Since the dawn of the nuclear age in 1945, global efforts to prevent proliferation of nuclear weapons centered on denying, restricting, or conditioning access to fissile materials by countries without nuclear weapons. Stemming the spread or acquisition of plutonium or highly enriched uranium facilitated a global trade in peaceful uses of nuclear technology for energy, research, medicine, agriculture, and industry, as countries that already possessed nuclear weapons supplied materials and technology. This basic bargain was enshrined in the Treaty on the Non-Proliferation of Nuclear Weapons (NPT) and implemented through the multilateral Nuclear Suppliers Group and other arrangements that together comprise a global nonproliferation regime.

This global regime has become riddled with exceptions, however, most of which are associated with the rules governing access to fissile material and related production technologies. Many of the carve outs were authored by the U.S. government as ad hoc measures to address regional policy compulsions and the interests of friends and allies at the expense of the universality of international rules. For example, the United States exempted India from policy restrictions on trade with countries that do not implement full International Atomic Energy Agency (IAEA) safeguards. It also permitted Japan to reprocess U.S.-supplied nuclear fuel, a permission that the United States otherwise withholds from countries without nuclear weapons.

The latest example of this nuclear exceptionalism is the security pact between Australia, the UK, and the United States (AUKUS), under which Canberra will receive from London and Washington the technology and fissile material for a nuclear-powered submarine program. Until now, naval nuclear propulsion has been the exclusive purview of countries with nuclear weapons. These countries keep their submarines and associated fissile material facilities outside of monitoring by the IAEA. Australia is expected to exercise a clause in its IAEA safeguards agreement that will exempt the submarine program from routine monitoring on the grounds that the program is for military purposes.

Ariel (Eli) Levite
Levite was the principal deputy director general for policy at the Israeli Atomic Energy Commission from 2002 to 2007.
More >

AUKUS therefore creates two related precedents that could be leveraged by countries that want to pursue nuclear weapons in the future. First, it will allow a state without nuclear weapons to exempt some activities and nuclear materials from the obligation to apply comprehensive IAEA safeguards arrangements over its entire nuclear program. Second, it will enable a state without nuclear weapons to utilize highly enriched uranium (a material that could be directly used to make nuclear weapons) for military purposes.

Most analysts do not expect Australia to seek nuclear weapons, but the precedent that AUKUS sets could be cited by others—most immediately and worryingly Iran—to use a naval nuclear propulsion program as cover for enriching uranium to a weapons-grade level and exempting it from IAEA safeguards. This would create a substantial nuclear weapons breakout capability. If protective measures are not taken, AUKUS could be used to legitimize activities that cross the primary boundary of nuclear weapons acquisition—possessing fissile materials.

Until now, Washington’s approach to managing the AUKUS precedent has been to argue that the supply of fissile material for nuclear-powered submarines is a singular initiative, only for Australia. Yet, Brazil is actively working on a nuclear-powered submarine; South Korea plans to build one and is seeking international assistance; Japanese leaders have expressed interest in nuclear submarines; and Iranian government officials periodically float the idea publicly.

Bearing in mind the potential ramifications of this precedent, Washington—alongside Canberra and London—needs to think beyond the confines of the trilateral initiative. AUKUS presents opportunities to reimagine the nonproliferation regime in ways that would better match evolving conditions and to develop new norms, tools, and practices that would help detect and deter future proliferators.

Toby Dalton
Dalton is the co-director and a senior fellow of the Nuclear Policy Program at the Carnegie Endowment. An expert on nonproliferation and nuclear energy, his work addresses regional security challenges and the evolution of the global nuclear order.

Using AUKUS to strengthen nonproliferation would require skillful diplomacy and creative technical spadework. First, AUKUS parties (and the IAEA) should agree on transparency and other behavioral restraints by Australia that set expectations for other countries that pursue naval nuclear propulsion. Second, the United States and like-minded partners should develop and promote new nuclear supply initiatives and practices that both make nuclear technology more available and clarify conditions for fissile material possession. And third, the IAEA and government partners should build a more holistic approach to assessing future nuclear proliferation.

