U.S. Nuclear Deterrence Faces a Two-Superpower Stress Test
Diego Garcia
David Jones
Matthew Kroenig
Asle Toje
Jane HarmanThe Aspen InstituteSunday, June 28, 202621 min readAt the Aspen Ideas Festival, Gillian Tett led a panel on whether mutually assured destruction can still stabilize a nuclear order no longer defined by a U.S.-Soviet rivalry. Christine Wormuth argued that the collapse of arms-control limits and China’s buildup have made this the most dangerous period of her national security career; Matthew Kroenig said the United States needs more nuclear capacity to deter Russia and China at once; and Jane Harman warned that weapons modernization without political strategy could deepen the risks it is meant to contain.

The same deterrent that reassures allies can accelerate the arms race
The central warning was not that nuclear weapons have returned after disappearing. It was that the habits, institutions, and assumptions built for the Cold War are being asked to carry a more crowded and less stable nuclear order. The same measures that can reassure allies and deter Russia and China — larger arsenals, forward deployments, flexible response options, missile defenses — can also increase costs, fuel arms racing, and make crises harder to control.
? gillian-tett framed the problem starkly: 80 years after Hiroshima and Nagasaki, the world has avoided another nuclear war, in large part because of what is often called mutually assured destruction. She also noted a major reduction in the total number of warheads since the mid-1980s, when she said there were “apparently 70,000” warheads around the world, compared with about 12,000 today. But that reduction sits alongside a more dangerous set of claims: today’s warheads are “between 20 and 80 times more powerful” than the weapons used against Hiroshima, spending on nuclear weapons rose 19 percent last year, and nuclear spending has increased for five consecutive years.
The geopolitical list was long, and Tett presented parts of it as live political discussion rather than settled fact: debate over whether Iran is racing to get a nuclear weapon; North Korea’s indication that it wants to upgrade its naval fleet with nuclear capability; Poland’s response to U.S. threats to withdraw troops by indicating that it may have to think about nuclear weapons; Finland’s announcement, as Tett described it, that it would allow nuclear weapons to be installed on Finnish territory; and broader interest among other states in nuclear capabilities. Her point was not only that more weapons may be built. It was that nuclear weapons are being normalized again as political tools.
She used Russian television to show what that normalization looks like. A clip attributed on screen to Russian Media Monitor showed Russian broadcaster Vladimir Solovyov tossing to a weather segment. The meteorologist, Evgeny Tishkovets, did not simply discuss rain and wind. He discussed how atmospheric conditions affect the consequences of nuclear strikes.
According to the transcript, Tishkovets said weather “doesn’t affect the impact of nuclear strikes” but is “a key factor determining their consequences.” Wind, precipitation, and atmospheric stability, he said, can change radiation contamination “by tens or hundreds of times” at varying distances from the epicenter. Rain during an explosion could increase pollution “by one to three times,” and wind speed could change the radiation dose in the area after an explosion “by up to 200%.” He then said a precise weather report was “critically important” for assessing strikes against military targets in Europe “where they are making weapons against us and sending them to Ukraine,” adding: “We should carry out the strikes and escalate.”
The visual mattered because it made the rhetoric concrete: a mainstream broadcast format, with weather graphics for Moscow and a meteorologist in front of a map, was being used to discuss the operational consequences of nuclear strikes. Tett stressed that this was not fringe content. She described Solovyov’s program as roughly comparable in prominence to “Morning Joe or Tucker Carlson” and said the segment came from mainstream Russian television. In her account, cartoons of nuclear weapons flying from Moscow to London and discussions of attacks on London have become common Russian television material.
The problem, then, was whether a doctrine designed for a bipolar confrontation between two major nuclear powers can still hold in a world of multiple nuclear actors, possible proliferators, and non-state threats. Tett asked whether mutually assured destruction can function when the system no longer consists of two principal adversaries deterring one another, and whether new technologies, including artificial intelligence, may make proliferation easier or more dangerous.
