Nuclear Fusion in 5 Years? What is Happening?

Fusion hype is colliding with financial reality: multiple companies and governments are accelerating fusion timelines and funding, yet there’s still little public evidence of reliable fusion performance or net energy gain. That gap—between bold claims like “five years” to grid power and the current state of experimental results—drives the skepticism at the center of this update.

Commonwealth Fusion Systems says it has sold more than $1 billion worth of power to Eni, even though the reactor doing the work isn’t operating yet. The contrast is stark: political and corporate announcements are moving faster than the underlying technology. US Energy Secretary Chris Wright has repeatedly suggested fusion power is about five years away, citing advances in artificial intelligence and progress at national labs and private companies. He also frames grid availability as an “eight to 15 years” prospect. The problem, according to the assessment here, is that the technology capable of delivering that outcome doesn’t yet exist in a demonstrated, dependable form.

Helion, backed by Microsoft and Sam Altman, is building in Washington using a hybrid concept combining magnetic confinement and inertial confinement. The approach involves shooting fuel beams at each other and then squeezing them in a magnetic field—likened to a particle collider, which raises concerns about energy efficiency. Helion’s target is ambitious: by 2028 it aims to deliver up to 50 megawatts to a Microsoft data centre. But the skepticism is rooted in what’s missing publicly: no clear evidence of reliably achieving fusion in the first place, let alone achieving net gain.

Government spending is also ramping up. In June, the British government invested more than $3 billion into a Tokamak. Germany’s fusion push is likewise expanding after Proxima Fusion’s stellarator plans, with Marvel Fusion raising over $380 million for a laser-ignited fusion plant. The underlying pattern is that fusion is being used as a climate and industrial promise—an “escape hatch” for policymakers—while also being positioned as a way to justify energy demand tied to AI.

Against that backdrop, First Light Fusion has shifted its strategy. Instead of its earlier “Big Friendly Gun” concept that fired solid projectiles at a fuel pellet, it now proposes FLARE, a two-step inertial confinement plan: slow compression via an electrical pulse, followed by rapid ignition with a short laser pulse. The system uses a large pool of liquid lithium to absorb heat, generate power, and breed tritium. First Light Fusion claims an energy gain of 200 to 1,000, but the timeline is still vague—“potentially by the mid-thirties”—and the critique is that companies aren’t always transparent enough about changing plans.

The bottom line is caution. Fusion may ultimately become the power source of the future, but the next five years are expected to be dominated by failures: many firms could run out of money before reaching breakeven. The most promising directions, based on current signals, are described as stellarators and laser-ignited fusion. Even then, progress is measured in incremental experimental milestones—such as NIF reporting a record shot with more than four times the energy out of the pellet than into it—where headlines may look dramatic while the path to practical, net energy remains long and uncertain.