Fusion energy is having a moment.
Last week, Commonwealth Fusion Energy Systems, which is backed by Google, announced it had raised nearly $3 billion in capital and would start developing a fusion power plant in Virginia. That came as a prominent Department of Energy scientist helped launch a new company to commercialize fusion.
The Nuclear Regulatory Commission, meanwhile, is planning to release a new rule this fall to develop fusion energy as investors pour unprecedented amounts of money into startups.
The moves raise a central question for the future of the grid: Is fusion finally at a turning point that will lead to its commercial development, or is it an industry that won’t live up to the hype?
For many investors, technical advances in the past three years signal that fusion can be an option for electricity by 2035 or earlier.
Investors “see the maturing of the ecosystem. They see the capabilities of the company, and they want to be a part of what will be the largest energy transformation in human history,” Commonwealth Fusion CEO and co-founder Bob Mumgaard told reporters last week.
But experts caution many technical barriers remain for fusion, and some observers question whether it will be a major factor for the grid, at least before 2050.
Developers “are going to be very unhappy when they find their investments are not yielding any return on any time scale of interest to them … that could be even bad for the longer term prospects of fusion,” said John Holdren, a research professor in Harvard University’s Kennedy School of Government and White House science adviser during the Obama administration. If a commercial fusion power plant is built, it won’t happen until the second half of the century, he predicted.
Which argument is correct could determine the fate of more than $9.7 billion in industry investments, which have grown fivefold since 2021, according to the Fusion Industry Association. More than 30 companies are eyeing the early 2030s as the likely period for pilot power plants to come online, although a handful of developers say they can start up new plants before the end of this decade.
Industry supporters range from Breakthrough Energy Ventures, spearheaded by Microsoft co-founder Bill Gates, to former Energy Secretary Ernest Moniz, who is on the board of fusion company TAE Technologies.
Commonwealth says it raised more than $863 million in its latest funding round to near the $3 billion mark and is on track for its fusion power plant near Richmond, Virginia, to be operational in the early 2030s. In a deal inked this summer, Google plans to buy 200 megawatts of power from the plant. The company has built 65 percent of its fusion reactor, Mumgaard said.
In addition to Commonwealth, Helion Energy said in July it had started construction of what it hopes will be the world’s first fusion power plant in Washington state. The company, which is backed by OpenAI CEO Sam Altman, said it was on track to deliver electricity within three years to Microsoft under a power purchase agreement. In its annual report this summer, the Fusion Industry Association detailed more than 50 companies aiming to develop fusion through various technological approaches.
One of the newest players is Inertia Enterprises, which launched last week and aims to advance technology from Lawrence Livermore National Laboratory that achieved “ignition” in 2022. In that experiment, a fusion reaction for the first time produced more energy than was put into it by lasers, long a technical barrier.
The company is led by Twilio co-founder Jeff Lawson, fusion power plant designer Mike Dunne and Annie Kritcher, who led the lab team at the National Ignition Facility in reaching the milestone three years ago. Inertia says it has licensed nearly 200 patents, plans to develop low-cost lasers and has project agreements with the national lab.
“There’s a great many startups pursuing fusion, obviously, but none of them [are] really taking this obvious path that is leveraging maximally what has now been proven at Livermore,” Lawson said in an interview.
While the energy yield during the Lawrence Livermore experiment lasted for a fraction of a second and was not intended as a commercialization exercise, it became a chief driver of the current cash flood, said Mike Campbell, a fusion expert at University of California, San Diego. Since the initial milestone, DOE scientists have reached ignition eight more times, according to the lab.
Advances in superconducting magnets the past five years that allow them to operate at higher temperatures also have helped spur investor interest, according to Campbell.
Magnets are important for one method of fusion because they can confine plasma created from heating up hydrogen isotopes to over 100 million degrees. The process used by Lawrence Livermore is different: It involves pelting fuel with powerful lasers to get atomic nuclei to collide.
Christofer Mowry, CEO of Tennessee fusion startup Type One Energy, pointed to supercomputing advances in the past 20 years as key to making fusion more commercially viable.
DOE has spurred the development of exascale computers, giving access to companies like Type One.
“You need modern supercomputing to model and simulate” fusion physics, Mowry said.
The relative abundance of deuterium, a hydrogen isotope envisioned for many fusion reactions, and lithium, which is used to make a second hydrogen isotope called tritium, also is generating enthusiasm, said Campbell. “You have access to a lot of fuel … it’s globally dispersed,” said Campbell.
A fusion plant, if built, wouldn’t have the same risk of accidents as traditional nuclear reactors do, he said. “You can never have a meltdown. If something goes wrong, the [fusion] reaction will just stop,” he said.
“There’s no question in my mind that fusion will eventually power the planet,” said Campbell.
‘Dangerous’ hype?
Unlike traditional nuclear plants, which involve splitting heavy atoms, fusion involves combining or fusing atomic nuclei to release energy — the same reaction that powers the sun and the stars. In theory, it could power the world several times over with carbon-free energy that produces less waste than traditional nuclear reactors, making it of interest to technology companies searching for power to support artificial intelligence in the decades ahead.
But fusion faces many technical challenges in moving to a stage where it provides electrons to the grid, particularly at scale and in an economic way.
