A New Nuclear Order

The Trump administration has been clear-eyed about the shift from a unipolar, hegemonic American-led global order to a multipolar one in which the United States will be, though still the strongest, one of many poles.¹ In recognition of this strategic reality, the administration has affected dramatic changes in domestic and foreign policy: its efforts at reform have encompassed world trade, military preparedness, relations with allies and adversaries, and internal governance. Untouched thus far in this raft of policy rewrites, however, has been America’s approach to great power nuclear negotiations. This is partially due to the multitude of fixed attitudes that need to be revised in any shift from one paradigm to the next; inevitably, something is changed last.

Nuclear weapons were present at the very beginning of the Cold War, and negotiations between the United States and the Soviet Union shaped the entirety of the period. The specter of the superpowers hung over every treaty signed, nuclear or non-nuclear. Every potential spark—from the Korean War to the Cuban Missile Crisis to random Soviet satellite glitches—could have ended in annihilation. The pattern of great power nuclear negotiations that emerged from this period was essentially repeated so often that it became entrenched, generating its own momentum: the United States and the Soviet Union (later Russia) bilaterally negotiating to reduce offensive or defensive armaments. These negotiations were occasionally accompanied by treaties restricting proliferation on a global scale, such as the Nuclear Non-Proliferation Treaty (NPT).

This pattern, which made sense amid the stark bipolarity of the Cold War, persisted long after 1989. It was carried on in bipartisan fashion by successive post–Cold War American presidential administrations desperate to hold onto the architecture (and the certainty and comfort that came with it) that had fostered decades of U.S. hegemony. But in extending this architecture, and specifically, this nuclear negotiation strategy, they made a series of mistakes that have damaged America’s national interest and put it in a weaker position entering into the era of multipolarity.

Yet multipolarity will bring chaos to the global order, rendering obsolete the rigid assumptions and relative strategic clarity of the old bipolar world.² Now is the time, therefore, to alter America’s nuclear strategies, both as they relate to great power negotiations and to nonproliferation. The vastly changed circumstances of the post–Cold War world, to which American leadership is only belatedly reacting to, will require bold and somewhat out-of-the-box adaptations, ones that will be in line with the spirit of institutional reinvention embodied by President Trump.

Accordingly, in this essay, I will recommend the withdrawal of the United States from the Nuclear Non-Proliferation Treaty, a piece of international law which, while highly useful in the Cold War context, no longer serves the interests of the United States in the multipolar twenty-first century. I will further argue for the replacement of the NPT with a new foundational, trilateral nuclear treaty.

But before these recommendations can be presented, it is worth addressing the errors inherent in U.S. policy as it stuck to Cold War-era negotiating axioms. All are interrelated, and all worked to damage, or at least postpone, America’s ability to prepare for multipolarity.

The First Error: Nonproliferation for Its Own Sake

The Cuban Missile Crisis, in which the world brushed with Armageddon, jolted American and Soviet leadership into signing a series of treaties to limit the global development of nuclear arms. The capstone to this effort was the Nuclear Non-Proliferation Treaty, signed in 1968 and effective in 1970, which formalized two “classes” of states: nuclear weapon states and non-nuclear weapon states. Nuclear weapon states³ were allowed to keep their weapons upon signing, but had to promise to not aid other states in procuring their own. Non-nuclear weapon states had to swear off seeking to obtain a nuclear device.

The NPT had a time-limit written into it: after twenty-five years, the state signatories would decide whether it was to be permanently extended. Extending the NPT into the post–Cold War years was a major foreign policy goal of the Bill Clinton administration, as the United States continued its Cold War–era approach of opposing all nuclear proliferation. This effort, in which the Clinton White House prided itself on playing a “critical role,” succeeded in 1995.⁴ The treaty had helped to create the norm that proliferation of nuclear arms was inherently “bad,” and the United States, on a bipartisan basis, prioritized the maintenance of this rule as a global norm.

