, however, is the speed of growth. Key IEA projections indicate that consumption will more than double, reaching approximately 945 TWh by 2030—slightly more than Japan’s current total consumption—and rising further to 1,200 TWh by 2035.

The United States accounts for nearly 45% of global data center electricity consumption, followed by China at approximately 25% and Europe at around 15%, according to the IEA.

While estimates vary regarding the level of data center electricity consumption in the United States by 2030, they all agree on a strong upward trend. Estimates from the Lawrence Berkeley National Laboratory suggest that data centers could represent approximately 11.8% of total electricity demand (within a range of 9.5% to 15.3%). The Electric Power Research Institute (EPRI) estimates the share to be between 9% and 17%, while McKinsey projects it to reach between 11% and 12%.

The irony is that these estimates start from a current level of just 3% to 5%, reflecting the sector’s exceptionally rapid growth. After data center electricity consumption represented a marginal share of national demand just a few years ago, it is now projected to account for roughly one-eighth of total U.S. electricity demand in the coming years.

Although the range of forecasts is wide and the final outcome depends on a set of factors, the base and most likely scenario points to consumption increasing several-fold over the next four to five years.

Physical Infrastructure

This demand is transforming the industry from a virtual, intangible field into a purely physical one. Operating artificial intelligence at scale requires power generation, transmission lines, transformers, cooling systems, water, and land, broadening the scope of investment beyond a narrow set of technology companies.

On the power generation side, no single fuel source dominates the landscape. The International Energy Agency (IEA) expects renewable energy to meet about half of the growth in data center demand, adding more than 450 terawatt-hours by 2035, supported by shorter construction times and lower costs.

Natural gas is expected to cover most of the remaining share, particularly in the United States, where it is bolstered by supportive policies and the availability of cheap supplies. Nuclear power will contribute over a longer horizon, as the sector revives interest in Small Modular Reactors (SMRs)—compact units built in stages that are expected to begin appearing by 2030, according to the IEA. The beneficiaries are spread across utilities, independent power producers, gas and nuclear supply chains, and equipment manufacturers.

Electricity grids have also become a major bottleneck to expansion, benefiting those able to provide effective solutions to these challenges. Developing new transmission lines in advanced economies takes four to eight years, and wait times for transformers and cables have doubled over the past three years. Equipment that was once a minor part of industrial manufacturing has suddenly become scarce, while utilities—long considered defensive assets with slow growth—are being reassessed as demand growth returns for the first time in decades.

Artificial intelligence can also help alleviate some of the pressure it creates. The IEA estimates that AI-assisted grid management could unlock up to 175 gigawatts of transmission capacity without the need to build new lines.

The report notes that the scale of spending is reshaping how these projects are financed. The estimated $3.9 trillion over the next five years is too large to be covered by corporate budgets alone. This gap between ambition and internal cash flows is attracting external capital, bringing the AI theme into the private markets and making it more significant for investment portfolios than any demand forecast.

The average annual total debt issuance by the top five technology hyperscalers was approximately $35 billion between 2020 and 2024, before surging to $93 billion in 2025 and reaching about $132 billion during the first seven months of 2026, according to Vanguard data. Broader estimates that include chipmakers, developers, and utility companies are several times higher. Companies that enjoyed structural cash surpluses just a few years ago are now classified among the largest borrowers in the market.

New financing structures are emerging alongside traditional debt. In August 2026, Nvidia announced partnerships with several global asset managers to launch platforms aimed at attracting more than $500 billion in investor capital, providing cloud operators and institutions with dedicated funding for Nvidia-based systems.

These arrangements treat computing power as productive infrastructure capable of generating long-term revenues tied to utilization rates. Although implementation is still in its early stages, it indicates that the expansion is attracting capital traditionally associated with the energy, transportation, and real estate sectors. For diversified investment portfolios, the most significant and direct exposure to this sector may lie in infrastructure, private credit, and physical assets, rather than being limited to publicly listed technology stocks.