As missile threats grow more numerous, diverse and technologically sophisticated, a recent US Senate hearing has exposed mounting concerns that the US’s homeland missile defenses are increasingly misaligned with the realities of modern warfare.
Assistant Secretary of Defense Marc Berkowitz delivered a stark assessment of current homeland capabilities, stating that the US relies on a “very limited” ground-based, single-layer defense system designed specifically to counter a small-scale rogue intercontinental ballistic missile (ICBM) attack from North Korea.
He emphasized that this architecture provides only “very limited capability” against other ballistic missile threats. Most critically, Berkowitz warned that the US currently has “no defense against hypersonic weapons or cruise missiles today,” later clarifying he was referring to advanced cruise missiles.
The Ground-Based Midcourse Defense (GMD) is currently the only system capable of defending the US homeland against ICBM attack. It utilizes integrated communication networks, fire control systems, globally distributed sensors, and Ground-Based Interceptors (GBIs) that can detect, track and neutralize ballistic missile threats.
However, the GMD’s performance may leave much to be desired. Data from the Missile Defense Advocacy Alliance (MDAA) indicate that between 1999 and 2023, 21 GMD tests were conducted, resulting in 12 hits and 8 misses, with a success rate of 57%.
Delving further into the limitations of the GMD system, Robert Peters notes in a March 2026 Heritage Foundation report that the US fields 44 GBIs for the GMD system. Peters points out that the GMD is limited both numerically and technologically, making it highly constrained in defending against more sophisticated or multi-missile attacks, including those with multiple warheads or advanced countermeasures.
A US Defense Intelligence Agency (DIA) assessment from March 2025 shows that traditional ICBMs remain the primary threat to the US homeland, with China fielding about 400, Russia about 350 and North Korea 10 or fewer, due to both their numbers and their ability to strike any part of the US. It notes that submarine-launched ballistic missiles (SLBMs) pose an additional risk, as no part of the US is beyond their reach.
The DIA assessment also indicates that hypersonic weapons — estimated at around 600 for China and 200 to 300 for Russia — complicate interception due to their speed and maneuverability, while land-attack cruise missiles (LACMs), estimated at 1,000 for China and 300 to 600 for Russia, exploit low-altitude flight and reduced signatures to evade detection.
It further highlights emerging systems such as fractional orbital bombardment systems (FOBS), which can approach from unexpected trajectories, including over the South Pole, to bypass early warning systems.
Together, these capabilities illustrate the scale and diversity of threats that current US missile defenses, such as the GMD, were not designed to handle.
Against such threats, US homeland missile defenses remain severely constrained. Using an ICBM attack as a performance indicator, a February 2025 American Physical Society (APS) report notes that, despite more than US$400 billion spent since 1957, no missile defense system is effective against realistic ICBM threats.
The report states that intercepting even a single nuclear-armed ICBM or its warhead in flight is extremely challenging, given short engagement windows and the difficulty of distinguishing warheads from decoys and other objects in space.
It adds that US missile defense systems are designed to address limited threats, such as small-scale attacks, and would face significant difficulties against more complex or larger-scale attacks.
The report emphasizes that missile defense must be nearly perfect to be effective, since even one nuclear warhead penetrating defenses would be catastrophic, yet current capabilities remain low and are likely to remain so for at least 15 years.
Taken together, these findings suggest that US missile defenses are structurally constrained not just by scale, but by fundamental technical limits. The Trump administration’s “Golden Dome” missile defense project aims to address these gaps but faces significant technical, cost and feasibility questions.
Golden Dome is a US initiative to build a layered “system of systems” integrating space-, air-, ground- and sea-based missile defenses to defend the US homeland against ballistic, hypersonic and cruise missiles, as well as other aerial threats.
However, Jeff Hecht writes in an article this month for Bulletin of the Atomic Scientists that Golden Dome faces severe timing constraints, with sensors unable to confirm an ICBM trajectory until about 75 seconds after launch, leaving only 25–35 seconds to decide and engage before an interceptor must be launched, while decoys and multiple warheads further complicate interception.
He adds that the system faces major feasibility challenges, noting that estimates suggest roughly 40,000 space-based interceptors would be required to counter even a limited salvo of 10 ICBMs, with satellites needing replacement every about 5 years due to orbital decay.
He also highlights the system’s potentially enormous costs, citing US estimates rising to about US$185 billion for initial deployment and analyses suggesting total costs could reach US$3.6 trillion over 20 years, raising serious doubts about affordability and practicality.
Against that backdrop, Jeffrey Lewis argues in a May 2025 Scientific American article that Golden Dome is “fantasy” rooted in the belief that the US can buy its way out of nuclear vulnerability. He argues that while nuclear vulnerability is difficult to accept, it is the logic of mutually assured destruction (MAD) that has prevented nuclear war.
Despite those criticisms, Kari Bingen argues in a January 2026 Center for Strategic and International Studies (CSIS) commentary that the Golden Dome can be justified because the threat environment has changed profoundly, leaving the US homeland increasingly vulnerable to a wider range of large-scale, diverse and coordinated missile and aerial threats.
Bingen contends that Golden Dome is not about perfect protection but about limiting damage, complicating adversary planning and raising the threshold for attack, thereby strengthening deterrence.
She adds that by integrating existing systems into a layered architecture, the system could provide policymakers with more response options short of nuclear retaliation, making it a necessary evolution rather than an unrealistic pursuit of invulnerability.
Going forward, US homeland missile defense will be defined less by its ability to stop attacks outright than by how effectively it reshapes adversary calculations in an increasingly complex and contested threat environment.
Yet the evidence points to a harsher reality: a defense architecture that must be nearly perfect to work is inherently vulnerable to failure, especially when faced with larger or more sophisticated salvos.
As even relatively unsophisticated missiles and drones have shown in the Iran war, US and Israeli defenses can be saturated and penetrated when attacked in sufficient numbers. At its core, this is a problem of physics and statistics — interceptor-based systems are finite, costly and vulnerable to being overwhelmed, while attackers can scale more cheaply.
Absent a shift toward fundamentally different technologies, such as directed-energy weapons, the current US approach to missile defense remains, in essence, a losing proposition — one that can mitigate risk but not eliminate it.
