Fusion energy has a reputation for being perpetually just over the horizon. For decades, the joke among physicists was that practical fusion power was always thirty years away, and always would be. It is the process that powers the sun and every star, and if we could harness it on Earth it would offer almost limitless clean energy with no carbon emissions, no risk of meltdown, and little long-lived radioactive waste. That prize has stayed tantalizingly out of reach. Yet something has shifted. The past two years have brought a run of genuine milestones, a flood of private money, and a growing sense that the horizon may finally be moving closer.

The Problem of Making a Star

Fusion works by forcing light atomic nuclei together so they merge and release energy. Doing this requires conditions similar to the core of a star: temperatures of many millions of degrees, at which matter becomes a turbulent, electrically charged gas called plasma. Two enormous challenges have always stood in the way. First, you must confine and control that plasma long enough for reactions to occur. Second, you must get more energy out of the process than you put in to start it, a threshold known as net energy gain.

For most of fusion’s history, both problems seemed intractable. The plasma is fiendishly unstable, and the energy required to heat and contain it kept swamping whatever the reactions produced. Progress was real but painfully slow, which is how fusion earned its reputation as a technology forever in the future.

Two Roads to Ignition

Researchers have pursued fusion mainly along two paths. One uses powerful lasers to compress a tiny fuel pellet in an instant, an approach called inertial confinement. The other uses gigantic magnets to trap plasma inside a doughnut-shaped chamber known as a tokamak, an approach called magnetic confinement.

The laser route produced a landmark result in 2022, when the National Ignition Facility in California achieved net energy gain for the first time, releasing more energy from the reaction than the lasers delivered to the target. In the years since, the facility has continued to push that output higher. The magnetic route has racked up its own records, particularly in China, where an experimental tokamak nicknamed the “Artificial Sun” sustained a superheated plasma for more than a thousand seconds, roughly seventeen minutes, a striking demonstration of stability at extreme conditions.

The Rise of Private Fusion

Perhaps the biggest change is who is doing the work. Fusion was long the domain of government laboratories and vast international collaborations, above all ITER, a project involving dozens of nations that is building the world’s largest tokamak in France. That public effort continues, but it now shares the stage with a swarm of startups. More than forty private fusion companies are pursuing the goal, and private investment in the sector had surged past several billion dollars, drawing strategic backing from major energy corporations.

One of the most watched is Helion, a company that has taken an unconventional approach and moved quickly. Its prototype became the first privately developed fusion machine to demonstrate measurable fusion using deuterium and tritium and to reach plasma temperatures of around 150 million degrees Celsius. The company has already begun building its first commercial machine, with the stated aim of delivering electricity from fusion to the grid for a major technology customer. Others, including firms working on advanced magnets and alternative reactor designs, are targeting commercial plants in the early 2030s.

The Unsung Hero: Magnets

Much of the recent momentum comes not from plasma physics alone but from a quieter revolution in engineering. Magnetic confinement depends on the strongest magnets ever built, and advances in high-temperature superconducting materials have transformed what is possible. These new magnets can generate far more powerful fields in a smaller, cheaper package, which changes the economics of building a reactor.

In 2025, engineers reported record-breaking results from superconducting magnet systems, and one company demonstrated a superconducting coil under conditions replicating the intense magnetic environment inside a real commercial reactor. These may sound like technical footnotes, but they are the kind of practical breakthroughs that turn a physics experiment into a machine that could actually run a power plant.

A Realistic Look at the Horizon

None of this means fusion has arrived. Demonstrating net gain in a single laser shot is very different from running a plant that produces steady, affordable electricity for years. Enormous challenges remain in materials that can withstand the punishing conditions inside a reactor, in maintaining plasma stability at scale, and in driving down costs. Industry analysts project that the first commercial fusion plants could begin operating sometime between 2030 and 2035, but such timelines have slipped before, and caution is warranted.

Still, the trajectory is unmistakable. Fusion has moved from a distant theoretical dream to a field with concrete records, serious commercial ambition, and billions of dollars behind it. Governments and private companies alike are now betting that the technology can be delivered within a couple of decades rather than a couple of generations. The old joke about fusion always being thirty years away may not be dead yet. But for the first time in a long time, it is starting to sound out of date.