New Paradigms Forum Proliferation Issues and Much More …


Iran, Nonproliferation, and the IAEA: A Legal History


This post follows up on the NPF posting of November 5, 2012, on the colloquy between Dr. Ford, Professor Daniel Joyner, and Andreas Persbo on the legal standards applied by the IAEA vis-à-vis Iran.  Dr. Ford published an additional paper on this subject in early November, and NPF provides its text here for readers' convenience.  The original paper, in PDF form, may be found on the Hudson Institute website.  Its text -- absent citations to the relevant underlying documentation -- is reproduced herein below. (Warning: It is rather long ....)


Ever since the first public revelations of Iran’s previously entirely secret nuclear program in 2002, the International Atomic Energy Agency (IAEA) has struggled to verify and document Iran’s degree of compliance with its safeguards obligations – as well as, for most of this period, Iran’s degree of compliance with its further claims or promises, and the additional obligations imposed upon it by multiple resolutions of the United Nations Security Council acting under Chapter VII of the U.N. Charter.  The IAEA and its member states have evaluated Iran’s behavior against a range of legal standards as the country’s obligations have developed.

With events in Iran at present (October 2012) giving every appearance of moving toward some kind of denouement in the next several months, it is important that this legal background be properly understood.  To be sure, the fundamental issues driving international reactions to Iran’s nuclear program are policy ones, being rooted in threats to international peace and security that are already presented by the clerical regime in Tehran, but which would be greatly worsened were that government empowered by nuclear weapons possession (and one or more of its neighbors spooked into engaging in nuclear weapons proliferation themselves).

The specifically legal issues and standards are hardly unimportant, however, and form a noteworthy aspect of the crisis that is seldom discussed or understood outside a narrow clique of diplomats, government attorneys, legal academics, and nonproliferation experts in the non‐governmental community.  The legal standards applied vis‐à‐vis Iran, moreover, have occasionally been contested – by Iran, at the very least, and sometimes by foreign observers who have tried to support Iran’s position against the IAEA.  Such attacks have so far not been persuasive, but they have an obvious and important bearing on the legitimacy of the international case against the Iranian regime. With the Iranian situation seemingly at a near‐breaking point, it is important, therefore, that these legal matters not be misunderstood.  This paper seeks to set forth and trace the development of the key legal issues that have arisen in the Iranian nuclear case.

I.           Nuclear Safeguards Agreement

After news surfaced in August 2002 of the hitherto clandestine enrichment plant that Iran was building at Natanz – as well as of the heavy‐water production plant it was building at Arak, apparently to supply a future plutonium‐production reactor – the Iranians conceded the existence of these facilities and formally declared them to the IAEA in February 2003.  At this point, the basic challenge for the IAEA became one of evaluating Iran’s compliance with its Comprehensive Safeguards Agreement (CSA), which had been signed by Iran, ratified by the Iranian parliament, and entered into force many years earlier.

A.           Safeguards Compliance

Much of the Agency’s investigations in the succeeding months thus concerned documenting the various ways in which Iran had failed to live up the obligations set forth in its CSA – such as, for instance, by failing to declare to the IAEA its importation in 1991 of a shipment of uranium hexafluoride (UF6) and other uranium compounds, a shipment which it declined to reveal to the IAEA until February 26, 2003.  In some instances, the Iranians tried to argue that the IAEA had no jurisdiction to look into certain questions because no nuclear material had been involved and the Agency’s jurisdiction vis‐à‐vis Iran at that point went to no more than verifying Iran’s declarations of declared material and facilities.  Embarrassingly for Tehran, however, many of these claims simply turned out to be lies.

