For more than a decade, the energy investment narrative was dominated by one question: how quickly could the world replace fossil fuels with cleaner generation? That question is changing. Electricity demand is accelerating as artificial intelligence, data centers, advanced manufacturing, electric vehicles and broader electrification converge on the same infrastructure. The International Energy Agency expects global electricity consumption to increase from 28,200 TWh in 2025 to 33,600 TWh by 2030, adding roughly 1,100 TWh of new demand every year—50% more annual incremental demand than during the previous decade.

Capital is already moving accordingly. Global electricity-sector investment is expected to reach approximately $1.5 trillion, around 50% more than investment in bringing oil, natural gas and coal to market. Ten years ago, that relationship was reversed.

But the investment cycle is becoming uneven. Roughly $1 trillion is now invested annually in electricity generation, while only around $400 billion goes into grids. More than 2,500 GW of generation, storage and large-load projects are consequently waiting in grid-connection queues around the world.

AI makes that mismatch harder to ignore. Global data-center electricity consumption reached approximately 485 TWh in 2025 and is expected to roughly double to 950 TWh by 2030. Data centers can also be developed substantially faster than the power infrastructure required to serve them. The result is a new form of scarcity. The most valuable asset in the next energy investment cycle may not simply be the lowest-cost source of electricity. Increasingly, it could be reliable power that is available in the right place, with an existing grid connection, at the right time.

Private capital is taking notice. Global infrastructure fundraising reached nearly $200 billion in 2025, while energy and power represented almost half of infrastructure deal value. The first phase of the energy transition was about building generation. The next phase may be about solving the bottlenecks around it.

The IEA expects global electricity demand to grow at an average annual rate of 3.6% between 2026 and 2030, compared with 2.8% during the previous decade. In absolute terms, the difference is enormous. Between 2015 and 2025, the world added approximately 700 TWh of electricity consumption per year. Through 2030, that figure is expected to reach approximately 1,100 TWh annually.

The drivers are also becoming more diverse. Industrial activity remains critical, particularly across China, India and Southeast Asia. But developed economies are returning to electricity growth as data centers, EVs, heat pumps, cooling and advanced manufacturing add new sources of demand.

The United States provides perhaps the clearest example. After nearly 15 years of muted growth, US electricity consumption is expected to rise close to 2% annually through 2030 under the IEA's latest outlook. Data-center expansion alone is expected to account for approximately 50% of incremental US electricity demand over the period.

This changes the economics of power infrastructure. Assets built for a world of flat demand are increasingly serving a world where electricity consumption is once again tied to economic growth—and increasingly to the growth of the digital economy. For investors, the implications extend beyond utilities. Greater electricity consumption requires more generation, transmission, substations, transformers, storage, electrical equipment and services. Electricity demand growth is becoming an infrastructure capex cycle.

The AI investment boom is rapidly becoming an energy story. Data-center electricity consumption reached approximately 485 TWh globally in 2025, growing 17% in a single year. Electricity consumption from AI-focused facilities grew even faster—approximately 50%.

The IEA now expects total data-center electricity consumption to approach 950 TWh by 2030, roughly double today's level and equivalent to around 3% of global electricity demand. AI-focused data centers grow faster still, with their electricity consumption expected to roughly triple between 2025 and 2030. Globally, that remains manageable relative to the entire electricity system. Locally, however, the story looks very different.

Data centers are geographically concentrated. Instead of millions of distributed loads appearing gradually across an electricity system, hyperscale facilities can require hundreds of megawatts—or eventually gigawatts—of capacity in a single location. The United States and China are expected to account for most of the incremental data-center power consumption through 2030. In the US alone, data centers could represent roughly half of total electricity-demand growth.

Meanwhile, AI infrastructure continues to become more power dense. The IEA estimates that the power density of AI servers increased approximately 11x between 2020 and 2025, with another substantial increase expected as advanced computing architectures scale.

There is an important investment distinction here. AI does not necessarily create a global shortage of electricity. It creates localized scarcity of power capacity, grid connections and electrical equipment. That is where the private-capital opportunity becomes considerably more interesting.

