AI’s Power Hunger: Why Energy Investments May Eclipse AI Itself

Venture capital has poured over half a trillion dollars into the artificial intelligence sector in the last five years. However, a latest report suggests the smartest AI investment today might not be in AI technology itself, but in the energy infrastructure needed to power it. Growing concerns about power constraints and delays in data center projects are shifting investor focus toward energy solutions, from grid-scale batteries to innovative power conversion technologies.

The Data Center Bottleneck

Sightline Climate researchers estimate that 30–50% of announced data center capacity may be delayed in 2026, with access to power being a primary obstacle. Sightline Climate is currently tracking 190 gigawatts (GW) across 777 large data centers and AI factories announced since 2024. Despite plans for at least 16 GW of capacity to come online in 2026, only 5 GW is currently under construction, and approximately 11 GW remains in the announced stage with no visible construction progress.

Projected delivery dates are proving unreliable. In 2025, 26% of expected capacity was delayed, and another 10% of projects pushed back their commercial operation dates. Given this track record, significant delays are anticipated for 2026 as well.

Hyperscalers Seek Independent Power Sources

The power supply squeeze is prompting hyperscalers to seek alternatives to relying solely on the traditional power grid. While grid-connected power remains the most common model, an increasing number of projects are incorporating on-site power sources or hybrid approaches. New Era Energy & Digital’s 7 GW project in New Mexico, for example, is large enough to justify its own dedicated power generation.

This shift is driven by shortages of power generation equipment, such as gas turbines, and the limitations of an aging grid infrastructure.

Investment Shifts Towards Energy Solutions

Tech giants like Google and Meta are already investing heavily in renewable energy projects, including solar, wind, and nuclear, to power their operations. They are also supporting emerging technologies like Form Energy’s 100-hour battery through direct investments and collaborations with utilities.

Numerous startups are developing technologies to address the power challenge. Companies like Amperesand, DG Matrix, and Heron Power are focused on new power conversion technologies, while Camus, GridBeyond, and Texture are building software to optimize power flow management.

The Rise of Grid-Scale Batteries

Grid-scale batteries are poised for significant growth. The U.S. Energy Information Administration projects that the U.S. Will have nearly 65 GW of battery storage capacity by the end of 2026. Form Energy, for instance, is raising a $500 million round in anticipation of an initial public offering (IPO).

Underrated Tech: The Transformer

Beyond batteries, the humble transformer – a technology over 140 years old – is receiving renewed attention. Traditional transformers, built with iron and copper, are becoming too bulky for the increasing power demands of data centers. Investors are backing startups developing solid-state transformers, which use silicon-based power electronics to offer greater flexibility and potentially lower costs despite a higher initial price.

A Hedge Against an AI Bust?

While investments in battery and transformer companies are currently smaller than those in AI, they offer a potentially more stable investment opportunity. As electrification expands across various sectors, the demand for power will only increase, providing a hedge against a potential downturn in the AI market. The necessitate for reliable and scalable power infrastructure may ultimately prove to be a more secure investment than the AI boom itself.