Wormuth’s risk model starts with the end of limits and the rise of China
? christine-wormuth did not hedge the risk. After roughly 35 years in national security, she said she considers the current moment “the most dangerous time in my professional life.”
Her first concern was the collapse of nuclear limits between the United States and Russia. Wormuth said the United States had arms control with Russia dating back to 1972, but that it no longer has such an agreement in force. She described New START as the last arms-control agreement and said it had expired. Her substantive point was that Washington and Moscow no longer operate under agreed limits in the way they did for decades.
Her second concern was China. For many years, she said, China had a small nuclear arsenal of roughly 200 warheads. Today, she said, China has about 600, is comprehensively expanding its nuclear enterprise, is building out its own triad, and is expected to have about 1,000 warheads by 2030.
| Actor | Warheads or posture described | Why it matters in the discussion |
|---|---|---|
| United States and Russia | Wormuth referred to the United States and Russia as having about 1,500; Kroenig separately treated roughly 1,500 as the current arms-control baseline for U.S. force-sizing. | The prior Russia-centered baseline no longer answers the question of how to deter both Russia and China. |
| China | About 600 today; about 1,000 expected by 2030, according to Wormuth. | U.S. planners can no longer treat China as a lesser included case. |
| North Korea | Described by Wormuth as essentially a nuclear weapon state with a small arsenal. | Current U.S. missile defense was described as designed around limited North Korean and Iranian threats. |
For Wormuth, China’s expansion changes U.S. planning even if Russia and China do not form a formal wartime alliance. She said she does not view a true Russia-China alliance attacking the United States as highly likely. The more plausible danger, in her example, is simultaneity: Russia invades a Baltic country, and Xi Jinping decides that the moment is favorable to move against Taiwan. In that situation, she said, U.S. nuclear planners would have to consider two sets of targets, not one. That puts pressure on the size of the U.S. arsenal and has contributed to more discussion in Washington about building it up.
Wormuth’s position was not unilateral disarmament. She said she favors “a strong American credible deterrent” as the best insurance against a nuclear exchange. But she repeatedly distinguished deterrence from an unconstrained arms race. If China seeks parity, the United States builds to account for both Russia and China, and Russia then builds in response to the United States, she warned, the result is a rapid upward spiral.
She also emphasized the cost. The United States, she said, is already recapitalizing the triad: a new Columbia-class submarine, a replacement for the Minuteman intercontinental ballistic missile, and new missiles for bombers. That modernization is “very, very, very expensive,” before accounting for any arms-race expansion. Wormuth argued that the United States must ask “how much is enough,” what the opportunity costs are, and whether there is a strategy for renewed talks with Russia and China. She said she is not optimistic right now, but does not believe talks should be taken off the table.
Kroenig argues that deterrence math now points upward
Matthew Kroenig accepted the premise that the threat environment is worsening and pushed further: in his view, the United States does not have enough nuclear capacity to deter Russia and China at the same time.
His first point was that mutually assured destruction was never a foolproof safety mechanism. It worked, he argued, because it created risk. The danger that a crisis could spin out of control made leaders cautious. But leaders have also exploited that danger for advantage. He cited the Cuban Missile Crisis as an example of nuclear “chicken,” and Putin’s present threats of limited nuclear strikes as a contemporary version of the same logic.
Kroenig said the United States now faces, for the first time in its history, the need to deter two nuclear superpowers simultaneously. China’s nuclear buildup compounds that problem.
The more technical part of Kroenig’s argument concerned what U.S. nuclear strategy actually targets. He said many people misunderstand mutually assured destruction as a simple reciprocal threat against cities: Moscow and Beijing threaten Los Angeles, New York, and Aspen; Washington threatens Moscow and Beijing; everyone is deterred. But U.S. strategy, he said, has long relied on counterforce targeting. That means the United States does not deliberately target cities; it targets military assets such as ICBM silos, naval bases, air bases, and command and control.