Steps for Australia

Technical, logistical, and other challenges mean that Australia will not operate its own nuclear-powered submarines (SSNs) until well into the 2030s, at the earliest. However, as Australia begins to build the physical and institutional infrastructure for its SSN program, AUKUS partners can help to author best practices that would address concerns about proliferation potential. These could then be adopted by other countries that would seek to exempt fissile material from safeguards for use in non-weapons-related military programs. These practices would be all the more critical for cases in which the naval propulsion program would utilize highly enriched uranium fuel, as appears to be the case with AUKUS.

First, AUKUS partners should articulate—and Australia should meet—nondiscriminatory requirements for military uses of fissile material outside of IAEA safeguards. Foremost among these is ratification and complete implementation of the IAEA’s Additional Protocol and the receipt of a broader conclusion, which would affirm that Australia’s nuclear materials are utilized exclusively in peaceful activities. These should reinforce Australia’s clean proliferation track record and its unblemished IAEA safeguards compliance (affirmed in annual IAEA internal reports) in recent decades. Australia should further stipulate its commitments not to re-transfer any fissile material and not to process any fuel it receives into material that could be used in nuclear weapons. Finally, the UK and the United States should agree to only transfer fissile materials in finished forms, such as fabricated nuclear fuel, and to require that handling of any spent submarine fuel not only meet strict safety and environmental requirements but also preclude any Australian reprocessing of the spent fuel.

Second, Australia should undertake enhanced engagement and transparency with the IAEA. Such cooperation should facilitate periodic updates by the agency to Australia’s safeguards’ broader conclusion. Building on this, the IAEA Secretariat should establish a practice to augment the frequency and scope of its reporting to its Board of Governors for countries that exempt fissile material from safeguards.

Third, in addition to enhancing transparency of its nuclear activities, Australia could pioneer best practices by committing to refrain from expanding or upgrading its nuclear material processing capabilities and foreswearing research, testing, and development activities for nuclear weapons design and manufacturing.

Australia regularly reiterates its firm commitment to the NPT and its policy to forego nuclear weapons acquisition, and it is not considered much of a proliferation risk today. Yet it is worth recalling its past investigation into nuclear weapons capabilities and the development of a laser-based uranium enrichment program. Abjuring activities that could be seen to revive a latent nuclear weapons enterprise would demonstrate Canberra’s unwavering commitment to maintain strong nonproliferation credentials and set an important precedent for others to follow, consistent with Canberra’s exemplary nonproliferation stewardship of recent decades.

Nuclear Supply Initiatives

Further diplomacy to translate steps by Australia into broader initiatives would be required to offset potential nonproliferation damage caused by AUKUS precedents. Canberra, London, and Washington should work with like-minded countries to advance new normative and regulatory initiatives in several areas.

To start, nuclear suppliers should augment international nuclear trade practices to address the erosion of the fissile material threshold. For example, suppliers could agree on a national basis to use bilateral nuclear cooperation agreements as a vehicle to establish new methods for transparency and reassurance—basically, to build upon the know-your-customer principle. State nuclear authorities and involved vendors could engage in reassurance dialogue regarding any concerning, anomalous behaviors or activities observed in a recipient state. If concerns could not be addressed, companies could convey relevant information to their national governments and the IAEA for further examination in the context of the recipient state’s safeguards activities. Another type of new trade practice could involve suppliers and their supply chain vendors working with recipient countries to incorporate nonintrusive monitoring technologies into products and systems used in the nuclear fuel cycle.

To be sure, any transparency or reassurance effort perceived as creating new trade barriers or conditions is likely to be rejected by suppliers and recipients alike—as seen in the attempt to institute a “gold standard” in which countries would agree to forego domestic enrichment and reprocessing activities. Thus, new nuclear trade initiatives would need to avoid additional constraints that are not tied to the liberalization of some conditions of transfer. In this regard, renewed interest in nuclear power in the context of climate change and the potential development of a market for small reactors could help motivate such a bargain.