The NPT did not, however, stop the proliferation of nuclear weapons. Out of the nine nuclear weapons states that exist today—the United States, Russia, China, the United Kingdom, France, Israel, Pakistan, India, and North Korea—at least three obtained nuclear weapons after the treaty came into force: Pakistan, India, and North Korea. Though not a nuclear weapons state, Iran has been able to remain perpetually “weeks away” from obtaining nuclear weapons.⁵ Libya also worked on developing nuclear weapons for years, though it did give up its program in 2003. Yet it did not do so out of loyalty to norms: it did so because the United States had just invaded Iraq over the same issue, and Libyan dictator Muammar Ghaddafi was interested in self-preservation.

This gives away the real reason why states obtain nuclear weapons: they do so out of national self-interest. If a state determines that it must have a nuclear weapon in order to protect itself, it will seek to obtain one. India and Pakistan saw each other as existential threats (and India felt threatened by China), causing them to seek a nuclear device in order to ensure neither could eliminate the other. North Korea felt the need to overcome its reliance on China, and it desired the capacity to deter anyone attempting to overthrow the ruling regime with the threat of nuclear retaliation. Iran, surrounded by hostile military forces, regularly threatens to obtain a nuclear device to gain leverage in negotiations and, should it ever need one, to deploy as a weapon of last resort.

States that have never initiated a nuclear weapons program may simply feel that they do not need one. For instance, some American allies within NATO and beyond, such as Spain or Japan, have adequate technical capacity to develop nuclear arms but choose not to do so, as they remain content to be under the protection of the U.S. nuclear umbrella. Sometimes, fear of another nuclear power can also keep a state from developing nuclear weapons. To take one example, Vietnam has not developed nuclear weapons because it did not want to upset relations with its significantly more powerful neighbor, China.⁶

None of this is to say that the United States should encourage all states to have nuclear weapons. But forcing a Cold War–era “one-size-fits-all” nonproliferation policy onto a fundamentally different era has been a serious error. The most blatant example of this was when the United States blocked Ukraine from obtaining working nuclear weapons and Germany from considering whether it should develop one. When the Soviet Union collapsed, a large portion of its nuclear arsenal was left on the territory of the new Ukrainian state. The George H.W. Bush and Clinton administrations pushed Ukraine to give its weapons to the nascent Russian Federation. This ultimately occurred, resulting in the Budapest Memorandum in which all parties, including the United States and Russia, agreed to respect Ukraine’s territorial integrity.⁷

At the time, there were few voices arguing against Washington’s actions. On the surface, the U.S. foreign policy establishment had good reasons for pursuing such a stance amid concerns that the post–Soviet states could descend into an orgy of violence. Plus, the weapons were, in the eyes of the U.S. government, property of the Russian Federation as the Soviet Union’s main successor state: as such, had Ukraine gained access to them, it would have broken the norm of nonproliferation.

But one dissenter from this consensus, John Mearsheimer, political scientist at the University of Chicago, argued that America’s pressure campaign was a mistake. In a Foreign Policy article published in 1993, Mearsheimer argued that Russia would inevitably try to recapture Ukraine, highlighting the likelihood of a war starting over “control of the Crimea,” pointing out that such a war could expand into Europe and result in millions of refugees. He further argued that this would in turn increase tensions between Russia and the West.⁸

Mearsheimer made the case that a nuclear Ukraine—and a nuclear Germany—would keep peace in Europe and allow the United States to move its troops out of the continent, something he believed would happen in short order. He expected this because he saw the prospect of continued U.S. military presence in Europe, once the Soviet threat had been extinguished, as irrational, and because he believed Germany would be well-prepared to defend itself and its neighbors. A nuclear Ukraine and Germany would effectively have penned Russia into its traditional sphere of influence in northeastern Europe but would have kept it from going any further.

Yet contrary to Mearsheimer’s hopes, the United States did not remove its troops from Europe, nor did it allow Ukraine or Germany to develop nuclear weapons. As a result, the worst of both worlds occurred: tensions with Russia steadily rose as NATO expanded eastward, and states which could have had strong militaries, like Germany, allowed their forces to atrophy. But at the same time, America’s desire to defend Europe was diminishing.⁹ Eventually, these currents crested together into a tsunami: the Russo-Ukrainian War, the largest war in Europe since World War II.

