Tech Titans Drive a Private Race Toward Carbon-Free Power
When Sam Altman first visited Helion Energy in 2014, the startup was focused on slow, methodical research. A few days later, the future OpenAI founder convinced the team to pursue aggressive commercial goals — a decision that would redefine the future of both companies. Within a year, Altman had invested $9.5 million and became Helion’s chairman. By 2021, his total stake ballooned to $375 million, making Helion one of the best-funded private players in the fusion energy race.
Once the domain of governments, nuclear fusion is now attracting private investors — particularly those driving the energy-hungry artificial intelligence boom. Global fusion funding soared from $1.7 billion in 2020 to $15 billion in 2025, according to Fusion for Energy. Alongside Altman, backers include SoftBank, Dustin Moskovitz of Facebook fame, and Nvidia, which also invested in Commonwealth Fusion Systems (CFS). Google, meanwhile, is supporting both CFS and TAE Technologies. “AI’s massive energy demand is accelerating fusion investment,” says Troy Carter of Oak Ridge National Laboratory.
From Scientific Breakthrough to Commercial Race
Fusion — the process that powers the sun — joins light atoms rather than splitting heavy ones, producing vast amounts of clean energy. For decades, scientists have failed to achieve more energy output than input, known as scientific break-even. That milestone was finally achieved in 2022 by the Lawrence Livermore National Laboratory. But the next step, engineering break-even — powering an entire generator — remains elusive.
Helion claims to be closest to crossing that line. Its upcoming commercial plant in Malaga, Washington, aims to supply the grid by 2028, under a deal to deliver 50 megawatts of fusion energy to Microsoft. The agreement includes financial penalties for delays. Helion’s seventh-generation prototype, Polaris, has already demonstrated near break-even results. CEO David Kirtley says their system “recaptures about 96% of its energy input,” using magnetic fields instead of turbines to generate electricity.
Rivals Closing In — From CFS to Pacific Fusion
The competition is heating up. Commonwealth Fusion Systems, a spinoff from MIT, is building a pilot reactor expected to hit scientific break-even by 2027, while also constructing a commercial plant for the early 2030s. Google has already agreed to buy 200 megawatts of its future fusion power. Pacific Fusion, backed by Eric Schmidt and Mustafa Suleyman, has raised $900 million to scale a design similar to the Livermore laser model. Even smaller startups like New Zealand’s OpenStar are gaining attention, achieving plasma generation with just $24 million in funding.
Despite the optimism, experts caution that overpromising could backfire. “You hope the hype doesn’t get too much,” says Carter, warning that one high-profile failure could cool investor enthusiasm. Still, with billions in private capital and AI’s energy appetite soaring, many see fusion as inevitable — not speculative.
The Energy Bottleneck Behind AI’s Growth
The AI industry’s drive toward artificial general intelligence (AGI) demands staggering energy resources. Microsoft and Google have committed to carbon neutrality by 2030, even as they build new data centers that consume gigawatts of power. U.S. energy output, meanwhile, has barely grown since 2010. Both Altman and Nvidia’s Jensen Huang now describe energy as the ultimate bottleneck for technological progress.
Fusion could break that constraint entirely. Beyond powering data centers, it could transform global economics and geopolitics by providing abundant, location-flexible clean power. As Oak Ridge’s Carter notes, “Most of our wars are fought over energy. If that’s no longer the driver, that changes things dramatically.”
