Zanskar, a conventional geothermal startup, is using field geologists to collect data, with the hopes of eventually training a generative model that locates new geothermal hotspots.
Zanskar
Could the answer to the AI industry’s prayers for abundant energy be just under our feet?
As the industry increasingly looks to unusual solutions to overcome energy and environmental constraints, geothermal energy companies see this as possibly their big moment.
The US electrical grid load, or demand for energy, has grown by 1.7% since 2022, according to the US Energy Information Administration, and is expected to increase by 25% over the next four years. By 2028, new data centers are projected to need an estimated 44 gigawatts of additional power capacity on the grid.
Adding to the pressure, several AI hyperscalers, including Microsoft and Alphabet, have set their own climate targets to reach net-zero by 2030. Meanwhile, these next four years are expected to be critical for both firms’ data center buildouts.
Next-generation nuclear technologies have benefited from the energy crunch. In 2025 alone, VCs invested $6 billion across 108 deals in nuclear startups globally. Even fusion, a nuclear technology yet to be proven in a laboratory, is gaining traction as startups focused on the approach raise larger rounds and secure bigger purchase agreements.
The timelines for getting more nuclear energy on the grid are uncertain. Even conventional nuclear fission, done in large plants rather than in next-generation small modular reactors, is beset by supply chain and geopolitical complications.
Enter geothermal energy, whose proponents say is getting overlooked.
“Geothermal has a very strong competitive advantage here,” said Diego D’Sola, head of finance at geothermal startup Zanskar. “The current alternatives on the table are next-generation nuclear—there, you have buildouts coming in the early 2030s in the best case scenario.”
Founded in 2021, Zanskar is building predictive heatmaps of the American West to identify new hotspots of previously undiscovered geothermal activity. The startup secured a $115 million Series C last month led by Spring Lane Capital.
The company says that “there are many, many more orders of magnitude more hidden, so-called ‘blind systems’ out there than there are ones that have obvious surface manifestations,” D’Sola said. Predictive AI technology will help uncover these blind sites, the company is betting.
One of Zanskar’s recent discoveries, a previously untapped geothermal hotspot that it nicknamed “Big Blind” in western Nevada, appears to hold promise for the energy source. “The models have started to make unintuitive predictions” that humans would not have made, according to Zanskar chief scientist Joel Edwards.
Despite its potential, geothermal energy has remained relatively under the radar, with VC funding concentrated in just a few companies. Fervo and Eavor, two enhanced geothermal startups, have accounted for all but one deal of $100 million or more in the geothermal category in the last five years. Unlike conventional geothermal technology, which taps naturally occurring hotspots, enhanced geothermal companies use artificial reservoirs to tap previously unreachable supplies of hot rock in impermeable sections of the Earth’s crust.
Fervo will be the first enhanced geothermal company to deliver large-scale electricity directly to the grid to go public.
“Conventional geothermal is being overlooked in the context of enhanced geothermal,” said Jason Scott, an investor at Spring Lane Capital who led its financing of Zanskar. “The VC industry tends to chase new and breakthrough technologies” while ignoring the potential of the not-as-new-and-shiny, he added.
Climate tech investors agree that the future grid will be a mix of energy sources, including some technologies that have yet to be commercially proven. But unlike the buzzy promise of fusion, geothermal energy offers a much faster timeline for scaling.
“Fusion timelines are so unknown—certainly claims and roadmaps exist, but the pathways to those depend on unpredictable technological breakthroughs,” said PitchBook senior emerging tech analyst John MacDonagh.
Geothermal energy’s timelines are more comparable to nuclear fission, which does not require the extreme temperatures of fusion, according to MacDonagh. There seems to be a greater appetite for small modular nuclear reactors that use fission technology, which, if scaled, could hypothetically be deployed faster than geothermal technology.
Geothermal energy is also held back by permitting laws, which can delay a project for months, if not years. But the energy source has a notable advantage: politics.
“Geothermal is sort of that bipartisan fuel source, where you can ‘drill, baby, drill,’ but you’re going down to get heat to spin a turbine with steam,” said Benjamin Baker, an investor at energy transition firm Greenbacker Capital Management.
Meanwhile, nuclear’s key feedstock, uranium, is already in a supply chain crunch and is highly subject to geopolitical winds. The bulk of the US supply comes from Russia and China. VCs are looking for routes to reshore that supply. Scott Nolan, a partner at Founders Fund, even stepped back from investing last year to found General Matter, a startup developing low-enriched uranium.
“I think without data centers, nuclear would not be as hot,” said Baker. But geothermal? “I think that it would be back, no matter what.”
Latest Research PitchBook Analyst Note: The Elevated Oil Stakes of Iranian Stability Iranian oil disruption could impact 3.5% of global supply. We examine risks to the Strait of Hormuz, US producer strategies, and PE investment opportunities in Iran’s energy sector if sanctions ease. January 21st, 2026
This article originally appeared on PitchBook News