Iranian officials at first declared, for instance, that since no nuclear material had been present in Iran that had not already been declared, the IAEA had no jurisdiction to inspect (as it had requested, apparently on the basis of intelligence information supplied by one or more member states) the Kalaye Electric Company facility or any other locations.  Even though the Iranians delayed the Kalaye visit while they conducted extensive renovations there in apparent bid to sanitize it, IAEA environmental sampling duly revealed the presence of particles of low‐ enriched uranium (LEU) and high‐enriched uranium (HEU).  Iran subsequently admitted that it had indeed tested centrifuges at Kalaye using its undeclared UF6 shipment, as well as to having pursued a secret laser enrichment program elsewhere, using uranium metal it had not previously declared, and conducting secret experiments with plutonium separation. (Iran also tried to conceal some of the tests at Kalaye by falsely claiming uranium loss due to “evaporation” from leaky valves.)  All of this was previously unreported to the Agency, but should have been.  Despite claims that its nuclear work had not involved nuclear material – and thus did not fall within the IAEA’s jurisdiction6 – it turned out that Iran had been committing multiple violations of its safeguards obligations for many years.

The information developed by the IAEA in Iran formed the basis of the IAEA’s subsequent finding of Iranian noncompliance.  It was, for example, found that Iran had failed to meet its obligations under the CSA with regard to reporting the possession, processing, and use of nuclear material, and with regard to declaring facilities where such material was processed and stored. (Additionally, the IAEA faulted Iran for its systematic concealment, provision of misleading or false information, and its lack of cooperation on safeguards verification.)  All in all, “[m]any aspects” of Iran’s nuclear program turned out to have been “not declared to the Agency in accordance with Iran’s obligations under its [Comprehensive] Safeguards Agreement,” and this “resulted in may breaches of its obligation to comply with that Agreement.”

The IAEA appears to have applied correct legal standards and fulfilled its own obligations up until the point at which noncompliance had been identified. Iran’s CSA requires it to provide the Agency with information concerning nuclear material subject to safeguards, and with information about the features of facilities relevant to safeguarding such material.  Imports of nuclear material must be declared no later than the time of their arrival, material must be placed under safeguards, and current information must be provided to the IAEA on the location and use of material used outside declared facilities.  (Inventory changes must also be declared.)  All of these provisions were flagrantly violated by Iran for years, until Tehran’s belated admissions in response to public leaks and the findings of the IAEA’s own investigations in Iran. The Agency’s finding of noncompliance was amply justified and painstakingly documented.

B.           Reporting the “Iran Dossier”

At the point that there was a clear finding of safeguards noncompliance, however, things became somewhat more complicated. Pursuant to the IAEA Statute, Agency inspectors have as one of their responsibilities the role of “determining whether there is compliance” with health, safety, and other measures prescribed in an agreement with a state, phrasing which includes compliance with safeguards obligations.  In the event that noncompliance is indeed found, the Statute says that the Board of Governors “shall” report this to all IAEA members, as well as to the U.N. Security Council and General Assembly.

At least from the point that the Agency had indeed declared Iran to have breached its safeguards obligations, therefore, the IAEA was required by its own statute to report this fact to the United Nations.  Thanks to the efforts of Director General Mohammed ElBaradei and some members of the Board to protect Iran from the prospect of facing Security Council sanctions, however, this reporting was not agreed, notwithstanding the fact that this technically put the Agency in violation of its own statute for some two years until a Security Council referral finally occurred. The Board did not request the Director General to report Iran to the Security Council until February 2006.  This problem was not one of the IAEA applying the wrong legal standard vis‐à‐vis Iran, of course – in fact, the Agency was ignoring its own statute to Iran’s benefit – but it was the only point at which the Agency can properly be said to have been acting ultra vires during the entire Iran episode.

C.           The Design Information Dispute

During this early phase of the Iranian nuclear crisis, controversy arose over Iran’s obligation to provide design information on its previously‐secret nuclear facilities. For the most part this dispute has been a fictitious one, concocted by Iran in an effort to shield its nuclear program from scrutiny and perpetuated by apologists whose eagerness to exonerate Tehran and malign the IAEA has outrun their legal judgment.  Nevertheless, in some quarters the particular timetables for providing design information remain controversial, so the issue bears note here.