It is unlikely that the Clinton administration believed their actions would lead directly to war. They were clearly unwilling, however, to give up American hegemony and did not see any wisdom in allowing non-nuclear states to develop nuclear weapons.

The problem is that history has shown how woefully wrong this way of thinking was. And this has had a direct negative effect on the United States. While the United States today does not need hegemony over all of Europe, Washington does not want to risk the entirety of Eurasia falling under a two-state alliance of Russia and China. While successive U.S. administrations may have seen a weakened Russia as a useful foil, the possibility of Beijing being able to divide the Eurasian landmass, or even be the dominant partner in an alliance with Moscow, was woefully underexplored by U.S. policymakers.

The Second Error: Groundhog Day

In June 1992, about six months after the collapse of the Soviet Union, Presidents George H.W. Bush and Boris Yeltsin signed a memorandum calling for a sequel treaty to START.¹⁰ The resulting agreement, START II, signed by both leaders in January 1993, was a sequel in two ways. For one, it cut nuclear arms in a similar manner to START I. But it was also explicitly linked to START I by holding that it would only go into effect once the earlier treaty had first gone into effect. A previous agreement, the Lisbon Protocol, had clarified that Belarus, Ukraine, and Kazakhstan, the post-Soviet states which hosted nuclear weapons other than Russia, were also party to START I and required them all to join the NPT as non-nuclear weapons states. Thus, America’s goal of broader nonproliferation was literally written into START II.

Even with the mistake of pursuing nonproliferation at all costs, however, START II was still a reasonable treaty: one could see it as almost putting a bow on the Cold War, a final wrap-up on decades of bilateral great power nuclear negotiations, which had started with SALT in the early 1970s. In a speech discussing his early plans for the treaty, President George H.W. Bush described it as in-line with his “post–Cold War strategy.”¹¹ With START II, the United States could turn a page and change course for the new world.

Of course, as discussed in the previous section, that did not happen: the United States insisted upon nonproliferation everywhere. But successive presidential administrations also insisted upon the continuation of one-on-one nuclear negotiations, exclusively between Russia and the United States.

START II also did not happen. While the treaty was signed, it never went into effect. This was because the Russian State Duma’s agreement to support START II was contingent upon the survival of the Anti-Ballistic Missile (ABM) Treaty; when the Bush administration withdrew from the ABM Treaty in 2001, START II was ended. It was soon replaced with a new treaty: the Strategic Offensive Reductions Treaty (SORT), also called the Moscow Treaty, negotiated between the administrations of Presidents George W. Bush and Vladimir Putin. SORT mandated that Russia and America reduce deployed warheads to no more than twenty-two hundred. The treaty, however, contained absolutely no means of verification.¹²

Nevertheless, in an interview, National Security Advisor Condoleezza Rice portrayed the treaty as “liquidating a legacy of the Cold War” and, elsewhere, argued that SORT was “the last treaty of the last century.”¹³ These were rather bold, if not hubristic, statements.

That was ultimately not the final treaty of the Cold War, as SORT was, a decade later, replaced with New START. Unlike SORT, New START included strict verification measures; like SORT, it was exclusively negotiated between the United States and Russia. Also like SORT, the treaty was portrayed by President Barack Obama as finally allowing Washington and Moscow to “[leave] behind the legacy of the 20^th century.”¹⁴

Of course, New START (like START II and SORT before it) did no such thing. Nine years later, in 2021, it would be quietly extended by Putin and then-President Joe Biden until 2026. But two years later, amidst Mearsheimer’s predicted Russo-Ukrainian War, Putin announced Russia would suspend its participation in the treaty.¹⁵

This chain of ill-fated treaties persisted due to inertia alone; SORT, in particular, seemed to be just checking a box: negotiating nuclear missile treaties with Russia was just something U.S. administrations did, and, therefore, it had to be done. But this verifiable Groundhog Day, in which each new treaty was portrayed as the actual end of the Cold War, just for both countries to repeat the charade five to ten years later, cost the United States. Yes, this focus on Russia was likely purposeful, as by focusing on a relatively weak Russia, the United States had an excuse to keep up the massive military-industrial complex underpinning its hegemony. But it also cost the United States in the long run and wasted valuable time it could have spent focusing on the real twenty-first century threat: China.

