Why investors are betting on nuclear fusion even as it remains elusive

NUCLEAR fusion has long captured the scientific imagination. The dream is to harness the same process that powers stars to produce cheap, abundant and carbon-free electricity.

The technology remains nascent, even after decades of research. But that hasn’t deterred some of the world’s richest investors from backing startups that hope to solve one of the toughest challenges in science.

Interest in fusion has been further fuelled by the proliferation of data centres, as tech firms hunt for sources of around-the-clock power for their cloud computing and artificial intelligence operations.

US President Donald Trump’s social media company is getting into the game, announcing in late 2025 a surprising plan to merge with fusion developer TAE Technologies.

What’s the difference between nuclear fusion and fission? 

Nuclear fission entails splitting heavy atoms, such as uranium, into smaller ones to release energy.

The energy is then used to heat water and produce steam that spins a turbine to generate electricity. This is the process that commercial nuclear power reactors have run on for decades.

Nuclear fusion involves fusing light atoms into a single heavier element.

For example, hydrogen isotopes deuterium and tritium can be slammed together to form a helium atom.

The mass of the resulting nucleus is slightly less than the sum of its parts, and that difference in mass is released as energy, in accordance with Albert Einstein’s famous E = mc² equation.

Fusion is what powers the Sun. Ordinarily, the positive charges of nuclei repel each other. But extreme gravity in the core of stars creates enough heat and pressure for atoms to smash together.

On Earth, much higher temperatures are needed for atoms to collide – in the range of 150 million degrees Celsius, a temperature that is 10 times hotter than the centre of the Sun.

Neither fission nor fusion release greenhouse gas emissions. Fusion has additional safety advantages: It doesn’t produce any high-level, long-lived radioactive waste from spent fuel rods, and there’s no risk of a reactor meltdown.

If fusion power plants become a reality, they could be located closer to the populated areas where electricity demand comes from.

How close is fusion to commercial viability?

Fusing atoms together requires vast amounts of power. After decades of trying, scientists were finally able to achieve what’s known as “ignition” in 2022, whereby a controlled fusion reaction produced more energy than was used to trigger it with a blast from a massive laser.

The breakthrough at the Lawrence Livermore National Laboratory in California suggested the core physics of fusion energy had been cracked. After several unsuccessful attempts, the lab was able to repeat the achievement in 2023.

However, the net energy gains were small and exceptionally brief – well short of the scale and consistency required to supply the power grid. And while the amount of energy generated exceeded what the laser put in, it paled into comparison to the energy needed to power the laser in the first place.

Fusion developers and the most optimistic experts say the first reactor could deliver electricity to the grid within the next decade. But many others don’t expect that to happen until at least 20 or 30 years from now.

Based on that timeline, fusion is unlikely to play a significant role in reducing global emissions to net zero by 2050 – a target seen as essential to avoiding the worst impacts of climate change.

What are the hurdles to fusion technology taking off? 

There are a number of technical challenges that must be overcome to bridge the gap between the current experiments and commercialisation – not least the ability to consistently initiate, contain and sustain fusion reactions.

At the high temperatures needed for fusion, hydrogen becomes a plasma: an electrically charged superheated gas that’s too hot to come into contact with any solid material.

Scientists have yet to perfect the technology to contain the plasma and keep it stable so that the fusion reactions continue.

Fusion systems are also currently very expensive to construct and operate, and supply chains will have to be expanded to bring down costs. In the meantime, there are much cheaper and already proven low-carbon technologies that are widely available, such as wind and solar power.

Who are the leading players working on fusion technology?

Fusion development was initially led by publicly funded research programmes, but the private sector has become increasingly active.

There are more than 50 fusion companies worldwide, according to the Fusion Industry Association, including 29 in the US, 12 in Europe and three in China. As at July, they had raised around US$9.8 billion, more than double the total from three years earlier.

Tech billionaires including Amazon founder Jeff Bezos and Microsoft co-founder Bill Gates are among the prominent investors in fusion startups.

Commonwealth Fusion Systems, which was spun out of the Massachusetts Institute of Technology, accounts for almost a third of all the money that has been poured into the fusion industry. It has raised around US$3 billion from investors, including the venture arm of chips giant Nvidia.

Commonwealth is building a demonstration system and aims to produce net energy from this machine in 2027. It’s also developing a 400-megawatt power plant in Virginia, where booming data centre installations are driving up electricity demand.

A growing number of big tech companies have already inked agreements to purchase power from fusion developers, even though those startups are at best a long way off from being able to deliver electricity at scale.

Microsoft agreed in 2023 to buy some of the energy generated by a planned fusion plant from Helion Energy, a startup backed by OpenAI’s Sam Altman.

Meanwhile, Google agreed in June to purchase half the output from Commonwealth’s Virginia project, from which the fusion developer expects to begin delivering power to the grid in the early 2030s.

What about public sector efforts?

The Trump administration is looking to increase support for fusion, contrasting its campaign to stifle the growth of renewable energy. US Energy Secretary Chris Wright has said commercial fusion electricity could be possible in as soon as eight years, and his agency created a new Office of Fusion in November.

China has stepped up its fusion ambitions. State-owned China National Nuclear Corporation said it could start commercial operation of its first fusion generation project by 2050.

The International Thermonuclear Experimental Reactor (ITER) project in southern France is a joint effort by more than 30 nations – including the US, China, European Union members and Russia – to build the world’s largest fusion experiment.

The aim is to prove the feasibility of fusion as a large-scale and carbon-free source of energy.

The gigantic demonstration machine has been under construction since 2010 and has suffered from a number of delays and cost setbacks. The final bill could be at least US$25 billion, and those behind the ITER project have said that it won’t be fully operational until 2039.

ITER, as well as many of the leading fusion companies, is using a so-called tokamak, a design that dates back to the Soviet Union. Lasers and powerful electromagnets are arrayed around a super-cooled doughnut-shaped chamber to hold the superheated plasma in place.

What are the main fusion technologies?

Two major competing technologies have emerged, although some startups are pursuing a hybrid of these approaches.

Magnetic confinement is the most common method. The fuel is heated to hundreds of millions of degrees until it becomes a plasma, which is contained by powerful magnets to sustain a fusion reaction.

In theory, if the plasma can be held in a steady state, it could produce energy for decades. Most efforts use a tokamak or a twisted variant known as a stellarator.

A project based on magnetic confinement has yet to demonstrate net energy gain, and there are challenges in developing materials that can withstand super-hot temperatures and bombardment from atomic particles.

The main alternative technique is inertial confinement, which was used by Lawrence Livermore to reach its ignition milestone. Hydrogen-filled pellets are blasted with high-intensity laser beams to fuse hydrogen isotopes into helium.

A commercial plant using this approach would need to repeat the process over and over again, and extremely rapidly, to generate enough energy to power the grid.

The hydrogen pellets are expensive and time-consuming to make. Some companies are exploring the use of alternative fuels to deuterium, which is widely available in seawater, and tritium, which is more scarce and radioactive.

TAE, for example, is developing a system that uses hydrogen-boron, which is relatively easy to procure and not radioactive. BLOOMBERG