The Broken Nuclear Umbrella

For decades, American allies in Asia and Europe have relied on U.S. extended nuclear deterrence for their safety. They forswore acquiring their own nuclear weapons, agreeing instead to live under the protection of the United States’ nuclear umbrella. This arrangement worked during the Cold War because the stakes of that competition were so high that the United States could credibly—if just barely—say that it would wage nuclear war, even at catastrophic risk to itself, to prevent the conquest of its allies.

But the world that created the U.S. extended deterrence system is long gone. The United States retains important interests in Europe, but Russia’s threat is far more limited than when the Soviet Union was poised to conquer much of Europe and Asia. A war today would be about the precise boundaries of Moscow’s influence in eastern Europe, not the global balance of power. Similarly, the Korean Peninsula was once a battleground for rival Cold War blocs; today, that conflict is largely local, with Pyongyang desperately seeking to hold on to its power. And the Trump administration’s view that the purpose of alliances is first and foremost to enhance the United States’ core national interests is at odds with promises to jeopardize American cities to defend allies. In a world in which wars are largely regional and Washington is pulling back, it strains credulity to believe that the United States would wage nuclear war for faraway allies.

U.S. allies are thus increasingly worried that the extended deterrence system is now a bluff. Their fears are justified. Leaders in Latvia, Poland, and South Korea cannot be confident that the United States would risk its own destruction for their safety. Those in North Korea and Russia may be drawing the same conclusion. If Washington’s nuclear commitments are no longer credible, U.S. allies need new arrangements to ensure their most critical security needs.

The crisis of extended nuclear deterrence has developed gradually since the end of the Cold War. When allies have questioned the credibility of American promises, Washington has sought to soothe their concerns. The United States issues official statements declaring its commitment to its partners and conducts military demonstrations, such as submarine port visits or bomber overflights, to underscore U.S. power and commitment. But it’s dangerous to try to convince allies that all is well when it’s obviously not. Refusing to address the weakness of the existing extended deterrence strategy could embolden adversaries to attack. And fearing for their safety, allies in Asia and Europe could rush to cobble together their own nuclear programs, which could in turn trigger preventive strikes and the rapid escalation of conflict.

The United States must accept that it can no longer be a credible nuclear guarantor for all of its allies around the world. It needs to get comfortable with its partners seeking solutions more suited for today’s geopolitical realities, including, in some cases, acquiring their own nuclear weapons. In Asia, this could mean establishing a nuclear sharing agreement with South Korea, modeled on the one between Washington and NATO, in which the United States would promise to transfer control of tactical nuclear weapons to Seoul if North Korea were to attack. Or if Seoul ultimately chooses to withdraw from the Nuclear Nonproliferation Treaty to build its own nuclear arsenal, Washington should support this decision and use its influence to insulate South Korea from the dangers of such a transition. In Europe, the best option for the United States and its partners is to modify the existing NATO nuclear sharing program to increase European control over the alliance’s collective deterrent. France and the United Kingdom—the two other nuclear-armed NATO states—could enhance their role in NATO’s nuclear plans to reduce Europe’s dependence on Washington and better protect the bloc’s most vulnerable members.

Helping U.S. partners develop stronger and more autonomous deterrence postures is anathema to many in Washington, who fear nuclear proliferation and do not want to give up the United States’ leverage as a global security provider. But this may not be Washington’s decision. U.S. allies are vulnerable, and they are right to take concrete steps to reduce the real dangers they face, whether that means developing their own nuclear deterrent or making other moves that violate outdated U.S. prohibitions. Washington should support them along whichever path they choose for protection.

CREDIBLE ARRANGEMENTS

The U.S. extended deterrence strategy grew out of two conflicting Cold War imperatives. The United States recognized that nuclear weapons were essential for preventing Soviet attacks against American allies around the globe. Yet Washington and its partners did not want scores of countries to have their own weapons because they feared a world of competing nuclear-armed states. The solution was a system of extended nuclear deterrence in which the United States opened its nuclear umbrella over its most important allies in Asia and Europe. Although France and the United Kingdom decided to acquire their own arsenals, American nuclear weapons served as the strategic deterrent for the rest of the U.S. alliance system during and after the Cold War.

