When Sam Altman Arriving at Helion Energy's small office in Redmond, Washington, in early 2014 with textbooks on nuclear fusion under his arm, the company focused its efforts on research and development. By the time he left, a few days later, he had convinced the fusion energy startup to chart a more aggressive path to deployment, CEO David Kirtley recalls. A year later, Altman, who co-founded OpenAI invested $9.5 million in Helion around the same time, taking on the role of chairman. He poured another $375 million into Helion in 2021, making it one of the biggest personal bets in his multibillion-dollar portfolio.
Nuclear fusion, once a government initiative, has now become a private capital race, much of it funded by the same people who create energy-intensive AI and pursue the goal of creating systems with human-like intelligence, known as artificial general intelligence (AGI). Total funding for fusion power jumped from $1.7 billion in 2020 to $15 billion as of September 2025, according to the study. report EU body Fusion for Energy. Along with Altman, who said The future of artificial intelligence depends on an energy breakthrough. Helion's investors include OpenAI backer SoftBank, as well as Facebook co-founder and early backer Anthropic. Dustin Moskowitz. Nvidia has backed Helion's competitor, Commonwealth Fusion Systems (CFS). So did Google, which also invested in another player, TAE Technologies. “AI is a big driver [due to] energy is needed … to power data centers,” says Troy Carter, director of fusion energy at Oak Ridge National Laboratory.
Recent advances in engineering and an influx of money from investors eager to chase trips to the moon have led some firms to promise energy capacity for years rather than decades. They have yet to prove that the technology works, but if fusion works, it will provide carbon-free energy without the seasonal fluctuations of solar and wind power or the long-lived radioactive waste of nuclear fission – a breakthrough that will not only lower energy bills, but also change what is possible.
Nuclear fusion, the same reaction that powers the Sun, produces energy through a process opposite to that of modern nuclear power plants, combining light atoms rather than splitting heavier ones. Deep in the star's core, this occurs in plasma—superhot, electrically charged gas. Recreating all this on Earth caused all the engineering problems.
For decades, even when scientists were able to trigger a fusion reaction, it produced less energy than was required to heat the plasma, which was called scientific break-even. But in 2022, researchers from Lawrence Livermore National Laboratory made history. Using giant lasers to briefly crush a tiny pellet of fuel, they demonstrated for the first time a nuclear fusion reaction that generates more energy than was used to heat the plasma. No private company has reached this milestone.
If or when they do, feeding electricity into the grid will require taking one more step: generating not just more energy than was used to heat the plasma, but enough to power the entire generator, which is known as engineering break-even. Of the firms aiming to reach this milestone, Helion is on the most optimistic schedule. The company expects a commercial version of its machine to provide electricity by 2028 from a facility in Malaga, Washington, where construction began in July. Helion has already signed an agreement to sell 50 megawatts of fusion energy to Microsoft and faces financial penalties if it deviates from schedule.
Unlike most fusion projects, which use boiling water to spin a turbine, Helion plans to harvest electricity by slamming two plasma rings together at about a million miles per hour. The collision will result in nuclear fusion, disturbing the magnetic field, which in turn produces energy. Kirtley says the setup currently recovers about 96% of the energy it consumes—kind of like how an electric car uses regenerative braking to charge a little while driving. This already puts him on the brink of breakeven. Polaris, the seventh-generation Helion prototype, was expected to achieve technical break-even in 2024. The prototype was first flown late that year. Kirtley declined to share the results.
Kirtley, who credits Altman with inspiring him to “move faster and on a bigger scale,” envisions more than just being the first to build a fusion plant. “Our goal is… to build one generator per day and deploy fusion systems around the world. And to do it quickly.”
According to Carter of Oak Ridge, such ambitious thinking has brought the arrival of nuclear fusion much closer. In 2020, he authored a Department of Energy report that said a pilot fusion plant could be built by the early 2040s, but he now believes that goal could be achieved by the mid-2030s. Beyond capital, AI is a useful tool for scientific progress. “The advent of artificial intelligence has made some very difficult problems in the plasma space more accessible,” says Nuno Loureiro, director of MIT's Center for Plasma Science and Fusion.
Read more: A breakthrough in nuclear fusion may be closer than you think
Even if Helion's plan is a few years behind schedule, it could still be a world first. And if his approach doesn't work, others will line up behind him.
California startup Pacific Fusion says it has developed a machine that will achieve engineering break-even using the same approach as the Lawrence Livermore National Laboratory device. It launched publicly in 2024, raising $900 million from investors including former Google CEO Eric Schmidt and Microsoft AI CEO Mustafa Suleiman.
CFS, which spun out of MIT in 2018 to become the world's most funded fusion startup, is taking a different approach: creating a magnetic bottle that holds super-hot plasma in place. CFS is building a pilot project that it expects to reach a scientific break-even point in 2027. The company is so optimistic that it has begun parallel work on a commercial power plant that hopes to supply that energy to the grid in the early 2030s. Google has already agreed to buy 200 megawatts. “[Having] those big hyperscalers behind us are very useful,” says Brandon Sorbomco-founder and chief scientist at CFS, adding that it signals to suppliers supplying superconducting magnets and other difficult-to-manufacture materials that “this is not a one-time science experiment.” (Investors in Commonwealth Fusion Systems include TIME co-chairman and owner Marc Benioff.)
Meanwhile, New Zealand startup OpenStar produced plasma with a relatively modest $10 million in funding in late 2024 and has since raised another $14 million. His prototype turns the “magnetic bottle” concept inside out, with a super-strong magnet at the core of the reactor around which the plasma is held.
While Carter is optimistic about the number of startups claiming to be industry leaders, he warns that high-profile failure could scare off investors and undermine confidence in the field. “You hope the hype doesn't get too big; [that] The collapse of one of the most visible companies does not stop the progress we have elsewhere,” he says.
However, the merger can't happen fast enough for companies like Google and Microsoft. Both companies are building new data centers to support artificial intelligence, while Microsoft plans to be carbon negative by 2030 and Google aims to be carbon neutral. Data centers powering AI operate 24/7; Without breakthroughs in energy storage, variable wind and solar will not be able to reliably cover this load. And supply is limited: U.S. power production has remained virtually unchanged since 2010, when China overtook China as the world's largest electricity producer. Even fossil fuels may find it difficult to scale as computing technology explodes. Nvidia's Altman and Jensen Huang now identify energy as a key bottleneck.
But the discovery of fusion power will have implications far beyond powering data centers. It could be built where the energy is needed, rather than where wind or solar conditions are best. And with abundant energy, global economic and geopolitical dynamics could be turned on their heads. “Most of our wars are fought over energy,” Carter says. “If that's no longer the driving force, it makes a big difference.”
With reporting by Billy Perrigo