The Third Error: Ignoring China

In a 2000 paper, a trio of researchers argued that China was a “forgotten nuclear power.”¹⁶ Under President Jiang Zemin, China, they wrote, was undergoing a nuclear modernization campaign. At the time of writing, China had “about 20 missiles . . . capable of reaching targets in the continental United States.” They further highlighted Chinese skepticism over “Washington’s professed commitment to de-emphasizing nuclear weapons,” and that China saw “U.S. reluctance to embrace a no-first-use doctrine as an indication of American planning for the use of nuclear weapons preemptively.” They further predicted that China would increase their ICBM count to a degree at which it could restore minimum deterrence with the United States.

Nearly twenty years later, almost the exact same concerns were being raised, with one analyst calling China “nuclear disarmament’s missing player” and further writing that China would seek to ensure that American and Russian nuclear stockpiles had a number “similar to its own.”¹⁷ Since the end of the Cold War, China has indeed proceeded with their nuclear modernization campaign (a campaign it is still undertaking, as of this writing), yet the United States still has not taken any steps whatsoever to bring China into nuclear negotiations.

This is not to say that American administrations have not been concerned about China’s nuclear arsenal: indeed, the concern has been a point of bipartisan convergence. In 2020, Secretary of State Mike Pompeo said that the Trump administration had “called on China to conform its nuclear capabilities to the strategic realities of our time.”¹⁸ A year later, his successor, Anthony Blinken, explicitly said that he had “deep concern with the rapid growth of the PRC’s nuclear arsenal.”¹⁹ Neither party actually sought to take any actions to convince China to change course or to bring them into great power nuclear negotiations.

When President George W. Bush pulled out of the Anti-Ballistic Missile (ABM) Treaty, China was extremely concerned, with some Chinese analysts saying that China would be forced to respond, as it would feel threatened if the United States decided to place missile defense systems in the Pacific.²⁰ And indeed, in a leaked speech to Goldman Sachs, Hillary Clinton revealed that the Obama administration had threatened to “ring” China with such systems.²¹ The Bush administration seemed aware of China’s concerns, and President Bush had even called Chinese President Jiang before the formal announcement of the ABM pullout, offering “high-level strategic talks” so as “to make sure that China did not feel that it was being frozen out as Washington builds a ‘strategic framework’ with Russia.”²² Nothing ever came of it.

As a result, we are now entering a multipolar era with no de facto treaties between the United States and Russia, which possesses the most nuclear weapons on the planet, and no de jure treaties with China, which is ratcheting up its nuclear forces and will likely be the most powerful pole after the United States.

The Fourth Error: Attacking Libya

The final mistake involves trust, specifically, earning the trust of the non-great powers. While great power nuclear negotiations are crucially important, earning the trust of medium-sized or even small states remains important for great power security. In the early 2000s, the Bush administration was explicitly concerned about nuclear proliferation not from larger powers but from smaller ones, such as North Korea. This fear, a relatively marginal concern throughout the 1990s, grew immensely after the September 11 attacks, and was, at least officially, the primary reason for the 2003 invasion of Iraq.²³

If a state is seeking a nuclear weapon, it is likely doing so because it feels threatened. To get a state seeking to develop nuclear weapons to stop its pursuit, the threat must either be removed, ideological leadership must change, or the threat of force must be so overwhelming that the risk of developing a nuclear device becomes greater than the risk of not developing one.

In 2003, Libya ended its nascent nuclear weapons program. This was likely due to the nearly concurrent U.S.-led invasion of Iraq and, possibly, to unconfirmed threats from the Bush administration. In a sense, this was one of the few success stories from the Iraq War: a rogue state gave up its nuclear program after determining that the risk of continuing it outweighed the risk of not doing so. It must be emphasized that Libya had already been a party to the NPT before it revealed and gave up its nuclear program. It was not the liberal democratic order which convinced them to change course. It was the threat of brute force.