The electricity investment boom contains a major imbalance. Around $1 trillion is invested annually in generation assets. Global grid investment, by comparison, remains near $400 billion. That gap is increasingly visible in the physical system. More than 2,500 GW of renewables, storage and large-load projects—including data centers—are currently stalled in grid-connection queues globally.

To put that number in perspective, this is not simply a pipeline of hypothetical renewable projects. Increasingly, the queue contains both sides of the power market: companies trying to connect new generation and companies trying to connect enormous new sources of demand. The IEA estimates annual grid investment needs to increase roughly 50% by 2030 from today's $400 billion level simply to accommodate expected demand.

Money is only part of the problem. Permitting can take years. Transformers and cables face supply-chain constraints. Utilities must coordinate large projects across regulatory jurisdictions. Grid equipment prices have increased. And the timelines involved are dramatically longer than those of the industries creating the new demand. A data center can move from planning to operation within a few years. Large transmission projects can require far longer.

The mismatch creates scarcity value around infrastructure that already exists. Existing grid connections become strategic assets. Brownfield generation can become more attractive than greenfield projects. Transmission equipment suppliers gain pricing power. Electrical contractors face expanding backlogs.

Capital deployment confirms the magnitude of the shift. BloombergNEF estimates global investment in the energy transition reached a record $2.3 trillion in 2025, increasing 8% from 2024. The largest category was electrified transport at $893 billion, followed by renewable energy at $690 billion and grids at $483 billion.

That composition matters. The energy transition is no longer simply a renewable-generation investment story. Grid infrastructure alone is attracting nearly half a trillion dollars annually, while storage, nuclear, electrification and other technologies broaden the investable ecosystem.

At the same time, the digital and energy capex cycles are beginning to converge. BNEF estimates roughly $500 billion was invested in data centers during 2025—an extraordinary amount for an infrastructure category whose expansion is increasingly constrained by access to electricity. Energy-transition M&A is also recovering. BloombergNEF recorded $99.1 billion of M&A activity during 2025, up 37% from the previous year, with clean-power and building-related transactions benefiting in part from global data-center buildouts.

This is the defining feature of the current cycle. The demand for power is not being driven solely by climate targets or government incentives. Some of the world's largest technology companies are becoming major buyers of generation, storage and energy infrastructure because electricity is becoming a prerequisite for their own growth. For investors, that introduces a powerful new source of demand.

The capital requirements are far too large for public-sector balance sheets alone. McKinsey estimates the world requires approximately $106 trillion of infrastructure investment through 2040. Energy and power account for around $23 trillion of that requirement. Private capital is already responding. Global infrastructure fundraising reached nearly $200 billion in 2025, an all-time high and almost 60% above 2024. It also surpassed the previous fundraising record of $180 billion set in 2022.

Perhaps more important is where that capital is being deployed. Energy and power represented nearly half of global infrastructure deal value in 2025. Digital and telecommunications represented another quarter. Together, the two verticals at the center of the AI-power convergence represented approximately 75% of infrastructure deal value.

LP appetite is moving in the same direction. In McKinsey's survey of roughly 300 global LPs, 51% said they intend to increase allocations to infrastructure over the next three years, compared with 35% for buyout and 30% for real estate.

Infrastructure is also moving beyond traditional core assets. Capital is increasingly targeting core-plus and value-added strategies, including brownfield projects, development platforms, infrastructure-adjacent services and technology providers. That evolution opens the power investment cycle to a much broader universe of private-equity investors.

The most obvious way to invest in the power supercycle is to own generation. It may not be the only—or even the most interesting—way.

Renewables remain essential because of their cost competitiveness and relatively short development timelines. Natural gas offers dispatchable capacity. Existing nuclear provides reliable baseload generation while new technologies could eventually expand the opportunity. The critical differentiator is increasingly time-to-power. An existing asset with generation capacity and grid access may command strategic value that cannot easily be replicated by a greenfield project.