CSAs, Iran’s among them, require the application of nuclear safeguards in the country in question. Since such agreements cannot supply the level of detail needed in such a complicated business, however – and since it is frequently necessary to alter safeguards arrangements as facilities are constructed, material balances rise or fall, and other changes occur – CSAs provide for the negotiation (and re‐negotiation, as necessary) of “Subsidiary Arrangements” (SAs) between a country and the IAEA. Iran’s CSA covers this process in Articles 39 and 40, pursuant to which Iran and the Agency are to work out SAs that “specify in detail ... how the procedures laid down in this Agreement are to be applied.”  Because of the need periodically to adjust specific measures (e.g., the specific steps needed to safeguard whatever materials happen to be present at various facilities at any given time), these SAs “may be extended or changed by agreement between the Government of Iran and the Agency without amendment of [the CSA itself].”

“Code 3.1” of the original version of the SAs agreed with Iran in 1976 required Iran to provide design information on new facilities “normally no later than 180 days before the facility is scheduled to receive nuclear material for the first time.” In time, however, this came to be seen as inadequate, and in 1992 the IAEA began asking countries to agree to a set of “modified” Code 3.1 provisions that gave the Agency more time in which to develop safeguards plans for new nuclear facilities.  Under the new standard, reporting of design information was to be made as soon as the decision had been made to construct a new facility. (Design information also had to be adjusted as plans evolved.)

Iran did not agree to “modified Code 3.1” at the time the IAEA first began requesting these provisions.  This did not emerge as a significant issue, however, because during these years Tehran was still concealing the extent of its nuclear program from the IAEA. (With no new construction supposedly going on, the issue of the timely provision of design information was largely irrelevant.)  When the scope of Iran’s deception and the ambitious nature of its growing nuclear program

began to become clear in 2002‐03, however, the Code 3.1 problem became acute. Embarrassed by the revelations about its effort and scrambling to avoid the possibility of Security Council sanctions over secret nuclear work at the very time that its neighbor Iraq was about to be invaded for supposedly possessing illicit weapons of mass destruction, Iran agreed with the IAEA – as provided for in Article 39 of Iran’s CSA – upon a Code 3.1 modification in February 2003.

Apparently later deciding that they did not wish to provide such transparency after all, however – and perhaps hoping to create a legal loophole that would allow them to construct turn‐key facilities for uranium enrichment or plutonium reprocessing that would be ready for “breakout” from the Nuclear Nonproliferation Treaty (NPT), but that they could claim did not have to be declared to the IAEA as long as no plans existed to use nuclear material there during the next 180 days – the Iranians declared in March 2007 that they had “suspended” the modification to Code 3.1.  Thereafter, Iran claimed that only the 1976 version of the SAs was operative.

This has remained a bone of contention ever since. Iran says it had the right to “suspend” the modified SA provisions because they had not been ratified by the Iranian parliament.  The IAEA, however, denies this, arguing that Iran remains bound by the modified Code 3.1 because SAs – which must be adopted “by agreement,” and for which no “suspension” procedure is provided – cannot be altered unilaterally.  (At present, Iran is only sate with CSA in force that hasn’t implemented Revised Code 3.1.)

This dispute lies at the heart of disagreements over whether Iran committed an additional safeguards violation by concealing the construction of its Fordow enrichment facility, near Qom, until 2009. Iran claims that it began to construct Fordow in 2007 – that is, after the “suspension” of modified Code 3.1 – though this seems to be false, with the result that Iran was in noncompliance with its SAs even if Tehran were within its rights to have “suspended” the modified Code 3.1.  (Satellite imagery shows construction at Fordow in 2002‐04, a temporary halt, and then a resumption of activity in 2006. The IAEA says has seen additional information suggesting that design work began there in 2006.)  Dispute continues over whether the modified SA still applies.

Some of Iran’s legal arguments in this regard have proven unsustainable even by the standards of that country’s customarily tendentious efforts at pseudo‐legal self‐exoneration.  It has sometimes invoked its “suspension” of modified Code 3.1, for instance, as a reason to deny IAEA design information visits (DIVs) to facilities Iran had previously declared to the Agency. (This was a problem with the plutonium‐production reactor at Arak in 2007‐08, for instance.)  This was entirely untenable, however, since modified Code 3.1 had nothing to do with DIVs, but instead concerned the provision of design information in the first place: once a facility has been declared and design information submitted, the IAEA has a continuing right to verify that information (e.g., track the progress of construction) through such visits.  (Iran has also invoked the Code 3.1 issue as a reason not properly to update previously‐submitted design information.) Additionally, Iran falsely claimed that Article 34(c) of its CSA did not require reporting of nuclear imports such as its 1991 uranium shipment.

