ACA’s journal, Arms Control Today, remains the best in the market. Well focused. Solidly researched. Prudent.

– Hans Blix,
former IAEA Director-General

Can a Fissile Material Cutoff Treaty Be Effectively Verified?
Share this

John Carlson

Capping the production of highly enriched uranium (HEU) and plutonium for nuclear weapons has long been a goal of the international nuclear nonproliferation agenda because it has been seen as putting a real-world limit on the potential for any nuclear weapons buildup. During the last decade, this goal has, at times, appeared close to being realized.

Negotiation of a verifiable fissile material cutoff treaty (FMCT) was one of the central principles and objectives that helped achieve the indefinite extension of the nuclear Nonproliferation Treaty (NPT) at the 1995 Review and Extension Conference, and it was again endorsed by states-parties at the 2000 NPT Review Conference. The five nuclear-weapon states have also contributed to this effort as they are all understood to have ceased fissile material production for weapons purposes.

Yet, for the last few years, negotiation of such a treaty has been blocked by the failure of the Conference on Disarmament (CD) to agree on its broader work program. Recently, China, the United States, and other countries have advanced proposals that could provide momentum to move the process of negotiating an FMCT forward. Before diplomats reach that stage, however, a new potential obstacle looms: ensuring that the negotiating mandate still has consensus.

Currently, the negotiating mandate for the CD, drawn from a 1993 UN General Assembly resolution,[1] outlines the following elements: an FMCT is to ban “the production of fissile material for nuclear weapons and other nuclear explosive devices” and the treaty is to be “nondiscriminatory, multilateral and internationally and effectively verifiable.”

The United States had supported this mandate for the last decade, but in July, officials in the Bush administration concluded that effective international verification of an FMCT was not realistically achievable. The administration still supports the early conclusion of an FMCT to establish cutoff as an international norm but is concerned that perceived technical verification difficulties will seriously delay conclusion of the treaty. In their view, for example, potential sticking points include how to distinguish new production from pre-existing material and how to verify that production said to be for naval fuel is really being used for that purpose.

There is no reason, however, that these difficulties should lead to a deadlock. There is a way to address the United States’ desire to conclude a treaty quickly as well as to meet the broad desire to have a treaty with effective verification: separating the questions of political objectives and commitments and verification measures into two separate negotiations. In fact, there is already a model for this in the NPT.

Verification Considerations
Is there anything in the FMCT concept that makes it inherently incapable of effective verification? This depends on the objectives of an FMCT, as reflected in its substantive provisions yet to be negotiated. We must take care that questions of verifiability are not confused with differences over objectives, or issues of treaty architecture.

For example, the generally held FMCT concept does not proscribe production of additional nuclear weapons from unsafeguarded stocks of fissile material existing prior to an FMCT’s entry into force. Rather, the objective is to ensure these stocks are not enlarged. Some may question whether the existence of large stocks of fissile material outside an FMCT reduces the treaty’s usefulness, but this is an issue relating to the appropriateness of an FMCT’s objectives, not its verifiability.

The maxim underlying nuclear “peaceful use” treaties, the principal one being the NPT, is “trust but verify.” The subject matter of the NPT is of such fundamental importance to national and international security that no state is prepared to rely on trust alone. The existence of a credible verification mechanism in the form of International Atomic Energy Agency (IAEA) safeguards, comprising inspections, surveillance, and other measures, is essential to maintaining confidence in the effectiveness of the NPT and reinforcing the commitment of treaty parties.

Most states approach an FMCT with a similar perspective: it must have a verification mechanism to be considered credible. Moreover, there is no need to start from scratch. There is already a highly developed verification regime for nuclear material and activities; we have more than 40 years experience with IAEA safeguards. In addition, there is considerable experience with bilateral verification and confidence-building mechanisms, some directly applicable to nuclear weapons and sensitive materials.

Treaty Architecture
In the area of multilateral verification treaties, there are two alternative precedents. The first is a single treaty containing both the basic treaty objectives and commitments and the details of the verification system—the approach taken with the 1993 Chemical Weapons Convention (CWC). The disadvantages of this approach include the time required to negotiate the treaty—a U.S. concern in the case of an FMCT—and the degree of inflexibility in the verification system. It would be a major political exercise to update the verification system at a later date.

