Golden Dome is Donald Trump's proposed missile-defence shield for America, an allusion to Israel's more modest Iron Dome. It is intended to protect America from attack using, among other things, hundreds or thousands of satellites that can both track and attack enemy missiles as they launch.
Trump said on May 20th 2025 that his tax bill included $25bn in initial funding and that the project would cost $175bn in total. The Congressional Budget Office reckons the bill could run to more than $500bn over 20 years. Trump's stated timeline of "two and a half to three years" is widely considered unrealistic. Experts calculate that a modest Golden Dome focused on parrying small incoming salvos might cost just over $250bn over 20 years, but a full version with tens of thousands of space-based interceptors (SBIs) in orbit could run to $3.6trn—a sum that would cannibalise America's armed forces.
Trump has claimed the system will offer "close to 100% protection". A report by the American Physical Society suggested that 16,000 space-based missiles would be needed to intercept a salvo of just ten North Korean Hwasong-18 missiles. If American leaders wanted 30 seconds of decision time before acting, they would need 36,000 interceptors. Defending very northerly cities, Alaska or the Midwest would require still more.
America's existing Ground-based Midcourse Defence (GMD) missiles in Alaska and California have a mixed record: of 21 live-fire tests conducted since 1999, nine have failed. The system has improved over time, however, with four consecutive successes from the most recent test in 2023. A completely impenetrable shield is almost certainly impossible, but advocates argue there is a difference between losing two American cities and 20.
The Missile Defence Agency requested proposals from firms in September 2025 and received so many that it pushed the deadline back a week. The competition pits established "primes"—Lockheed Martin, L3 Harris—against newer Silicon Valley-style defence firms: SpaceX, Anduril and Palantir. The likely outcome is a compromise: primes building radars, interceptors and launchers, while the newer firms focus on the command-and-control software. Lockheed Martin plans to test a space-based interceptor in orbit by 2028; Anduril is said to be mulling space-based lasers.
Golden Dome's most plausible use may be as part of America's long-standing "damage limitation" strategy: if a pre-emptive counterforce strike destroys most of an adversary's nuclear weapons before launch, a missile-defence shield need only mop up the residual salvo. Vipin Narang, a political scientist at MIT who led nuclear-weapons policy at the Pentagon during the Biden administration, argues the system makes most sense as "a limited system that is greater than and more flexible than what we have now" but still sufficient to handle a residual Chinese or Russian force. Critics such as James Acton of the Carnegie Endowment warn this could generate an arms race in which America wastes enormous sums failing to achieve its objective.
The shield responds to the Pentagon's concern that America's adversaries are building large numbers of new and more diverse missiles. American radars and defences have historically focused on missiles travelling over the North Pole, but long-range hypersonic missiles and "fractional orbital" systems that can partly encircle the Earth take more unpredictable routes. Canada, which already has a joint aerospace-defence command with America, is in talks about joining Golden Dome.
The same space-based interceptors designed to destroy enemy missiles could in theory also target enemy satellites. General Stephen Whiting of US Space Command has described the need for "space fires" and "orbital interceptors" across space-to-space, space-to-ground and ground-to-space domains.
Golden Dome depends on a network of sensors around the world and in space. America's early-warning station at Pituffik in Greenland is an important node, both for tracking missiles and communicating with satellites. RAF Fylingdales in the north of England is another key site. The system's dependence on allied territory means that Donald Trump's pursuit of Greenland risked alienating the very allies whose co-operation the project requires.
The concept echoes Ronald Reagan's Strategic Defense Initiative. Ronald Reagan and George H.W. Bush both toyed with space-based interceptors as part of their missile-defence initiatives in the 1980s, though the scheme fizzled out. Trump has said that "Ronald Reagan wanted it many years ago, but they didn't have the technology." General Michael Guetlein was appointed as the head of Golden Dome.
The project's scope remains dangerously vague. It is unclear whether Golden Dome is meant to parry small salvos from a second-tier nuclear power like North Korea, or to block hundreds of Russian and Chinese warheads in a full-blown nuclear exchange. A degree of mutual vulnerability is an inherent part of stable nuclear deterrence; an over-ambitious shield could induce adversaries to expand their own arsenals.
Three may keep a secret, if two of them are dead.