He gave several reasons. First, adversary leaders such as Putin and Xi care about their leadership and military forces, not necessarily their populations. Second, the United States abides by the law of armed conflict, which permits attacks on legitimate military targets but not civilians. Third, counterforce targeting determines the size of the arsenal required. If the goal were simply to kill large numbers of people in Moscow and Beijing, Kroenig said, perhaps two or four nuclear weapons would be enough. If the goal is to hold at risk every Russian ICBM silo, naval base, and air base, the number rises. Add China’s expanding arsenal, and the current number no longer works.
If you want to target every Russian ICBM silo, every Russian naval base, every Russian air base, then that puts upward pressure on numbers.
Kroenig said the current rough arsenal size of 1,500 was set in arms control with Russia in 2010. The world has changed since then, and China now has to be included. His conclusion was blunt: “It’s a simple math problem. We need more than 1500.”
That claim was one of the sharpest points of tension. Wormuth agreed on the need for a strong deterrent but warned against unconstrained escalation. Jane Harman accepted the need to modernize the triad and increase defense spending in some contexts, citing her work chairing the Commission on National Defense Strategy. But she rejected what she called matching China and Russia “nuclear widget for nuclear widget.” In her view, that may make the United States less safe, not safer, unless it is embedded in a broader strategy using all elements of national power: hard power, diplomacy, humanitarian aid, and alliances.
Missile defense raised the related question of whether technology can reduce vulnerability without creating false confidence. Diego Garcia, in an audience question, raised “Golden Dome,” space-based interceptors, and projections that he said put the cost as high as $4 trillion. Kroenig said the Strategic Posture Commission on which he served recommended that the United States develop a homeland missile defense system to deal with Russia and China. He acknowledged that a perfect shield is impossible. If Russia and China launch a massive attack, any system can be overwhelmed. But a limited system could take “cheap shots” off the table. In a war over Taiwan, for example, if China launched a handful of missiles at the United States, Kroenig argued Washington should not be vulnerable to that by policy choice.
He said moving to space could make sense because the best time to intercept a missile is during boost phase, when it is slow and trying to escape Earth’s gravity. Space-based interceptors, he said, were once considered impossible or too expensive, but cheaper commercial satellite launch has made them more plausible.
Harman joked that Golden Dome was the perfect name for a new ballroom, then added more seriously that boost-phase intercept is a technology worth investing in. Wormuth was more skeptical about cost and feasibility. She said the current limited missile defense system can handle a handful of incoming weapons, the kind of threat associated with North Korea or Iran. It cannot handle a strategic attack by Russia or China. She also said it cannot deal with many of the advanced conventional cruise missiles that China and Russia are developing. She was skeptical of Golden Dome’s space-based interceptor concept on cost grounds, but supported space-based sensors and a system to counter advanced conventional cruise missiles.
The narrow area of agreement was that missile defense has a role against limited attacks and some emerging conventional threats. It does not solve the problem of large-scale nuclear vulnerability.
Harman’s fear is not only weapons, but the absence of strategy
Jane Harman’s central critique was political and institutional. She said she has “very little hope” that the current U.S. administration can marshal other governments into a durable nuclear arrangement, not because of partisanship as such, but because she sees the administration as “totally transactional and not strategic.”
Her indictment extended to Congress. Harman said the United States Congress has “totally failed” to act strategically and in bipartisan fashion, contrasting the current environment with the era of Sam Nunn and Dick Lugar. She invoked Nunn-Lugar, the post-Cold War cooperative effort to reduce the Soviet arsenal, and noted that Ukraine was one of the countries that gave up nuclear weapons. In the present environment, she said, Ukrainians may look at that decision with regret, especially when comparing themselves with North Korea, which kept and expanded its nuclear capability, or Iran, which may seek one.
Harman’s broader claim was that in a world where states feel defenseless without nuclear weapons, proliferation becomes more attractive. She pointed to reported or possible cases in deliberately cautious terms: Saudi Arabia, she said, had reportedly purchased a nuclear bomb from Pakistan but did not yet have it; Israel reportedly has nuclear weapons; South Korea and Japan had, as she remembered President Trump’s position, been encouraged to develop stronger defenses, possibly including nuclear capacity. She added rogue actors to the concern, warning that limited nuclear actions could be misunderstood and trigger larger ones.