The alternative approach, which was demonstrated very successfully by the NPT, is to include the basic political commitments in a principal treaty and to set out the verification system in a secondary agreement or series of agreements (in the NPT’s case, each party concludes a safeguards agreement with the IAEA based on a model IAEA safeguards agreement). This approach separates the largely political from largely technical subject matters and allows for an adaptable verification system.

The NPT was concluded in 1968 and entered into force in 1970. A model safeguards agreement[2] was not concluded until 1972. Yet, in a demonstration of how such an approach pays off in the long term, these safeguards were then expanded in 1997 after revelations of Iraq’s pre-Gulf War nuclear program led to the successful negotiation of the Model Additional Protocol.[3] Another advantage of having separate negotiations on the technical details is that such negotiations can proceed quite expeditiously. Despite their complexity, these two model safeguards instruments each took only about 18 months to conclude.

Given the commonalities in subject matter between IAEA safeguards and FMCT verification, it would seem sensible to follow the NPT route, thus enabling the political commitment to be established quickly and the details of the verification system to follow.

Objectives of an FMCT
The basic objective of an FMCT would be to proscribe production of fissile material for nuclear weapons or other nuclear explosive devices. Accordingly, when the treaty enters into force, the parties would undertake not to produce fissile material for nuclear weapons. They would accept international verification on relevant facilities and nuclear material to verify this commitment. They would also pledge not to use any fissile material that is subject to verification under an FMCT for nuclear weapons, that is, the principle of irreversibility would apply and subject material could not be withdrawn for weapons use.

For the purposes of an FMCT, it is expected that “fissile material” would encompass separated (unirradiated) plutonium and HEU. It would also include uranium-233, designated for IAEA safeguards as a “special fissionable material,” and would probably also need to include separated neptunium because the IAEA has recognized that this material has the potential to be used in nuclear weapons.

Nonproscribed Activities
A final FMCT would permit production of fissile material for civil purposes and also for non-explosive military purposes such as naval propulsion, but only under verification to ensure fissile material is not diverted to weapons. The recycling (or cleanup) of plutonium from weapons—an established stockpile stewardship practice—that did not involve new production of fissile material would also be permitted.

States Affected
An FMCT would apply to three groups of states: non-nuclear-weapon states party to the NPT; the five NPT nuclear-weapon states; and the three nuclear-capable states outside the NPT—India, Israel and Pakistan.[4]
Non-nuclear-weapon states are already committed not to produce or use nuclear material for weapons purposes and to accept IAEA safeguards on all their nuclear material and activities to verify this commitment (comprehensive safeguards).[5]

In principle, therefore, an FMCT should not involve any additional commitments from non-nuclear-weapon states that are implementing both an NPT safeguards agreement and an additional protocol. The principal effect of an FMCT and its verification task relate mainly to the nuclear-weapon states and the non-NPT states.

Scope of an FMCT
A major issue to be resolved in the negotiations is the scope of an FMCT—the facilities and material to which verification would apply. There are two basic options. One path would be to negotiate a treaty of wide scope. This would cover all nuclear facilities and nuclear material except nonproscribed military activities such as naval propulsion and existing stocks of nuclear material. The other option would be to seek a more narrowly focused treaty, concentrating on the most proliferation-sensitive fissile material production facilities, i.e., reprocessing and enrichment facilities, and relevant production from those facilities.

There would be substantial problems in trying to extend the comprehensive safeguards system to the nuclear-weapon states and non-NPT states. It is impossible to apply truly comprehensive safeguards covering all nuclear material in the nuclear-weapon states and the non-NPT states because, when the treaty enters into force, they will still have nuclear weapons and be permitted to retain pre-existing nuclear material outside verification. Further, the cost of verification based on the comprehensive model in the nuclear-weapon states would be very high.