From its inception, this strategy had credibility problems because it seemed far-fetched that the United States would risk nuclear catastrophe to defend its allies. Indeed, in 1961, French President Charles de Gaulle scoffed at the notion that the United States would sacrifice New York City to defend Paris. If anything, de Gaulle greatly understated the extent of the challenge: U.S. policy effectively required Washington to put every major American city at risk to protect its allies.

To enhance the credibility of extended deterrence, the United States devised an array of military actions and diplomatic maneuvers. Washington deployed troops and tactical nuclear weapons near the border between East and West Germany and just south of the dividing line between North and South Korea. By stationing nuclear weapons directly in the path of a potential invasion, where they would surely become entangled in any fighting, the United States made it clear that a major attack risked catastrophic escalation.

Leaders of both U.S. political parties buttressed those deployments with rhetoric that signaled the inseparability of American and allied interests. In 1963, when President John F. Kennedy, a Democrat, stood in Berlin and declared, “Ich bin ein Berliner,” he was emphasizing that Washington and its key NATO allies were one. More than two decades later, President Ronald Reagan, a Republican, struck the same note when he told a European audience, “An attack on you is an attack on us. . . . An attack on Munich is the same as an attack on Chicago.”

Anxiety about America’s commitment is palpable.

Underlying these military deployments and this diplomatic rhetoric was a deeper truth: the geopolitical stakes during the Cold War were so great that allies and adversaries alike could believe the extraordinary threat of waging nuclear war. A successful Soviet invasion of Western Europe would have given Moscow control of the territory stretching from the Pacific Ocean to the English Channel—an outcome that American strategists believed posed a grave threat to U.S. security. North Korean advances, too, could be construed as a major victory for the Soviet-led bloc. In that context, it was plausible that the United States would risk nuclear Armageddon to hold the Soviet Union and its allies at bay.

The collapse of the Soviet Union in the early 1990s began to erode the credibility of the U.S. extended nuclear deterrent. Facing regional security challenges, rather than near-existential threats, the United States became less inclined to risk nuclear war on behalf of its partners. But few allies paid attention. In Europe, for instance, nuclear weapons were seen as a legacy problem from the Cold War until renewed Russian aggression reminded countries on the continent of their need for a stronger deterrent. Moscow’s threat to use nuclear weapons against NATO countries after it invaded Ukraine, in 2022, was a particularly jarring lesson.

The recalibration of American foreign policy, which has played out over two Trump administrations, has further driven allies to question the reliability of the United States’ nuclear promises. The essence of President Donald Trump’s “America first” policy is to assess alliances in terms of how much they contribute to U.S. national security. The administration’s 2025 National Security Strategy, for instance, declared that U.S. policy is now aimed singularly at promoting core American interests. A pledge to put “America first” directly contradicts a pledge to wage nuclear war on behalf of allies.

Countries around the world may hope that whoever is elected president in 2028 will recommit Washington to supporting their defense. But regardless of who sits in the Oval Office, it is no longer sensible for the United States to wage a major nuclear war to help far-flung allies enmeshed in regional conflicts. Trump’s reorientation of U.S. foreign policy has also shattered any illusions that Washington is permanently committed to its allies’ defense. It was always a risk for allies to rest their core security needs on American promises of nuclear war, even when a bipartisan consensus of U.S. leaders publicly supported such a position. With Washington divided over the direction of its foreign policy, allies would be taking an extraordinary gamble to continue to rely on U.S. extended nuclear deterrence.

Indeed, this system is under increasing strain worldwide. Rising tensions in the Taiwan Strait, for instance, have raised concerns in Japan about being caught in a nuclear crossfire if Washington and Beijing were to come to blows, and in Australia about being drawn into the fighting if it were to allow the United States to use its bases in a potential war. But nowhere are extended deterrence concerns as urgent as on the Korean Peninsula and in eastern Europe, the frontlines of the two theaters where aggressive, nuclear-armed adversaries are threatening escalation.