Second, like-minded countries should build normative clarity around fissile material practices by countries without nuclear weapons. For instance, without infringing on rights to peaceful uses of nuclear technology, countries could identify a set of behaviors that would be understood to be concerning (or even raise alarms) or which should trigger additional transparency or reassurance requirements. For example, enrichment at a scope that lacks any credible peaceful use should invite further scrutiny, as well as the enrichment of uranium beyond 20 percent U-235, accumulation or stockpiling of low-enriched uranium (outside fuel rods) in quantities sufficient to build one or more nuclear weapons, or converting highly enriched uranium into metal (the form in which it is used in nuclear weapons).

Third, leading nuclear states should step up efforts to expand adherence to and associated reporting under core international agreements and of international codes of conduct by all countries with peaceful nuclear programs. These include the Convention on Nuclear Safety, the Convention on the Physical Protection of Nuclear Material, the Vienna Convention on Civil Liability for Nuclear Damage, and others pertaining to environmental protection and early notification of incidents. Expanded adherence and enhanced compliance with these instruments (including in meeting their respective transparency modalities) could improve overall nuclear safety and security practices but also help highlight anomalous behavior that might be indicative of proliferation interests.

Reimagining Nonproliferation

The precedent set by AUKUS also requires some thinking about how to address other nuclear-weapons-related activities. This area holds the most opportunity—and also the greatest challenges—in reimagining a nonproliferation regime for the future.

For the last sixty years, the primary proliferation warning indicators were fissile material acquisition activities. This warning worked well in a world in which very few countries without nuclear weapons undertook such activities. But if even a handful of additional countries follow the AUKUS precedents in the future, then fissile material will become less behaviorally relevant to generating timely proliferation warnings. This should provoke new modes of analysis and new ways of thinking about nonproliferation norms and the corresponding requirements for rollback and disarmament cases.

There are several possible ways to compensate for the decreased salience of fissile material indicators for proliferation analysis. One is to complement existing models of nuclear weapons acquisition analysis conducted by the IAEA under its state-level approach—which focuses on connections between individual activities using the so-called Physical Model)—with analysis that broadly compares peaceful nuclear-related activities in a country against generic benchmarks for such activities. Such analysis would turn up programmatic anomalies. For example, an anomaly might involve a fuel cycle activity that has no clear linkages to existing or planned programs, such as conversion of uranium into metallic forms that have no connection to existing or planned nuclear reactors.

A more ambitious approach would be to develop a comprehensive model that focuses on other, non-fissile-material elements of nuclear weapons programs. This would include research and development of nuclear explosion capabilities, weapon systems integration, nuclear-capable delivery vehicles, and changes in military orientation and infrastructure preparatory to the induction of nuclear weapons. Such an approach, considered further in Carnegie’s nuclear firewall project, would enable assessing the consistency of these activities with their stated purpose and their cohesion in relationship to nuclear weapons development. This type of modeling could be utilized by national governments or by the IAEA.

More ambitiously, countries could begin to build a more holistic nonproliferation system by translating this broader set of indicators into norms of restraint. These restraints would be especially apt in countries that pursue advanced fissile material capabilities. Australia could be a leader in this effort by, for example, voluntarily committing to pertinent transparency measures in these domains and foregoing certain weapons research and development activities, namely those that were included in Section T of the Joint Comprehensive Plan of Action, which capped Iran’s uranium enrichment program prior to the deal’s unraveling following U.S. withdrawal.


To prevent future proliferation, it is incumbent on the United States and its partners to take steps along the lines described here. If done well, they could mitigate the potential deleterious effects of the fissile material precedents set by AUKUS and even use this initiative as an opportunity to build a better nonproliferation regime for the future. Heading into the January 2022 NPT Review Conference, a positive agenda like this—which applies not just to nonproliferation but also to setting benchmarks that could be applied in future cases of nuclear rollback or disarmament—could help to assuage concerns about further nuclear exceptionalism.

There are two other important reasons to advance this agenda. First, there is value in reassuring China and others that this cooperation will not enable Australia to deploy nuclear weapons on its submarines in the future. And second, far from being interpreted as another effort to restrict technology sharing at a time when nuclear power is increasingly a consideration to mitigate climate change, these ideas could attract support by major vendor countries (such as China, France, Japan, Russia, South Korea, and the United States) and their clients as part of a package to liberalize nuclear trade.