Taking these and related considerations into account, Australia has concluded that a separate and distinct FMCT verification regime would be required for the nuclear-weapon states and non-NPT states. These states would undertake the same commitment as the non-nuclear-weapon states, not to produce fissile material for nuclear weapons, but the verification approach would be more appropriate to their particular circumstances.

Accordingly, Australia has proposed a “focused” approach. One aspect would involve the monitoring of fissile material production facilities such as uranium-enrichment and reprocessing plants. The other would involve verification that fissile material subject to an FMCT—separated plutonium, uranium-233, HEU, and separated neptunium produced after the treaty’s entry into force—is indeed being used for peaceful or nonproscribed purposes. This would require verification measures at downstream facilities handling these materials after production.

In addition, an FMCT verification regime will need to include measures aimed at detection of possible undeclared production facilities.

A major issue fundamental both to the substance of an FMCT and to the prospects of successful negotiation is how to treat past production of fissile material, i.e., stocks existing when the treaty enters into force.

The issue of stocks does not arise in the case of states that have accepted comprehensive safeguards under the NPT. All holdings of nuclear material in these states are subject to IAEA safeguards. The issue relates to those states outside comprehensive safeguards and having the capability to produce fissile material—nuclear-weapon states and the non-NPT states.

The term “stocks” can have a wide meaning, ranging from fissile material in weapons to bulk material declared surplus to defense needs, and to civil stocks of fissile material. Clearly, an FMCT could not apply to all pre-existing stocks held by the nuclear-weapon states and the three non-NPT states as this would amount to instant disarmament—an unrealistic objective. It must be recognized that past production of separated plutonium, HEU, and so forth in the nuclear-weapon states and non-NPT states would be outside verification. The task of verification will be to provide assurance that there is no new, undeclared production.

If the parties so wished, however, an FMCT could contain a mechanism by which parties could place surplus and civil stocks under the treaty in accordance with the principle of irreversibility. Some parties may wish to do this as a mutual confidence-building measure.

Verification Methods
A verification model for an FMCT has yet to be established. Based on experience from IAEA safeguards, however, an effective regime can be expected to comprise three basic elements: routine verification activities for declared facilities and material; verification activities aimed at detection of possible undeclared fissile material production; and complementary measures aimed at transparency and confidence building.

One of the issues to be decided in the negotiation process is the verification agency. Given the commonalities between FMCT verification and IAEA safeguards, it would make sense for the IAEA to verify an FMCT, but at present, there is no consensus on this.

Declared Activities
In concept, this aspect would be very similar to IAEA safeguards. FMCT parties would be required to declare all relevant facilities such as enrichment and reprocessing facilities and downstream facilities handling subject material.
Declared facilities would be monitored through inspections, containment and surveillance, and other measures to verify that there is no undeclared production of fissile material and that declared fissile material is not diverted to nuclear weapons (or purposes unknown).

For enrichment plants, verification would be applied to all facilities, including those producing low-enriched uranium (LEU), to ensure there is no undeclared HEU production. In principle, verification would not need to be applied to LEU, but in view of the advantages of LEU as a feedstock for undeclared HEU production, some verification measures for LEU may need to be considered, particularly in the case of states with smaller arsenals.
To ensure that the use of plutonium conforms to treaty commitments, verification would be applied to separated plutonium product leaving a reprocessing plant, as well as to any facilities in which plutonium separated after the treaty enters into force is present. Verification would cease to apply once plutonium has been returned to a reactor as mixed-oxide fuel and irradiated, because irradiated plutonium is of no further strategic value until it has been reprocessed.

Undeclared Activities
Here, two broad forms of verification activity are envisaged: routine activities aimed at evaluating the completeness and correctness of FMCT declarations and inspections based on suspicion of a breach of FMCT commitments.

A major challenge for IAEA safeguards is the detection of undeclared nuclear facilities, particularly centrifuge-enrichment facilities. In the past, IAEA safeguards techniques for the detection of undeclared facilities were limited, and detection of undeclared nuclear material was seen as the major indicator of the existence of undeclared facilities. Now, the increasing availability and capability of techniques for detection of undeclared facilities has led to a revolutionary change in safeguards. In the ongoing program to strengthen IAEA safeguards, emphasis is being given to detection of undeclared facilities.