Her most repeated word was “strategy.” She acknowledged that the defense strategy commission she chaired unanimously and on a bipartisan basis recommended increasing defense spending and improving the nuclear triad. But she said those recommendations came in the context of a full national strategy. In her view, the present problem is that the United States is not using hard power, soft power, diplomacy, aid, and alliances coherently.
Harman also challenged the assumption that leaders will behave rationally. The logic of deterrence depends heavily on rational calculation: adversaries understand consequences and step back from unacceptable risks. Harman said she does not assume rational behavior is the norm. She worried about miscalculation, weakened institutions, and the removal of senior expert military leaders from top positions.
Her concern extended beyond deliberate nuclear use to nuclear accidents. She raised Zaporizhzhia, the nuclear reactor in Ukraine, saying Russian control and bombing around the site have made the situation increasingly dangerous. If the core becomes more exposed, she suggested, a “nuclear accident-ish” event could trigger very bad responses.
Proliferation pressure rises when the umbrella looks weak
North Korea and Iran mattered less as isolated threats than as examples of why countries and allies calculate differently when nuclear guarantees appear uncertain.
Asked about Annie Jacobsen’s book Nuclear War, which Tett said left her unable to sleep, ? christine-wormuth said it is a terrifying and effective dramatization of nuclear consequences and the compressed timeline facing a U.S. president. But she had “quibbles” with some specifics, including the scenario of a North Korean submarine off the California coast. North Korean submarines, she said, are very noisy, unlike Russian submarines in particular.
Wormuth said that, now outside government, she can say North Korea is essentially a nuclear weapon state with a small arsenal. The current U.S. missile defense system, she said, is built to defend against that kind of threat. She still worries about Kim Jong-un, but said there is “some relative stability” with him at the moment. Her greater concern is Putin: Russia is on the back foot in Ukraine, and a cornered actor may lash out.
The Iran discussion was more specific. An audience member asked whether the United States is prepared if Iran had a bomb and sent it to the United States. Wormuth responded that the current missile defense system was designed for North Korean and Iranian threats, and that she is not overly worried at this point about Iran putting a warhead on an ICBM. But she also said it is important to be accurate about Iran’s material. The highly enriched uranium buried underground, she said, could be made into a crude nuclear weapon on the scale of the bombs dropped on Hiroshima or Nagasaki. In her view, it is not accurate to say Iran lacks the material to make a weapon. What does not keep her up at night right now is mating that material to a missile.
For Matthew Kroenig, U.S. extended nuclear deterrence is not merely a defense promise. It is one of the main reasons nuclear weapons have not spread further. Asked whether Europe can still count on the American nuclear umbrella, Kroenig answered yes. He described U.S. extended deterrence as “one of the greatest forces for good in the world over the past 80 years,” because the United States uses its nuclear weapons not only to defend itself but to defend what he called “the entire free world.” More than 30 formal treaty allies, including NATO allies, South Korea, Australia, and Japan, are told they do not need their own nuclear weapons because they can rely on U.S. weapons.
Without that promise, he said, Germany, Taiwan, and other countries likely would have nuclear weapons today. He acknowledged European concerns about the current administration but said its position has been consistent: European allies should do more on conventional defense, while U.S. extended nuclear deterrence remains in place. He also said a second Trump administration would resist allied attempts to build their own nuclear weapons.
Jane Harman was much less reassured. She said statements by the current president have not clearly reinforced the nuclear umbrella over Europe, and that the administration’s comments about NATO have been negative enough to drive some countries to conclude the United States will not be there for them.
A Polish participant asked about extending NATO nuclear sharing, including certifying Polish F-35s for such a role. Kroenig said such an idea could make sense. The United States has non-strategic nuclear weapons deployed on NATO territory, but he said they are essentially still located where they were left at the end of the Cold War, in Western Europe. During the Cold War, their presence signaled to the Soviet Union that an invasion of NATO could quickly go nuclear and reassured allies that they were protected. Today, if Putin invaded NATO, Kroenig said, it would not be Western Europe but Poland or the Baltic states. In his view, deploying some non-strategic nuclear weapons farther east could strengthen deterrence, though he emphasized that this is not current U.S. policy.
