A lot can change in a few months.
The climate tech world hasn’t exactly been turned upside down, but it’s definitely more askew than it was in the summer. The U.S. federal election results may have imperiled the startup-friendly Inflation Reduction Act, likely throwing a wrench into many companies’ business plans.
Yet at the same time, AI’s skyrocketing computing needs have driven data center operators scouring the earth for sources of electricity, bringing a surge of interest in a range of power sources, including nuclear, renewables, batteries, and even fusion.
As 2025 dawns, it’s a good time to look at the trends that are likely to define the coming 12 months.
Nuclear power received a lot of love this past year, from Microsoft restarting a reactor at Three Mile Island to Google signing a 500-megawatt deal with startup Kairos. The driver? Data centers, data centers, data centers. With AI servers facing a power shortage as soon as 2027, tech companies have been racing to get their hands on electricity wherever they can find it.
Nuclear power is one of those places. Historically, adding nuclear capacity meant big power plants that take a decade or more to build. But a new wave of startups has been proposing smaller designs that can be more easily mass produced, or so the thinking goes. They haven’t been tested at scale yet, and the success of nuclear startups will depend on how the first few go.
In their favor, those companies have the benefit of a newly streamlined regulatory process, which should help speed the time from proposal to construction.
But they’re also facing stiff competition from renewable power sources, which are proven and quick to deploy. Unless there’s a breakthrough in efficiency for AI model training or inference, expect to hear more about tech’s love affair with nuclear in the coming year.
We’re just over two years out from the National Ignition Facility’s groundbreaking announcement that it had produced the world’s first controlled, net-positive fusion reaction. Fusion startups undoubtedly used the news to kickstart their fundraising efforts. Among the winners this year: Acceleron Fusion, Marvel Fusion, Marathon Fusion, Type One Energy, Xcimer Energy, and Zap Energy.
Expect more this year, too. Building a fusion power plant, even a demonstration unit, is expensive. Several startups have begun work on prototypes, demos, and even commercial reactors, including Commonwealth Fusion System and Zap Energy. Many have goals of hooking power plants up to the grid in the early 2030s, which means they have a lot of work to do in the coming years. And that means they’ll need more money soon.
It’s a risky technology, but the rewards include remaking the trillion-dollar energy sector. If companies are able to hit scientific and engineering milestones, expect more investors to line up in 2025.
Few sectors are as exposed to potential changes to the Inflation Reduction Act as hydrogen. Many startups are hoping to eventually deliver the gas at $1 per kilogram, but not until later this decade or early next.
To get there, they’ve been optimistic that the two-year-old IRA can help them bridge the gap by way of a $3 per kilogram subsidy for hydrogen produced by renewable electricity. If that provision is nixed, a number of hydrogen startups could be in danger of going belly up. Large companies have already grown skittish.
At the same time, scientists and investors have warmed to so-called geologic hydrogen, or hydrogen that’s produced naturally within the Earth. Could it save the industry? The next 12 months might be a make or break moment.
The coming year will almost certainly bring more changes, especially as politicians and regulators grapple with growing power demand from AI. Changes in the permitting process could drive a wave of investment in grid-related technologies, but if those efforts stall, expect more companies to sign deals with power providers to sidestep the grid and connect directly to data centers.
Investors have told me that it will probably be challenging for many startups to raise new funding in the coming year. The most exposed companies are those that are overly dependent on vulnerable subsidies.
But 2025 is just as likely to throw a curveball — it’s helpful to remember that the current climate tech wave emerged during the first Trump administration. Next year might have some surprises in store, too.