The Lawrence Livermore 2022 experiment demonstrated that a reaction can create more energy than lasers provided, but that gain doesn’t account for the energy needed to power the lasers in the first place. To match power coming from the grid, the reaction would have had to produce 100 times more energy. For fusion to eventually provide electricity affordably, it would need to consistently provide more power than the energy going into a plant — a feat that has yet to be demonstrated.
Fusion using deuterium and tritium also generates high-energy neutrons that can damage equipment and make it radioactive, said Andrew Sowder, a senior technical executive at the Electric Power Research Institute. It’s an engineering challenge to make material that can withstand a large number of neutrons hitting it, he said.
Additionally, fusion developers have to figure out how to harness energy from plasma at scale in a form that can be used for power. This typically would require a machine called a blanket that “has to do multiple things in a small space,” including converting kinetic energy from neutrons into heat, said Sowder.
To be viable, the industry would need to overcome other obstacles, such as developing cooling equipment that works with fusion, using materials that have not been tested extensively, building a supply chain and ensuring an adequate supply of well-trained workers.
In 2024, Holdren wrote a paper saying hype about fusion electricity is “dangerous” because it could set back the timetable for eventual success by frustrating investors, as well as drive a “false hope” of a silver bullet for climate change.
Among other things, radioactive equipment created from neutrons pelting a fusion machine might have to be replaced too often to be economic and would require advanced robots for maintenance, he said.
“Assuming robots are going to be good enough by 2035 to maintain the inside of a radioactive fusion reactor is a stretch,” he said in an interview.
Producing large amounts of tritium in a fusion reactor, as some companies are envisioning, could create a “significant radiological hazard,” added Holdren, who supports fusion development despite skepticism about current timelines.
“The degree of tritium control needed to stay within current guidelines for public radiation exposure at a nuclear plant boundary would be extremely challenging to achieve,” his paper said.
Nonetheless, companies like Google say the possibility of a breakthrough in fusion for electricity is worth the financial risk. Some fusion developers are working on new technology they say will address major technical challenges, although experts have questioned their technical predictions.
Helion, for example, claims it will be able to pull electricity directly from plasma, rather than trying to heat water to create steam to power a turbine. It’s also planning to use a fuel called deuterium-helium 3 that it says will “reduce neutron emissions.”
“It’s a hard path ahead — no doubt about it. But if we’re successful, this is world changing,” said Lawson of Inertia Enterprises.
Trump, DOE and the megalaw
The flood of investments comes as Energy Secretary Chris Wright has been pushing for fusion, saying in June the technology is at a tipping point “where things are going to happen fast.”
“We need to step up our national efforts in fusion,” he said, warning that the U.S. is in a race with China to develop the technology. Advances in AI would help develop fusion faster, he predicted.
The NRC is expected to release a rule this November that would give fusion developers greater regulatory certainty, carving out a path that regulates fusion differently from nuclear fission.
A person familiar with internal NRC operations told POLITICO’s E&E News that NRC commissioners were ready to advance the staff fusion rule in May, but confusion around the implementation of a February executive order from Trump instructing independent agencies to submit “significant regulatory actions” for White House review delayed progress for months.
In July, the proposed rule was received by the White House, where it is still pending review, according to the Office of Management and Budget website.
The person, granted anonymity due to the sensitivity of ongoing proceedings, said the commissioners were planning one big change to draft guidance released last year, which would swap out the staff definition of a “fusion machine” to one preferred by the industry.
Adopting the industry’s definition would simplify licensing, but it could also limit the NRC’s ability to tailor safety and environmental reviews for commercial fusion facilities. States are already moving forward on companion regulations in anticipation of the NRC rule, said Andrew Holland, CEO of the Fusion Industry Association.
But Trump’s budget request for fiscal 2026 called for a roughly 6 percent cut in fusion research at the Department of Energy. While that is significantly less than some other proposed cuts at the department, it falls short of what industry leaders say is needed.
Holland said he would like to see an infusion of DOE cash of $5 billion to $10 billion — an ask that may be unlikely considering the administration’s plans to cut DOE spending. He’s also pushing for DOE to focus more on commercialization, and not just research, with fusion. “There’s some work that needs to be done with the Office of Science,” Holland said.
DOE’s milestone program, which was authorized by the Energy Act of 2020, selected eight companies two years ago to develop fusion pilot plants. Those companies are making “great progress,” DOE fusion official Colleen Nehl said in June, according to NuclearNewswire.
In an emailed statement, DOE said it is still developing its strategy on fusion, adding it would soon be releasing a road map that will “outline the path to U.S. fusion energy dominance by highlighting key goals, actions to take and metrics and milestones to track progress.”
The Republican One Big Beautiful Bill Act signed on July 4 allows fusion companies to obtain tax credits, and unlike wind and solar, the time lines were not changed from prior to the law’s passage. “It did nothing positive for fusion, but it also didn’t do anything negative for fusion,” Holland said of the megalaw.
In the meantime, it remains unclear what Trump or DOE may ultimately do, he said.
“Will there be more than just, when asked a question, [Chris Wright] says something nice” about fusion? he asked.
“Are we going to move off the status quo here … or is the U.S. government basically just going to say ‘we’re going to let the private sector lead on this,’” Holland said.