SEOUL FLYING SOLO

Since North Korea tested its first nuclear device in 2006, Pyongyang has amassed between 50 and 90 nuclear weapons, and its arsenal continues to grow. The North Korean government has ordered its armed forces to integrate nuclear forces into the country’s conventional war plans, which implies that any major war on the peninsula will become nuclear. Technological advances are making North Korea even more dangerous. In 2017, North Korea tested a device that was an order of magnitude more destructive than its previous atomic bombs, suggesting that the country is on the path to developing potent fusion weapons capable of destroying entire metropolitan areas in a single blast. And progress toward intercontinental nuclear delivery systems will soon put American cities directly in Pyongyang’s cross hairs. If North Korea’s nuclear weapons can target American cities, U.S. policymakers will think twice before using nuclear forces on behalf of South Korea.

Meanwhile, the United States has gradually dismantled the arrangements that once bolstered the credibility of extended deterrence on the peninsula. Washington has reduced the number of troops stationed there and moved them farther south, away from the demilitarized zone dividing the country from North Korea. And in 1991, when the Cold War ended, Washington removed its tactical nuclear weapons from South Korea.

The Trump administration’s rhetoric and policies have further fanned South Korean fears of abandonment. Trump has adopted hardball trade policies toward Seoul and denigrated the value of the alliance, complaining that the United States has been protecting South Korea “free of charge.” In 2025, U.S. Immigration and Customs Enforcement officers raided a Hyundai-LG battery plant under construction in Georgia and detained hundreds of South Korean engineers, many of whom were in the United States on business visas. Images of their shackled compatriots enraged South Koreans. Most recently, Trump’s decision during the Iran war to move U.S. missile defense batteries from South Korea to the Persian Gulf, over Seoul’s objections, exemplified South Korea’s growing concern that the United States would put its own security over the needs of its ally. Extended nuclear deterrence requires the opposite: a willingness to sacrifice dearly on behalf of one’s partners.

As conditions have eroded, South Korean strategists have debated three main options for strengthening nuclear deterrence. Seoul could ask the United States to redeploy tactical nuclear weapons to South Korea. Alternatively, Washington and Seoul could create a nuclear sharing system, which would mirror the arrangement between the United States and its NATO allies: the United States would station nuclear bombs at U.S. military bases in South Korea and promise to transfer control of those weapons to Seoul in the event of a crisis or war. Or in what would mark a more significant shift, South Korea could exercise its legal right to withdraw from the Nuclear Nonproliferation Treaty and build its own nuclear arsenal.

Seoul has the main elements of a nuclear arsenal—minus the weapons.

So far, South Koreans have been committed to the principle of nonproliferation, and many South Korean security experts still reject nuclearization because it risks economic sanctions, diplomatic isolation, or even a North Korean preventive attack. But given the uncertain credibility of the United States, leaders in Seoul may soon decide that they cannot gamble their country’s survival on the promises of a wavering ally. South Korea may accept a nuclear sharing agreement as enough protection, but it is increasingly possible that Seoul may feel it needs an independent nuclear deterrent. Numerous polls over the past decade have revealed that a supermajority of South Korean citizens support such a move.

Seoul has been laying the groundwork for an independent nuclear capability for decades. It has already built the delivery systems needed for a secure and effective nuclear retaliatory force. South Korea is the only nonnuclear country in the world to possess ballistic missile submarines; it owns three of them and is building three more. Additionally, its air force operates a large and growing fleet of modern fighter aircraft, dispersed across a dozen military airfields and protected by hundreds of hardened shelters—the ideal infrastructure for an air-delivered nuclear deterrent. And South Korea operates dozens of mobile land-based missiles. In short, it has already paid for and built the main elements of a nuclear arsenal minus the weapons themselves.

U.S. policymakers have long opposed an independent South Korean arsenal and instead sought to reassure Seoul that the existing system remains strong. In 2023, for instance, the Biden administration responded to questions from Seoul about the reliability of the U.S. deterrent by hosting the South Korean president at a summit, reiterating the United States’ commitment to South Korea’s defense, and sending a nuclear missile submarine to visit the South Korean port of Busan. U.S. officials and foreign policy elites have also tried to dissuade Seoul from plotting an independent course by noting the consequences required by U.S. law for countries that seek nuclear weapons, including automatic sanctions and the loss of access to civilian nuclear fuel and technology. But as American credibility wanes, South Korea’s growing vulnerability may lead Seoul to proceed anyway. The result would be a higher risk of war during the period in which Seoul tries to develop nuclear weapons because North Korea would interpret such a move as a threat and might attack.

