The global system to prevent nuclear proliferation and promote disarmament is beginning to fray. Although the nonproliferation regime has held together for more than half a century, more countries are acquiring sensitive nuclear material and technology through illicit acquisition and preferential trade. In May 2021, for instance, the International Atomic Energy Agency (IAEA) reported that Iran had accumulated ten kilograms of highly enriched uranium and severely restricted access to its nuclear sites. And in October 2021, Australia, the United Kingdom, and the United States announced a new strategic partnership (AUKUS) that will make Australia the first ever nonnuclear state to receive highly enriched fuel for nuclear-powered submarines. It is unlikely Australia would divert this uranium to make bombs, but it establishes a dangerous precedent.

These two cases typify the growing challenges faced by the nonproliferation system. Historically, the framework set in place by the 1970 Nuclear Nonproliferation Treaty (NPT) relied on development limits and international monitoring of uranium enrichment and plutonium reprocessing to succeed. But the spread of these sensitive nuclear materials and technologies is substantially degrading both processes. It is increasingly difficult to distinguish between nuclear programs designed for peaceful purposes and those that aim to yield bombs. The IAEA toolkit to detect concerning activity, flag it, and address it diplomatically risks obsolescence.

To restore the nonproliferation regime’s role as a bulwark of global stability, international nonproliferation institutions and states need new ways to track and tackle the development of nuclear weapons. This requires an innovative approach to monitoring and constraining dangerous activity. But given that more countries are acquiring or producing highly enriched uranium, material constraints alone are not enough. Monitors will need fresh tools to credibly track additional indicators of potential bomb activity that are hard to pass off as peaceful in nature, such as weaponization: the development and manufacture of nuclear warheads for missiles or other delivery vehicles. Monitoring this kind of activity, in particular, goes beyond the traditional focus of nuclear observers, but it may now offer the best, most reliable way to know whether states are trying to acquire nuclear weapons.

Given the global rise in nuclear activity, the world should move quickly to create such a system. When existing powers are less able to prevent uranium enrichment—and even hand over highly enriched material to other countries—it incentivizes competitors to double down on their programs. When new states develop bombs, it further encourages proliferation. Especially in an era of growing geopolitical competition, the international community needs more indicative information on the spread of nuclear weapons so that diplomacy can head off destabilizing proliferation and arms races.


The NPT is the cornerstone of the nonproliferation order. It requires that nonnuclear states avoid acquiring weapons (Article II), allows all states access to nuclear technology for peaceful purposes (Article IV), and commits the signatories with nuclear weapons—China, France, Russia, the United Kingdom, and the United States—to eventually disarm (Article VI). In practice, this regime limits state possession of and the operation of technology used to produce the fissile materials needed for nuclear weapons: highly enriched uranium and plutonium. And it empowers the IAEA to stringently scrutinize nuclear research and energy programs, enforce nuclear safeguards, and detect clandestine production. Although its powers of enforcement are limited, the NPT has endured due to a combination of IAEA monitoring, strict nuclear trade regulations by multilateral institutions, and commitments by individual member states to abide by the rules.

The United States has historically played an important role in leading and augmenting this regime, mostly with the Soviet Union and later with Russia. To uphold the NPT bargain, Washington cooperated with many partner countries through the Atoms for Peace program, sharing technology and materials to advance nuclear energy and research in states that swore off atomic weapons. U.S. agencies also operated programs to constrain the spread of enriched uranium and nuclear production technology, promote transparency in civilian nuclear material, and convert research reactors from high- to low-enriched fuel. By doing so, Washington further slowed the accumulation of fissile and sensitive technology. Other nuclear states supported or at least acceded to these efforts, tightly restricting the flow of key materials and technologies.

The system isn’t perfect. There are inherent tensions in the NPT bargain, which allows some states to hold weapons while banning others from acquiring them. The world has experienced periodic proliferation crises that exposed gaping holes in the regime, such as in 1991, when observers discovered that Iraq had a clandestine program for producing nuclear weapons. Sometimes, the NPT has failed to stop weapons development altogether, as happened in North Korea, which withdrew from the treaty when its program was exposed. And three nuclear-armed states—India, presumably Israel, and Pakistan—never joined the NPT.

But remarkably, this package has generally held together for decades. Most countries that at one time had nuclear weapons aspirations walked them back, such as South Korea, Sweden, and Switzerland. South Africa voluntarily disarmed, dismantling the six nuclear bombs it had secretly built, and joined the NPT as a nonnuclear state. The world averted close calls, including one resulting from the demise of the Soviet Union, which could have easily yielded four nuclear states instead of one, and another in Iran, which was stopped en route in 2003. If anything, the norm against proliferation has grown stronger over time. Today, only six states without nuclear weapons have the indigenous capability to produce fissile materials—Argentina, Brazil, Germany, Iran, Japan, and the Netherlands—a testament to the efficacy of the regime. But there are multiple signs that this record may not continue.