In the main, however, the dispute over Code 3.1 revolves around whether Subsidiary Arrangements are binding at all.  Iran’s position that SAs may be modified or “suspended” unilaterally because they have not been subjected to independent ratification is a claim that none of the provisions in any safeguards arrangements of this sort need to be followed by any host government, and that they may be modified (or scrapped) at will.  Were this true, the entire IAEA safeguards edifice would fall apart, for most of the actual operational details of nuclear safeguards arrangements are provided by SAs rather than in CSAs themselves.

It is through SAs, after all, that the detailed “technical and administrative procedures” are established for “how the provisions laid down in a safeguards agreement are to be applied.”  SAs provide both a general set of default procedures for nuclear safeguards in a given country (in their “General Part”) and detailed provisions for each specific facility that will be subject to safeguards in any way (in “Facility Attachments”).  If compliance with these provisions were deemed “optional,” governments would be permitted to change what specific safeguards procedures (if any) are applied at nuclear facilities at their whim, making nuclear accountability – the raison d’être of the IAEA – an impossibility.  Fortunately, however, this Iranian position is legally incorrect.

To the extent that it can be coherently expressed at all, the case against IAEA Subsidiary Arrangements being legally binding seems to revolve around the idea that (a) they are not conventional treaties subject to ratification and that (b) the parties involved did not intend SAs to be legally binding because they provided for an entry‐into‐force procedure (“by agreement”) that differs from that governing a CSA.  (This is the argument of Daniel Joyner, an American law professor who has worked hard to provide legal substance to the various vague and conclusory legal assertions made by Iranian officials in this regard.)  Fortunately for the very idea of nuclear safeguards, this argument is unsustainable.

There is no evidence that any of the parties that have negotiated SAs over the years have ever regarded them as anything but legally‐binding, and indeed – until Iran’s 2007 “suspension” of modified Code 3.1, at least – no one had ever asserted anything to the contrary or behaved accordingly.  (Indeed, Iran has itself repeatedly modified its SAs by simple agreement with the IAEA in order to accommodate the application of safeguards to its growing nuclear activities at Natanz and elsewhere, never suggesting that compliance with these procedures is optional, or that they require ratification.  Iran also seems to accept that it is bound by the 1976 version of Code 3.1, even though that provision was adopted in no different a way than the 2003 modification.)  More importantly, the text and structure of Iran’s CSA – and indeed all CSAs – make quite clear that SAs are intended to be legally binding, for they are integral to the structure and function of the safeguards mechanism that CSAs set in place, and this framework would make no sense otherwise.

Taking Iran’s CSA as an example, there are numerous points at which the agreement provides that the parties must follow the detailed procedures established by agreement between them in Subsidiary Arrangements. It is expressly provided in the CSA, for instance, that SAs agreed between the parties will define the scope of nuclear safeguard requirements with regard to: the characteristics of the material accountancy and control system for nuclear material (Article 32); the required time limits for the submission of design information to the IAEA (Article 42); what records must be kept in each material balance area (Article 51); whether and how small batches of material must be reported (Article 65); what must be done in the event of the loss of nuclear material or changes in containment or surveillance techniques (Article 68); the IAEA’s right to install surveillance gear and seals (Article 75); and the access points for IAEA inspectors (Article 76); and the intervals at which inspections are to take place (Article 90).

Without SAs, there would also be no provisions for how to handle “components of fuel elements containing nuclear material sealed in metal cladding (e.g. subassemblies and fuel rods, pins or plates) ... for batch definition and reporting purposes,” what items and batches will be handled at a nuclear facility, how uranium is to be accounted, what activities the IAEA is authorized to perform at an access point, how to apply seals and tamper‐resistant devices, or what reporting formats are to be used in submitting information to the Agency.