That is the dilemma: reassurance can prevent allies from building nuclear weapons, but forward-deploying nuclear weapons can also look escalatory. Wormuth put the proliferation concern plainly: she worries about South Korea, Japan, or Poland seeking their own weapons if U.S. guarantees weaken. Her phrase was: “Friends don’t let friends proliferate.”
The tactical nuclear distinction does not make nuclear use small
The difference between strategic and tactical nuclear weapons matters because Russian threats often imply that smaller, battlefield nuclear use could be limited and manageable. Tett said Russian television discussions often center on tactical weapons and the idea that a “nice little, small tactical nuclear weapon” used against Poland or Ukraine would have contained fallout. That, she said, is why weather appears in the propaganda.
Matthew Kroenig described the Russian doctrine often called “escalate to de-escalate.” If Russia were losing a war on its border, whether in Ukraine or against NATO, it might use one, two, or a dozen lower-yield nuclear weapons on the battlefield to frighten Washington and other Western capitals into backing down. He asked the audience to imagine being the U.S. president after six nuclear weapons are used in Europe. Does the United States respond with everything it has, risking suicide? Does it back down because the war is not worth nuclear escalation? If it backs down, he argued, Putin learns that nuclear use works.
Kroenig’s answer was flexible response. The United States, he said, has options between doing nothing and launching everything. It also has lower-yield nuclear weapons. The message to Putin, in Kroenig’s account, is that using one, six, or twelve nuclear weapons does not produce victory. The result may instead be a limited nuclear war with NATO. The goal of possessing such options is not to fight that war, but to convince Putin not to start down the path.
? christine-wormuth agreed that U.S. planners want rungs on the escalation ladder and options for the president. But she rejected the idea that tactical nuclear weapons have meaningful battlefield utility. Anything needed tactically on the battlefield, she said, can be achieved with conventional weaponry. The real significance of tactical nuclear use would be normative and strategic: it would break the nuclear taboo that has held since 1945.
A tactical battlefield use would fundamentally break the seal on the nuclear taboo and none of us want that to happen.
She credited past U.S. diplomatic messaging and Xi Jinping’s intervention with helping signal to Putin that nuclear use would be unacceptable. She said that when Putin rattled the nuclear saber a couple of years earlier, the U.S. government began intense diplomatic messaging, and Xi also made clear that Putin should not go there. Wormuth said she had recently been in Shanghai for Track II talks with Chinese nuclear experts and made a point of thanking them for Xi’s intervention. If Putin resumes this behavior, she said, China and others must again send the message that nuclear use is a terrible idea.
Jane Harman’s objection was that all of this assumes rationality and functioning communication. She questioned whether comparable strategic conversations are occurring now and worried about miscalculation in a degraded institutional environment.
China is both a nuclear problem and a possible channel for restraint
David Jones asked for “a shot of optimism” and focused on China. He said he sees a bipartisan consensus emerging that China is America’s enemy, ideologically and militarily. But he asked whether shared interests in global trade and decades of working together might provide a basis for cooperation on nuclear issues.
? christine-wormuth offered the clearest opening. She said she views the U.S.-China relationship as a mix of competition and possible cooperation. In her recent Track II discussions in Shanghai, she said Chinese nuclear experts appeared quite worried about artificial intelligence and what AI could mean for nuclear systems. That may create an opportunity for official conversations. After the last Trump-Xi summit, she said, a working group was announced to discuss AI and related issues.
She also saw possible openings on missile launch notification agreements and other confidence-building measures. Wormuth did not suggest China is eager for arms control. She said China is not currently eager to enter any arms-control agreement. But communication channels and confidence measures may still be possible.
Her emphasis was practical: the Chinese are not the Russians, and they are not the Soviet Union. They have a different strategic framework. Without sustained dialogue, she warned, the United States and China could easily misunderstand each other. She argued for more channels of crisis communication, not only between heads of state but also between military leaders, so the two sides are not speaking for the first time when a crisis begins.