Some of the recent problems with the nonproliferation system derive from the stalled progress toward nuclear disarmament by states with nuclear weapons. After Russia and the United States dramatically slashed their Cold War nuclear arsenals by retiring obsolete systems, arms reductions have come to a standstill in both countries in the last decade. Now, they are modernizing their arsenals, as are China, India, Pakistan, and the United Kingdom. This has prompted many nonnuclear states to push forward a nominally complementary but practically competing UN Treaty on the Prohibition of Nuclear Weapons. (The treaty calls for a categorical ban on the possession of nuclear weapons by the signatories, and it has been rejected by all nuclear weapons states and their allies.) Other issues come from the behavior of nuclear states outside the system, most problematically North Korea. But the most worrisome problem is that the barrier between peaceful nuclear activity and weapons development is eroding.

Most of the damaging erosion has been done by the weapons states themselves, via ad hoc arrangements to advance other strategic interests. Notably, the 2005 U.S.-Indian nuclear deal, endorsed in 2008 by the Nuclear Suppliers Group—one of the principal institutions that regulates nuclear sales—enabled the United States and others to trade technology with India alongside carve-outs that permitted unsafeguarded Indian nuclear activity for weapons development. Then, the NPT’s 2010 Review Conference affirmed that states pursuing nuclear energy had an unconditional right to fully access nuclear technology, regardless of necessity.

The barrier between peaceful nuclear activity and weapons development is eroding.

The 2015 Iran nuclear deal, the Joint Comprehensive Plan of Action, capped Iran’s uranium enrichment program but walked back an earlier understanding (in the preliminary 2013 Joint Plan of Action) that Iranian enrichment activities would be limited to what Tehran needed for its peaceful program. Both the United States and Iran have since undermined the deal, leaving Iran, according to IAEA Director General Rafael Grossi, enriching uranium to concentrations that “only countries making bombs are reaching.”

There are other prominent carve-outs in the system. In 2010, China violated supplier rules by agreeing to construct additional nuclear power reactors in Pakistan, and in 2010 and 2017, Russia signed nuclear deals with Turkey and Egypt, respectively, without requiring (as far as is publicly known) that Ankara or Cairo forego developing fissile materials. Washington’s efforts to promote a “gold standard” for nuclear trade, in which recipients renounce the ability to produce enriched plutonium and uranium, foundered after it was introduced, in 2009, due to opposition from most prospective clients, including Saudi Arabia, and the U.S. government’s own desire to revitalize the domestic nuclear industry through power plant exports.

Collectively, this decay in the NPT’s rules has made it easier for nonnuclear states to obtain fissile materials, blurring the primary distinction between peaceful nuclear programs and military ones. Countries with weapons aspirations can now more easily hide their ambitions—and progress—in plain sight.


To stop proliferation and reestablish a pathway for rolling back nuclear weapons, states and nonproliferation institutions will need to take multiple steps. They can start by promoting better norms around the acquisition of enriched uranium. Responsible countries should unilaterally commit to avoid procuring or enriching nuclear materials not needed for peaceful purposes. States that insist on exercising their “inalienable right” to possess sensitive nuclear materials and technologies for peaceful purposes should affirm their bona fides by undertaking additional transparency and verification measures, such as implementing the IAEA Additional Protocol, which requires states to provide the agency with far greater access to and information about their nuclear activities. They should also accede to and properly implement every nuclear convention and best practice in the areas of nuclear safety, security, liability, and environmental protection.

To enforce these practices, the international community could use positive and negative incentives, including trade rewards and penalties, to induce compliance. Over time, states could codify such behavior in bilateral cooperation agreements and in the outcomes of multilateral meetings, such as G-20 communiques, NPT Review Conference statements, or UN Security Council resolutions.

Although it may seem to run contrary to the aims of nuclear nonproliferation overall, the new AUKUS deal offers an opportunity to strengthen the nonproliferation system. Australia is a state that had past interest in nuclear weapons but has evolved over the years into a champion of nonproliferation with an unblemished track record. It would therefore be well within Australia’s character and interest to assume unprecedented formal constraints and transparency measures on its indigenous nuclear pursuits both within and outside the submarine deal in exchange for the new security guarantees that the deal provides. These constraints would not only reassure other states about Australia’s exclusively peaceful nuclear intentions but also establish an approach for others to emulate if they pursue nuclear-powered submarines. 