If South Korea requests a nuclear sharing arrangement, the United States should welcome that solution. This would enhance deterrence on the Korean Peninsula without forcing Washington to deal with the diplomatic and strategic challenges that would arise from an ally acquiring nuclear weapons. Alternatively, if South Korea decides it needs its own nuclear weapons, the United States should help Seoul acquire them safely. Washington could strengthen extended deterrence during the transition period in which South Korea nuclearizes by temporarily positioning tactical nuclear weapons on the peninsula and redeploying American troops closer to the demilitarized zone. The United States could agree to staunchly defend Seoul in international forums in which sanctions might be debated. (There is a legitimate basis for South Korea’s need for nuclear weapons because the threat it faces comes from a North Korean arsenal that was obtained by flouting numerous UN Security Council resolutions.) Washington could also waive the sanctions written into U.S. law, similar to what it did for India and Pakistan in 1998–99.

If U.S. leaders acknowledge the real deterrence challenges that have developed on the Korean Peninsula and help South Korea do what it needs to protect itself, the United States can emerge with a stronger alliance—one that resembles the close partnerships it maintains with nuclear-armed countries such as France, Israel, and the United Kingdom.

SPARE KEYS

In the past two decades, a reenergized Russia has invaded Georgia, seized Crimea, and launched a full-scale attack on Ukraine. But despite Moscow’s renewed aggression, the Trump administration has stepped back from its NATO partners, arguing that Europe can defend itself from a Russia that is much weaker than its Soviet predecessor. Furthermore, in stark contrast to the Kennedy and Reagan declarations that the United States and its allies were firmly on the same team, the Trump administration has highlighted the areas in which U.S. and allied interests diverge. It has criticized NATO partners for free-riding on U.S. military spending, rebuked European governments for what the White House deems to be repressive restrictions on free speech, and threatened to seize Greenland, which is administered by Denmark, a NATO ally. And in early May, Trump responded to the lack of European participation in the war against Iran, as well as German Chancellor Friedrich Merz’s criticism of the conflict, by withdrawing 5,000 troops from Germany.

U.S. nuclear weapons remain the cornerstone of NATO’s nuclear deterrent. But given the major change in stakes from the Cold War to the present and the shifting political landscape in the United States, European governments can no longer assume that American nuclear promises are ironclad. Anxiety about the United States’ commitment is palpable. European countries, particularly those in eastern Europe, have started to discuss how to strengthen NATO’s nuclear deterrent.

For vulnerable countries in eastern Europe, pursuing independent nuclear capabilities is unlikely to be the best solution. Even Poland, the most militarily capable country in the region, would encounter significant hurdles to developing a secure nuclear arsenal. Warsaw lacks the delivery systems that are the foundation of a strong and reliable nuclear deterrent. Poland has no modern submarines; it is negotiating to buy some from Sweden, but these would be primarily for coastal patrol, cannot carry large numbers of missiles, and are many years from operational deployment. Poland does own a few dozen land-based mobile missile launchers, but they can scarcely reach halfway to St. Petersburg, never mind Moscow. The most up-to-date branch of Poland’s military is its air force, which is integrating F-35 fighter jets into its force structure. But relying on air-delivered nuclear weapons to deter an attack is dangerous for a country so close to Russia. The Polish arsenal would be vulnerable to a Russian missile barrage on its airfields, which could leave Warsaw without the ability to retaliate.

Poland also lacks the building blocks to quickly create the fissile material needed for nuclear weapons. Compared with Seoul’s civilian nuclear program, Warsaw’s is small, and Poland has neither large stockpiles of nuclear spent fuel nor the facilities needed to reprocess it and build weapons. Over time, Poland could build a capable nuclear arsenal if it decided to do so. But the pieces are not yet in place. And these significant hurdles are even higher for the other vulnerable eastern European members of NATO, which have less money, worse delivery systems, and smaller territory on which to deploy and protect their weapons.

Rather than encourage member states to pursue independent nuclear arsenals, NATO could modify its existing nuclear sharing plans to enhance deterrence. Under the current arrangement, the United States positions a small number of nuclear bombs in Europe in peacetime. Those weapons remain under exclusive American control: they are locked, and only the U.S. president can unlock them. In the event of a serious nuclear crisis, NATO could vote to disperse these weapons, and the United States would—at least in theory—release them to European air forces operating under NATO command.