Transmission lines, substations, transformers, switchgear and related equipment sit directly in the path of the coming capex cycle. The thesis is technology-agnostic: whether incremental generation comes from gas, solar, wind or nuclear, electricity still needs to move through the grid.

A more complex electricity system requires flexibility. Battery storage can shift supply across hours, reduce congestion and help integrate intermittent generation. The IEA estimates 20–25 GW of batteries could be installed directly at data centers by 2030.

For hyperscalers and large industrial users, waiting years for grid upgrades may not be economically viable. On-site generation, microgrids, battery systems and dedicated power arrangements can allow large users to move faster.

This may be the most traditional PE opportunity. Electrical contractors, engineering companies, transformer and switchgear suppliers, maintenance businesses, grid software and specialist infrastructure-services companies can provide exposure to rising power capex without directly underwriting electricity prices. Many operate in fragmented markets where sponsors can deploy familiar buy-and-build strategies.

The energy supercycle therefore extends well beyond mega infrastructure funds. Middle-market PE can sell the picks and shovels.

Power has many of the ingredients private capital looks for: enormous capital requirements, long-duration demand, barriers to entry and essential infrastructure. But a structural tailwind does not automatically produce attractive returns.

Permitting remains slow. Grid projects can take years to approve. Transformers and other equipment remain supply constrained. Electricity-market regulation varies dramatically by geography. Construction costs can change before projects become operational. AI introduces another variable. Data-center electricity demand is growing rapidly, but forecasting AI several years forward remains unusually difficult. Hardware becomes more efficient. Models change. New applications can dramatically increase consumption. Some announced data-center projects may never be completed. The IEA therefore emphasizes a range of potential outcomes rather than treating today's pipeline as guaranteed demand.

There is also a more familiar private-equity risk: price. Once investors recognize that grid access, generation capacity or electrical equipment is scarce, that scarcity can quickly become embedded in valuations. Buying the right theme at the wrong multiple remains a perfectly effective way to generate mediocre returns.

The key underwriting question is therefore more nuanced than whether power demand will rise. Investors need to identify where demand growth creates durable scarcity, where that scarcity produces pricing power or contracted cash flows, and where those economics have not already been fully capitalized into entry valuations.

The global economy is entering a new electricity investment cycle. Demand is accelerating. AI is creating enormous concentrated loads. Generation investment is booming. Grid investment is struggling to keep pace. And private capital is increasingly financing the infrastructure connecting all four.

The numbers reflect the shift: approximately $1.5 trillion of annual electricity-sector investment, 2,500+ GW sitting in connection queues, almost $200 billion of infrastructure fundraising and nearly half of infrastructure deal value flowing into energy and power.

For mega infrastructure managers, the opportunity includes generation, transmission and large-scale energy platforms. For traditional private equity, the opportunity may be one layer removed: electrical equipment, engineering, storage, grid technology and services required regardless of which generation technology ultimately wins.

The first phase of the energy transition was about producing cheaper electricity. The next phase is about getting enough of it to the right place at the right time. AI gets the headlines. Power gets the capex.

International Energy Agency (IEA), World Energy Investment 2025, 2025.
https://www.iea.org/reports/world-energy-investment-2025

International Energy Agency (IEA), Electricity 2026: Analysis and Forecast to 2030, 2026.
https://www.iea.org/reports/electricity-2026

International Energy Agency (IEA), Key Questions on Energy and AI, 2026.
https://www.iea.org/reports/key-questions-on-energy-and-ai

International Energy Agency (IEA), Energy and AI, 2025.
https://www.iea.org/reports/energy-and-ai

BloombergNEF, Energy Transition Investment Trends 2026, 2026.
https://about.bnef.com/insights/clean-energy/bloombergnef-finds-global-energy-transition-investment-reached-record-2-3-trillion-in-2025-up-8-from-2024/

McKinsey & Company, Global Private Markets Report 2026: Infrastructure, 2026.
https://www.mckinsey.com/industries/private-capital/our-insights/global-private-markets-report/infrastructure