Venture capitalists have placed bigger bets on AI startups, investing more than half a billion dollars in the field over the past five years.
But these days, the smartest AI investment may be in the offing, according to a Sightline Climate report. Researchers have found that up to 50% of announced data center projects may be delayed. One of the main reasons is access to power.
Of the 190 gigawatts worth of data centers the company is tracking, only 5 gigawatts are under construction. About 6 gigawatts of data center projects in Sightline’s database came online last year. The largest percentage – about 36% – saw their periods decrease by 2025. The delay may slow down and affect large enterprises and other companies that use AI in their businesses.
That supply-demand squeeze is an opportunity for investors. Here is the reason.
Big tech companies like Google and Meta have dedicated large portions of their balance sheets to developing solar, wind, and nuclear projects. These companies also support emerging technologies such as Form Energy’s 100-hour battery by investing directly and partnering with utilities to accelerate their adoption.
A number of startups are pursuing technologies that address the energy crisis. For example, Ampersand, DG Matrix, and Heron Power are developing new energy conversion technologies, while companies like Camus, GridBeyond, and Texture are developing software that can control the flow of electrons.
Power remains one of the most important constraints for data centers, a shortage that won’t change anytime soon. AI is expected to drive data center energy utilization to 175% by 2030, according to Goldman Sachs.
Techcrunch event
San Francisco, CA
|
October 13-15, 2026
These grid shortages are unprecedented in modern times, and have driven up electricity prices across the country. That has pushed many technology companies to explore other ways to power their data centers. (The Trump administration, sensing a looming political crisis, urged tech companies to build their own energy sources, pay higher prices, or both. Most had already made plans to do so, of course.)
Other grid methods
Amazon, Google, Oracle, and other big tech companies have been working to reduce their dependence on the grid. Several data centers are planned using on-site power or a hybrid approach that combines on-site power with grid connection.
Large data centers are leading the way. Less than a quarter of projects have identified a local or hybrid energy source; together they represent 44% of the total volume.
The change was motivated in part by a shortage of power generation equipment – namely gas turbines – and an antiquated grid. That paved the way for other energy sources.
Google’s recent deal to power a new data center in Minnesota shows one way to tackle the problem. The company will combine wind and solar with a large 30-gigawatt-hour battery from Form Energy. Google has also partnered with Xcel Energy to create a new pricing structure that it says will help encourage the adoption of new technologies in the utility’s grid.
The Form Energy battery is not the only example. Grid-scale batteries are poised to take out a big nest in the electricity market. By the end of this year, the US should have about 65 gigawatts of battery storage capacity, according to the US Energy Information Administration. Like many of its peers, Form Energy is looking to capitalize on this momentum by raising a $500 million round ahead of the eventual IPO.
Underlying technology
Power supply is only part of the story. When power hits the grid or data center, it needs to be managed, a task that falls heavily on the humble transformer.
Most of today’s transformers use large coils wrapped in copper wire, a technology that is about 140 years old. It’s reliable, but it’s getting bigger as data center capacity demands. By the time rack servers hit 1 megawatt in maximum power, the electrical equipment needed to run them will take up twice as much space as the rack itself, one expert told TechCrunch.
That’s why investors have been flocking back to solid-state transformer startups recently, hoping that silicon-based electronics can replace the old iron-and-copper technology. They are more expensive than existing transformers, but they are also flexible enough to replace several pieces of equipment in a data center, which should make them cost competitive.
Overall, the scale of investment in battery and transformer companies has been much smaller than other blockbuster cycles we’ve seen in the AI industry.
That’s not a bad thing – those rounds are attractive to investors. And, as the world electrifies everything from transportation to heavy industry, demand for energy will only grow, giving investors a hedge against an AI bust. Perhaps the best AI investment is not in AI at all.