A whole suite of new measures is being established, including more effective information collection and analysis; satellite imagery; and, through the Additional Protocol, wide-ranging complementary access to apply verification measures such as environmental sampling. Another technique under study is wide-area environmental monitoring. It can be expected that methods similar to these would apply under an FMCT.

An issue raised for an FMCT is how to recognize noncompliance. Because the nuclear-weapon states and non-NPT states already have undeclared material, how would inspectors be able to determine if any particular material was produced before or after the treaty’s entry into force? There are some direct technical answers to this question— nuclear material can be dated—but, as is increasingly the case with IAEA safeguards, the principal focus of FMCT verification to counter undeclared activities would likely be the detection of undeclared enrichment and reprocessing facilities.

Detection of undeclared enrichment and reprocessing facilities will no doubt be a challenge for FMCT verification as it is with IAEA safeguards, although in some respects the problem is more manageable for an FMCT. The relevant states possess extensive intelligence information on each other, and most have a limited motivation to cheat as they have nuclear arsenals they consider adequate and in some cases are actively reducing. The more difficult cases might be states with small arsenals, such as India and Pakistan, where bilateral confidence-building measures could have an important role in complementing international verification.

Similar to the Additional Protocol, under an FMCT, the verification agency should have the right to request access to locations to resolve questions and inconsistencies arising from information analysis. In addition to access initiated by the verification agency, FMCT parties may wish to have a verification mechanism they can initiate directly, along the lines of the CWC’s challenge inspection mechanism.

The challenge inspection mechanism reflects that the CWC verification agency—the Organization for the Prohibition of Chemical Weapons (OPCW)—does not look for undeclared chemical facilities. CWC parties are expected to use national means for this: if a party becomes aware of a suspect site, it can request that the OPCW undertake a challenge inspection. However, a challenge inspection can be blocked by a three-fourths vote of the CWC Executive Council. Given that to date the CWC’s challenge inspection mechanism has not been used, for an FMCT a “technical” inspection mechanism that can be initiated by the verification agency might be seen to have greater utility.

National Security Aspects
The nuclear-weapon states and non-NPT states are concerned about protecting national security and proliferation-sensitive information relating to past nuclear weapons programs and ongoing stockpile stewardship activities. Hence, verification methods and procedures will need to be carefully defined. Managed-access provisions will be essential, probably elaborated in greater detail under an FMCT than in the Additional Protocol.

Open Skies and Regional Access Arrangements
In addition to measures drawn from IAEA safeguards experience, other approaches that could be relevant to an FMCT include open flight corridors for verification purposes and bilateral or regional access arrangements. Such measures could have an important confidence-building role, complementing international verification.

Verification of HEU Produced for Naval Propulsion
This is often raised as a problem for an FMCT, because the design of naval fuel is highly classified. Appropriate verification is required to ensure this does not become a route for diversion. This is complex but not insurmountable—the Trilateral Initiative among the United States, Russia, and the IAEA demonstrates the practicability of innovative approaches to verifying fissile material of sensitive composition, shape, and mass. Under the Trilateral Initiative, verification methods have been developed by which IAEA inspectors can determine that fissile material inside containers has certain attributes (e.g., plutonium within an agreed isotopic range, a mass of at least an agreed value) without the inspectors seeing the material or ascertaining classified characteristics. For naval fuel, it should be possible to devise verification approaches that do not reveal classified information.

Verification Intensity
Although some of the methods of IAEA safeguards are readily applicable to FMCT verification, the treaty objectives are very different. This would be reflected in the design of an FMCT verification regime.

Comprehensive safeguards are designed to provide assurance against horizontal proliferation, that is, the acquisition of one or more nuclear weapons by a non-nuclear-weapon states. The acquisition of just one or two nuclear weapons would be a dramatic change in the strategic status of such a state. The sensitivity of IAEA safeguards, reflected in technical parameters such as goal quantities (e.g., the significant quantity of 8 kilograms plutonium), detection probability, timeliness goals, and inspection frequency, has been set accordingly.