One shortcoming of the current strategy is that NATO may be unwilling to wage nuclear war to protect the alliance’s eastern members. The United States and its NATO partners could address this problem by placing some of the alliance’s nuclear forces where they would be entangled in any fighting if Russia were to attack an eastern European ally such as Poland. Warsaw, for instance, could build modern nuclear storage sites at several of its airfields, either as locations to hold NATO nuclear weapons in peacetime or as places to which they could be dispersed during a crisis. Deploying nuclear weapons on Polish territory would strengthen deterrence by making Russia confront the danger that an invasion of Poland would be met with a nuclear defense.

But the existing nuclear sharing arrangements face an even more fundamental challenge: a single person—the U.S. president—can veto the transfer of the weapons. NATO nuclear sharing in its current form is therefore only as credible as the United States’ commitment to its European allies.

The best solution lies in expanding the list of countries that hold the keys to NATO’s nuclear arsenal. In a modified arrangement, France and the United Kingdom could each contribute about two dozen nuclear weapons to the alliance’s stockpile. Those weapons, which would need to be suitable for delivery by NATO aircraft, could be deposited into existing NATO bunkers that are spread across Europe. In peacetime, these additional French and British weapons would remain under the full control of Paris and London, respectively, just as U.S. weapons are under Washington’s control, but they could be released to other members in times of crisis.

NATO should increase European control over its collective deterrent.

This updated sharing arrangement would retain the logic of the existing plan but convert NATO’s deterrence from a “one key” to a “three key” system. Currently, NATO’s nuclear force would evaporate if a single person, the U.S. president, were to withhold the key that unlocks the bombs. Under a three-key system, the deterrent would be operational if any of the three leaders—of France, the United Kingdom, or the United States—were to give the go-ahead.

A three-key arrangement would comply with the Nuclear Nonproliferation Treaty because France, the United Kingdom, and the United States would maintain peacetime control over their weapons. France already owns air-launched cruise missiles suitable for inclusion in the sharing arrangement; to fulfill these new obligations, it would need to build only two dozen more nuclear-tipped missiles and share the software and physical connectors that NATO aircraft would need to employ them. Because the United Kingdom has only submarine-launched nuclear weapons, it would need to build bombs suitable for air delivery, which is a prerequisite for use in the NATO arsenal. But these could be lightly modernized versions of the nuclear gravity bombs that it deployed until the late 1990s.

A revised system along these lines is within reach. France and the United Kingdom have the manufacturing facilities and the know-how to build these weapons; constructing additional warheads would mainly require hiring extra workers, running more shifts, and perhaps delaying some routine warhead maintenance because the countries’ plutonium-handling facilities would be manufacturing new weapons. France is already expanding its arsenal of nuclear cruise missiles, although it has thus far declined to allocate any weapons to an enhanced NATO sharing program. An improved sharing arrangement would work best if both France and the United Kingdom contributed weapons to the NATO stockpile, but a contribution from either country would bolster deterrence relative to the current arrangement.

LIKE IT OR NOT

For Washington, the easiest move is to stick with the existing system of global extended deterrence. This leaves allies dependent on the United States for their security, maximizes American influence in the world’s geopolitical hotspots, and—at least in the short term—limits the proliferation of nuclear weapons. It also avoids generating new tensions: after all, U.S. adversaries would cast any changes to the status quo as provocative threats and fiercely oppose them.

But continuing with the extended deterrence system courts catastrophe for both Washington and its partners. A security strategy based on a bluff will feel like a bargain until the moment it is challenged. It is better to bolster deterrence today before a military crisis erupts.

Even if Washington doesn’t want to change course, it might not have a choice. U.S. allies may simply reject the U.S. extended deterrent and pursue their own alternatives. South Korea may decide that it must take responsibility for its own national security, and Europeans may conclude that something must be done to enhance nuclear deterrence in eastern Europe. Washington’s best option is to help them develop their own approaches to deterrence. Supporting allies as they do so would reduce the risk of war during the transition to a new deterrence status quo, and it would increase the odds of closer and more productive relationships with them afterward.

The global system of extended nuclear deterrence was an audacious strategy that was built for the unique circumstances and stakes of the Cold War. That era is over. The United States needs to refashion deterrence strategies for the world not as it once was, but as it is.