Since its announcement in late 2021, the AUKUS security partnership has sparked heated debate about its impact on global security. Critics of the partnership argue that it would provide nuclear-powered submarines fuelled with high-enriched uranium to Australia, a non-nuclear weapon state under the Treaty on the Non-Proliferation of Nuclear Weapons (NPT). Non-nuclear weapon states can conclude a so-called Article 14 arrangement in such situations, which means that routine safeguard measures by the International Atomic Energy Agency (IAEA) to ensure that the fuel is not diverted for the production of nuclear material for a weapons programme would temporarily not be applied. Some states have called this a nuclear proliferation risk.

The political and legal considerations in Article 14 arrangements have been, and continue to be, extensively discussed. Relatively little attention has been paid to the technical factors related to the nuclear-powered submarine programme that would influence an Article 14 arrangement. Exploring technical issues shows that the main potential proliferation risks associated with an Article 14 arrangement are located outside of the actual use of nuclear material to fuel the submarine, and that the IAEA will need to ensure that classification concerns do not stand in the way of adequate verification measures during this period.

Article 14 and diversion

Article 14 refers to a standard part of the safeguards agreement that non-nuclear weapon states must conclude with the IAEA. Under an Article 14 arrangement, routine safeguards procedures are not applied to nuclear material to be used in non-proscribed military activities (as opposed to the proscribed use as nuclear explosives) since applying them would reveal classified military information. They are replaced by other measures that allow the IAEA to provide credible assurance that this nuclear material is not diverted. When evaluating the risk of diversion, much of the current literature focuses on the scenario where a state uses the non-application of safeguards as an opportunity to covertly remove the nuclear material from the submarine.

Looking at technical issues shows the challenge associated with such diversion. In the case of AUKUS, to remove nuclear material, the metal submarine hull designed to withstand tremendous water pressure would need to be cut open with heavy machinery. The submarine’s fuel would then be extracted from the reactor, requiring specialised facilities. Fuel for a nuclear submarine, however, cannot easily be used for the production of nuclear material for a weapons programme: it comes in the form of fuel rods surrounded by metal or ceramic cladding rather than the uranium or plutonium metal form used in weapons programmes. The uranium in this fuel would need to be chemically separated from other materials before it could be used to produce nuclear material for a weapons programme. All the above steps cannot be carried out quickly enough to outpace international reaction, so it would have to be done in covert facilities without alerting other states to the fact that a submarine worth billions of euros had disappeared and an underground weapons programme had been launched. Hatches in the hull can provide easier access to the nuclear material, but the fuel used by submarines with hatches consists of uranium that is lower enriched – and thus less proliferation-sensitive – than the uranium AUKUS submarines will use.

This suggests that two other technical issues will decide the diversion risks of an Article 14 arrangement. Firstly, how easy it is to use the fuel in question to produce nuclear material for a weapons programme. In addition to the ease of separating uranium from other materials mentioned above, this ease is determined by the enrichment of uranium. This refers to the percent of the total material that is fissile. Nuclear-powered submarines make use of uranium enriched to levels between around five and 97 percent, while weapons programmes generally require enrichment of 90 percent or higher. Secondly, how much access the state has to the type of nuclear facilities needed for the production of nuclear material for a weapons programme. Enrichment and reprocessing facilities play a key role in this regard.

The ability of the IAEA to carry out verification related to these two technical issues may be limited by classification concerns. Knowing the technical specifications of submarine fuel can help outsiders deduce what the submarine’s capacities, such as speed or operational range, might be. To avoid this, states may try to limit verification measures that could reveal technical specifications, such as routine safeguards. This could also apply to activities outside of the fuel’s use in the submarine, for example when the fuel is being fabricated.

What diversion risks should Article 14 discussions focus on?

Considering the above technical concerns, three main diversion risks present themselves. First, a state could use an excuse to remove nuclear fuel from the submarine when it returns to port. For instance, the state could claim that the submarine is undergoing maintenance unrelated to the nuclear material, which would reveal classified information if observed. A believable excuse may allow the state to gain a head start in the lengthy process of removing nuclear material described earlier by reducing international scrutiny.

Second, a state could attempt to divert nuclear material that is still in the fuel cycle. If it successfully argues that safeguards should not be applied to some nuclear facilities, reduced oversight offers an opportunity: for instance, the state could try to divert nuclear material being converted into fuel.

Third, a state could use the nuclear-powered submarine programme as an excuse to develop its nuclear capabilities. If a state domestically produces fuel for a submarine that requires high-enriched uranium, it has a chance to build a reserve of nuclear material—not yet converted into submarine fuel—that could be diverted before the international community has an opportunity to respond.

These diversion risks suggest that an Article 14 arrangement should pay close attention to four key measures:

• There should be minimal and ideally no non-application of safeguards outside of the use of fuel in the submarine.
• Oversight should be given over the transportation of nuclear material, and its presence in facilities should be verified, including in a classified form.
• Verification measures should be carried out when nuclear material is placed in and removed from the submarine.
• The nuclear material’s presence in the submarine should regularly be verified.

Furthermore, discussions on Article 14 arrangements should consider a submarine programme’s impact outside the arrangement itself. In this context, any potential increase in a state’s ability to produce nuclear material for a weapons programme should be met with increased international monitoring.

What could the IAEA’s approach to Article 14 negotiations be?

The closer verification measures get to the finished form of the fuel and to the submarine, the more a state will object to them due to their potential to reveal information about the submarine’s operational capacity. When the IAEA pursues its goal of providing credible assurance that nuclear material is not diverted, the main obstacle it will encounter is the need to balance its objective with Article 14’s enshrinement of the protection of classified knowledge.

The IAEA can insist on at least the first three of the four points laid out above. The negotiations of the document on which Article 14 is based clearly established that the non-application of safeguards does not extend to activities that are not intrinsically military, specifically naming enrichment and reprocessing. While the status of fuel fabrication activities is less clear, this gives the IAEA solid arguments to apply safeguards to nuclear material when it is not used as fuel in the submarine, including during transportation between facilities. It also suggests that the IAEA should be able to verify that fuel has entered an intrinsically military activity, namely when it is installed in the submarine. Regarding the fourth point, it is unlikely that the IAEA will regularly be able to carry out verification measures in or around the submarine. However, seeing the submarine in operational use would confirm the presence of nuclear material on board. The IAEA could, therefore, seek to ensure that it can carry out some verification measures whenever the submarine returns to port for longer-than-usual periods of time, adjusted based on how long the extraction of nuclear material from the submarine is estimated to take.