Yet eight years later, the United States helped to overthrow Libya’s government. Its leader, Muammar Ghaddafi, was killed, and the country descended into a civil war from which it has yet to emerge. That Ghaddafi was killed was not necessarily a tragedy: his dictatorship was, of course, brutal and oppressive. But the way he was killed—via a “responsibility to protect” (R2P) mechanism passed into law by the United Nations Security Council, after he had given up his nuclear weapons program—evaporated any sort of coercive effect the Iraq War may have had on rogue states considering nuclear weapons. After all, what is the point of giving them up if you will be overthrown anyway? That the liberal democratic order could be turned so clearly to these ends (the Russians had reportedly been promised that R2P would not be used to effect Ghaddafi’s overthrow) also likely weakened trust in widespread multilateral agreements.

A key aspect of the NPT is that states agree to give up possessing nuclear weapons in exchange for certain rights, including the ability to develop nuclear energy for peaceful purposes. Implicit also is the “right” to not be attacked, for which states give up the protection that a nuclear weapon could provide. But from the end of the Cold War to the advent of the second Trump administration, the United States sought to have it both ways: to continue using liberal democratic language and the hegemonic structures of the Cold War while also enforcing its will selectively and wherever it wished. All this achieved was a lack of trust with all parties. In a multipolar world, states will already have less reason to stay away from nuclear weapons—but they will now have less reason to trust the United States when it says it will not let them be harmed in exchange for abandoning nuclear weapons.

The developments outlined above are the tragic results of multiple U.S. administrations doggedly attempting to engage in what could be called “stretched bipolarity,” a better term to describe the last three decades than “unipolarity.” Long after the dissolution of the Soviet Union, the United States pulled its Cold War institutions like silly putty around the rest of the world. But that “world” order had been intended only for a bipolar, Cold War–era planet, and the clear failure to adapt it to the twenty-first century increasingly showed up via these various errors.

This series of mistakes—keeping a perpetual policy of nonproliferation everywhere, refusing to break out of the bilateral approach to great power nuclear negotiations, excluding China, and weakening trust with middle-tier countries—has left America with a rudderless nuclear negotiation strategy in the multipolar world. The Trump administration has been nothing if not proactive in recognizing the past mistakes made by America’s leadership in other areas, invoking these blunders as motivation to pursue bold corrective actions: it should be equally bold here.

The Case for Ending the NPT

The United States should announce that it intends to leave the NPT. The NPT’s withdrawal process, as established by the text of the treaty, takes three months. As New START is set to expire in February of 2026, the administration should consider announcing their withdrawal this year, to coincide with the end of that treaty.

The NPT was built for the twentieth-century Cold War order. It is the rotting foundation upon which almost the entirety of all nuclear disarmament architecture rests.

Nuclear negotiations set the tone for almost the entire duration of the Cold War. Because of the continued centrality of nuclear weapons, whatever foundation exists once the multipolar world is firmly set will likely be the foundation (rotting or otherwise) for the rest of the century. The NPT cannot be that foundation, as it no longer makes sense in the multipolar context.

Leaving the NPT will dramatically increase America’s potential for shifting forces and resources. While the United States no longer needs hegemony over Europe, it still requires, for its security, some extended version of the Monroe Doctrine. The concept has various names, from Monroe+ to the American Sphere Doctrine, but it essentially holds that the United States should not allow explicitly anti-American governments to become dominant in the Americas, in Western Europe, or in East Asia, with the Americas taking priority.²⁴ Until now, it has sought to ensure this with the placement of military forces. But that course has been costly and is increasingly unpopular, especially in the face of Europe’s refusal to spend on its own defense.²⁵

Allowing the proliferation of nuclear weapons in key areas, or even the threat of it, would decrease pressure on the United States. Countries like Poland could adopt an Iranian strategy of getting close enough to developing a bomb to be taken seriously, or it could develop one altogether.²⁶ A Polish bomb would essentially keep a Russo-Chinese alliance from dominating Europe, as no serious threat of invasion could be made. This would have the effect of recreating Mearsheimer’s prescription, just over thirty years later.