Jane Harman invoked Henry Kissinger’s frame for China: confront, compete, and cooperate. She said that is the right approach. She also suggested that, despite rhetoric, Donald Trump wants a positive relationship with Xi, and that at the last summit Xi appeared commanding while Trump appeared somewhat cowed. Harman presented that, cautiously, as “kind of positive.”
Wormuth also corrected herself on treaty status. She said she was “quite sure” China is a signatory to the Non-Proliferation Treaty and that she may have been thinking of the Comprehensive Nuclear-Test-Ban Treaty. The distinction mattered because the nonproliferation bargain depends on two sides of restraint: non-nuclear states agree not to acquire weapons, while nuclear weapons states are expected to behave responsibly and move toward disarmament. Wormuth said the last NPT review conference could not reach consensus because non-nuclear weapons states are upset with nuclear weapons states.
Arms control has to contend with extended deterrence
Asle Toje of the Norwegian Nobel Committee pressed the question of what a rebuilt arms-control framework might look like and whether No First Use could be a useful first step.
Matthew Kroenig said he has written about what a trilateral arms-control arrangement among the United States, Russia, and China might look like, while acknowledging that getting there would be very hard. China, he said, will not even engage in serious arms-control talks right now.
On No First Use, he was opposed from a U.S. perspective. A No First Use policy would mean the United States pledges not to use nuclear weapons first and would use them only in response to nuclear use by someone else. But Kroenig argued that U.S. nuclear weapons also deter conventional attacks against allies. If Washington adopted No First Use, he said, it would effectively tell Putin he could invade Estonia and commit atrocities like Bucha as long as he did not use nuclear weapons. Kroenig said the United States wants Putin to worry about U.S. nuclear weapons precisely to deter that kind of conventional aggression against NATO allies.
That disagreement fed directly into what citizens can plausibly do. Kroenig urged support for increased defense spending. He said the nuclear submarines, missiles, and bombers are old, many dating from the end of the Cold War, and compared relying on them to depending on a car built in the 1980s. Modernizing the triad, he said, is estimated to cost more than $1 trillion, and in his view even that is not enough for the current threat environment. Given war in Europe, war in the Middle East, and daily Chinese aggression in the Indo-Pacific, he argued, a strong U.S. military and nuclear deterrent are necessary not only for the United States but for the broader free world.
Harman accepted modernization and higher spending in principle, as her commission had recommended, but again only inside a strategy. She said she does not see that strategy, and worried that some of the “good thinkers” no longer work in government.
Wormuth urged citizens to talk to members of Congress and political leaders. Nuclear issues, she said, have fallen off the radar. In the era of Nunn and Lugar, members of Congress took these issues seriously across party and chamber lines. Representatives and senators, she argued, need to hear that voters care about nuclear risks and expect them to pay attention.
She also urged support for organizations such as the Nuclear Threat Initiative, which she leads. NTI, she said, is no longer trying in the near term to rid the world fully of nuclear weapons; she does not see that happening in the foreseeable future. Instead, it is working on stability, strategy, and dialogue. That includes efforts to open conversations with Chinese nuclear experts so that each side better understands how the other thinks, and to create opportunities for launch notification agreements and military crisis-communication channels.
Harman noted that the United States used to have military-to-military communications with China and Russia and spoke to the Russians throughout the Cold War. She said the danger of miscalculation is huge if leaders cannot pick up the phone and ask what is really happening. Wormuth said there was a Pentagon phone line during the Obama administration, though she was unsure how often the Chinese answered or what channels exist now.
The unresolved tension is structural. Arms control, restraint, and norms are necessary to prevent spirals and accidents. But extended deterrence deliberately preserves ambiguity and fear to prevent conventional war. Kroenig leaned toward capability and deterrent credibility. Wormuth leaned toward a strong deterrent bounded by cost, strategy, arms-control efforts, and crisis communication. Harman repeatedly insisted that neither weapons nor doctrine can substitute for strategy and functioning political institutions.