Nonproliferation institutions must increase their tracking of missile development.

Norms, however, only go so far. Some countries will want to evade increased standards, and national intelligence agencies and international institutions—primarily the IAEA but also nuclear supplier regimes (such as the Nuclear Suppliers Group)—need better ways to evaluate the purpose of nuclear programs, especially now that the possession of fissile materials is no longer a sufficient or timely indicator of weapons intent. Thankfully, there are promising alternative metrics. For countries that acquire fissile materials, the primary technical barrier remaining to nuclear weapons acquisition is creating a workable warhead. That means nonproliferation institutions must increase their tracking of weaponization and delivery vehicle development.

There are precedents for how they can do so. In 2006, for instance, the Security Council empowered a comprehensive IAEA effort to ascertain the nature of Iran’s nuclear program. As part of its work, the agency thoroughly investigated Tehran’s entire program to design, develop, build, and test the mechanisms that package fissile materials into a warhead. Future, similar efforts could be naturally folded into the IAEA’s State-Level Approach: a methodology for assessing nuclear and nuclear-related activity in a specific country using a broad range of information. Independent national intelligence efforts would complement and inform IAEA activities, and both could advance special data mining and analytical tools to help strengthen tracking and analysis, even when on-the-ground monitoring is limited.

Warheads are not the only item that states need to transform fissile materials into deliverable nuclear weapons. They must also build or acquire advanced missiles, and transnational institutions can focus on identifying which states appear to be pursuing nuclear-capable ones, concentrating especially on features that distinguish them from other types of launch vehicles, including those for satellites. An effective detection system would prove immediately valuable, given that a number of states, ranging from Saudi Arabia to South Korea, are making or acquiring the means to produce ballistic and cruise missiles capable of delivering nuclear bombs. Many of these same states have also flirted with developing nuclear weapons.

There are additional indicators that nonproliferation monitors can use to distinguish peaceful nuclear programs from military ones. A working nuclear arsenal requires a setup beyond just warheads and missiles, including an elaborate system of command and control; rigorous personnel screening; nuclear safety procedures and emergency response capabilities; exceptional security measures; employment and use doctrines; trained personnel; and dedicated facilities that can store, maintain, and launch weapons. All of these activities leave traces, and as international monitors evaluate nuclear programs, they should seek to assess whether governments are advancing several of these capabilities, as well.


Making this enhanced nonproliferation system work won’t be easy. Australia, the United Kingdom, and the United States will first need to adopt new, specific nuclear monitoring measures that go beyond AUKUS. They will then need to secure the cooperation of China, Russia, and other countries that have nuclear weapons to make sure they all adopt the same rigorous practices and support international efforts to institutionalize them. And it will be difficult to get nonnuclear states to agree to new transparency expectations let alone empower any international body to routinely collect more information about programs with possible weapons aims.

But that doesn’t make it impossible. States might find more monitoring acceptable if it is incorporated into the IAEA’s State-Level Approach and used to inform the agency’s routine work surrounding safeguards, with additional inquiries only on a case-by-case basis. These measures could all eventually be enshrined by the IAEA Board, the annual NPT Review Conference, the G-20, and in UN Security Council resolutions. To obtain buy in from nonnuclear states that might be subject to new levels of scrutiny, leading nuclear powers could guarantee that they will not push for any new international legal instruments or institutions; the revised regime would run entirely through existing frameworks. They will also have to make clear that nonweapons states will not be subject to any new expectations, so long as they do not scale up uranium enrichment. For states that do opt to use enriched uranium, the expectations of greater forbearance and transparency will be tied to the scope and scale of their proliferation activities.

Finally, leading nuclear powers could provide new incentives to adhere to the enhanced framework by expanding export policies and assistance programs to states seeking nuclear technology for peaceful purposes. Liberalizing exports of nuclear material might seem to go against the aims of nonproliferation, but it could serve as a powerful inducement for nonproliferation compliance and also enhance the international response to the climate crisis. Nuclear power has been identified as an important element in the transition to renewable energy, but it presents difficult financial and technological hurdles for many countries.

Before the nuclear regime frays further, states must make an urgent effort to reinvigorate the NPT bargain. This work is essential both to stopping and rolling back future proliferation and to eventual disarmament—in other words, to a world in which nuclear technology doesn't contribute to geopolitical instability.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • TOBY DALTON is Senior Fellow and Co-Director of the Nuclear Policy Program at the Carnegie Endowment for International Peace.
  • ARIEL (ELI) LEVITE is Senior Fellow in the Nuclear Policy Program and Cyber Policy Initiative at the Carnegie Endowment for International Peace.
  • More By Toby Dalton
  • More By Ariel Levite