There is also the argument to be made that the treaty is already effectively dead. Each time a new country credibly threatens to get a nuclear weapon (such as in the case of Iran) or attains one (such as in the case of North Korea, Pakistan, or India), it underscores that the NPT is not really self-enforcing and carries no import on its own merit. To those who are already skeptical of the liberal international order, this is not Earth-shattering news. Yet many of those skeptics still have not come around on the uselessness of the NPT.

And this is further damaging to America’s credibility in a multipolar world. For one, the NPT is clearly identified as an American-led treaty. Yes, Russia and the rest of the UN Security Council are members, but it was the Clinton administration which spearheaded the treaty’s perpetual extension in 1995, and it is the United States which has sought to restrict nuclear proliferation the most. This means that every time the treaty is violated, or even whenever a non-signatory credibly threatens to obtain a nuclear device, American credibility is put on the line.

The NPT also requires the threat of American police action to enforce. And while, in a multipolar world, the United States may need to take military action to defend its national interests, that national interest is not inherently threatened by any nation anywhere ever getting a nuclear weapon. The Trump administration and realists have correctly determined that this Gordian knot—of America being forced into the role of world police in order to save credibility—can simply be sliced in two by walking away from the NPT and the anachronistic Cold War thinking underlying it.

U.S. withdrawal will have other benefits: it would prove that Washington is no longer interested in clinging on to legacy Cold War structures but is instead interested in treating the other poles with the serious recognition they merit, as opposed to secondary states to be pushed around.

Many smaller and larger countries around the globe, some of whom we will need to work with in the coming multipolar order, simply do not trust the United States because of its past desperate attempts at prolonging hegemony. Leaving the NPT would make clear that that time has passed. It would likewise send a message to the “unaligned” states between the poles that we are no longer interested in being the world’s police. Instead of holding on to the sand of stretched bipolarity as it slips through our fingers, American leaders can instead build the multipolar order in terms that are far more realistic as well as broadly conducive to American security and global stability.

Toward a New Trilateral Treaty

The architecture of the Cold War’s bipolar world was shaped by nuclear negotiations. The architecture of the multipolar world should be likewise. To that end, the Trump administration should suggest a new trilateral treaty, to be signed with Russia and China. The treaty should have two main points:

First, the treaty should include a no-first-use commitment from all signatories. Unlike Russia or the United States, China already has such a policy in place. But an embrace of the concept by Washington and Moscow would weaken the fear of significant military escalations in the case of a conflict. It would also turn no-first-use into a definitive international norm, a useful thing in a multipolar world in which nonproliferation is not a norm.²⁷

Second, the new treaty should include a ceiling for how many warheads can be possessed, one which is substantially lower than New START’s. One study found that even a limited nuclear war could end billions of lives, and others have reached similar results.²⁸ While the number could be up for debate, it should be in the hundreds. With hundreds of nuclear warheads, the United States can still destroy any country that targets it and have leftover warheads to spare.

This would achieve much. It would save the United States money in the long run by no longer requiring the maintenance of thousands of nuclear warheads. It would also allow the United States to appear magnanimous, climbing down to China’s level.

Such a treaty would also likely be appealing to China, as the country would no longer be pressured with having to catch up to America and Russia and could therefore cease its nuclear build-up. This arrangement would also lessen pressure on the United States (and others, such as India) to react to China’s build-up, and it could help avoid a new multipolar arms race altogether. This new treaty could be extended to include other nuclear powers as well, such as France and India. As China and Russia, however, will be the two main rival poles in the new world order, it is imperative that America ensure that those two join the treaty.

The world has not been truly multipolar since the early twentieth century, before the dawn of the nuclear age. Old strategies regarding multipolarity must be updated to include lessons learned from the nuclear era. Multipolarity will bring instability and new challenges but will also open new opportunities. In order to maximize those opportunities and minimize chaos, the United States should abandon Cold War–era thinking and revitalize its approach to great power nuclear negotiations. The proposals put forward here–leaving the NPT and replacing it with a new trilateral agreement–will help realize this goal.

This article is an American Affairs online exclusive, published November 20, 2025.

---

Notes