On the other hand, with states outside comprehensive safeguards, the concern is essentially vertical proliferation, that is, additions to existing arsenals. An FMCT would address this by providing assurance that stocks of fissile material held outside international verification will not increase. For states that already hold nuclear weapons, their concern will be with treaty violations that could substantially alter strategic relativities: for states with thousands of weapons, a strategically significant violation might involve hundreds of weapons. Thus, the design of verification approaches to deter vertical proliferation could be qualitatively different to those directed at horizontal proliferation.

A major development in IAEA safeguards implementation is a move away from uniformity to safeguards based on a “state-level approach,” taking into account appropriate state-specific considerations. This development remains a work in progress,[6] but the concept is readily applicable in the FMCT context.

Decisions on verification intensity could consider several factors, particularly the true risks of vertical proliferation. Because the nuclear-weapon states will retain military stocks outside an FMCT, it is reasonable to assume they will have concluded, before joining the treaty, that those stocks are sufficient for their foreseeable needs. Hence, they should have little incentive to cheat. This is especially the case for the United States and Russia, which are dramatically reducing nuclear-weapon numbers and are committed to transferring substantial quantities of fissile material out of weapons programs irreversibly. In these circumstances, rigorous verification is not required. On the other hand, for states with small arsenals, verification intensity will need to reflect the fact that small-scale violations could have a serious effect on strategic relativities.

The United States, Russia, France, and the United Kingdom have announced cessation of fissile material production for nuclear weapons, and it is understood China has also ceased such production. An FMCT would make an important contribution to nuclear nonproliferation by formalizing this situation and making it irreversible. A further important contribution would be capping the fissile material available in the non-NPT states, which are otherwise under no restraint. An FMCT would also help establish conditions under which further nuclear disarmament, involving all relevant states, would be possible—a significant consideration in the context of the NPT review process.

Achieving these benefits will require the treaty to be effective, which requires a credible verification regime. There is a substantial foundation on which to build, drawing on experience from IAEA safeguards and bilateral verification arrangements.

Undue delay in concluding a normative treaty can be avoided by separating negotiation of the principal treaty from negotiation of the verification system. The latter would be a largely technical negotiation, which can commence in parallel with the principal negotiation but be concluded subsequently.

Whether a particular verification regime provides the degree of assurance required by the parties is a matter for judgment based on many factors: the verification objectives, the verification methods and standards, related confidence-building measures, other information including intelligence available to the parties, incentives and deterrents reinforcing compliance, and so on. Only when we have defined the objectives and main features of FMCT verification will it be possible to design the verification system and to judge whether it will be sufficiently effective, but there seems no reason in principle why this should not be achievable. ACT


1. UN General Assembly Resolution 75, General and Complete Disarmament, A/RES/48/75/L, 1993.

2. International Atomic Energy Agency (IAEA), The Structure and Content of Agreements Between the Agency and States Required in Connection With the Treaty on the Non-Proliferation of Nuclear Weapons, INFCIRC/153, 1972.

3. IAEA, Model Protocol Additional to the Agreement(s) Between State(s) and the IAEA for the Application of Safeguards, INFCIRC/540, 1997.

4. In 2003, North Korea announced its withdrawal from the NPT. The validity of this withdrawal has not been determined, and for the purposes of this discussion, it is assumed that North Korea is still bound by the NPT.

5. The safeguards applying to non-nuclear-weapon states-parties to the NPT used to be termed full-scope safeguards, but the usual term now is comprehensive safeguards.

6. For a discussion of some ideas in this area, see Annette Berriman, Russell Leslie and John Carlson, “Assessing Motivation as a Means of Determining the Risk of Proliferation,” 2004 Annual Meeting of the Institute of Nuclear Materials Management.


John Carlson is director-general of the Australian Safeguards and Non-Proliferation Office. He is also chairman of SAGSI, the International Atomic Energy Agency’s Standing Advisory Group on Safeguards Implementation. This paper reflects the views of the author and does not necessarily represent the views of the Australian Government or of SAGSI.

Posted: January 1, 2005