These are most certainly not all things intended to be left to the whim of the host government, any more than they are to be left solely to the discretion of the IAEA.  (If the former could rewrite or “suspend” them at will, nuclear safeguards would be meaningless; if the latter could do so, having this rule as the default basis for safeguards around the world would be an intolerable imposition upon national sovereignty.  It is very hard to imagine either alternative being the agreed intention of the parties to a CSA, nor is there the slightest evidence to this effect.)  To the contrary, it is integral to the structure and logic of the CSA framework – and nuclear safeguards in general – that such details of safeguards implementation be both agreed between the two parties and legally binding upon both.

In effect, therefore, Subsidiary Arrangements get their legal legitimacy and binding force from the CSA’s own ratification, for that document expressly requires the creation of SAs, provides a mechanism for their establishment, and obliges parties to follow their provisions. Whatever might (or might not) be the legal status of something like SAs if they were simply “agreed” on their own absent the overarching framework of a CSA, there is nothing inherently problematic about having one binding treaty instrument establish a mechanism for the creation and modification of further detailed arrangements that will also be binding under the framework created by that first instrument.  This is just what the CSA process does – indeed, rather explicitly – and indeed the CSAs structure and text would otherwise be both inexplicable and incoherent.

By ratifying the CSA, therefore, Iran approved its process for the creation of Subsidiary Arrangements and agreed to be bound by the results thereof.  (The CSA at one point even authorizes SAs to supersede the provisions of the CSA itself, requiring that Iranian declarations to the IAEA be made in at least one of several specified languages unless otherwise specified by agreement in Subsidiary Arrangements.  This makes no sense unless the SAs are themselves legally‐binding instruments.)  The specific mechanisms for each type of text are different, but they are intimately related, and ultimately rest upon the same legal foundation.

The IAEA, therefore, has been legally correct in continuing to apply modified Code 3.1 to Iran, and Iran’s continuing refusal to provide information according to its requirements constitutes yet another breach of Iran’s safeguards obligations.  These violations include Iran’s refusal for nearly two years to provide preliminary design information on the reactor it admitted having decided to construct at Darkhovin, as well as its refusal to provide revised design information on its Natanz and Fordow enrichment facilities in a timely fashion, before the installation of new centrifuge cascades, and its failure to report construction of a tunnel complex at Esfahan before construction was already underway.  Furthermore, though Iranian officials have publicly announced ambitious plans for additional nuclear construction – such as ten new enrichment facilities and four or five new research reactors that will run on HEU – Iran has also refused to confirm or clarify to the IAEA even whether or not it has taken the decision to construct any additional nuclear facilities.  (Iran has additionally claimed to possess laser enrichment technology and “third generation” centrifuges, but has refused IAEA requests for further information about such work.)  The IAEA has also faulted Iran for failing to provide timely design information on the previously secret enrichment facility under construction at Fordow that was publicly revealed only in 2009.

(Iran also later failed to comply with Article 45 of its own CSA by failing to provide information to the IAEA on its work to increase uranium enrichment levels from 5 percent to 20 percent in time for safeguards to be adjusted accordingly. By the time new safeguards were agreed and the IAEA arrived on the scene, Iran had already started to feed LEU into a cascade at Natanz for higher‐level enrichment.  At least for a period in 2006, moreover, Iran also refused to provide multiple‐entry visas for IAEA inspectors even through required to do so by its Subsidiary Arrangements.)

Because of Iran’s refusal to honor its obligations under modified Code 3.1, the Security Council acted in 2010 under Chapter VII of the U.N. Charter to require (in Resolution 1929) that Iran will “comply fully and without qualification with its IAEA Safeguards Agreement, including through the applications of modified Code 3.1 of the Subsidiary Arrangement to its Safeguards Agreement.”  Legally speaking, this was in no way necessary in order to create an Iranian obligation to comply with modified Code 3.1, but it certainly emphasized the importance of doing so, and would have created such an obligation even if Iran were right in its specious legal arguments about having “suspended” those provisions of its Subsidiary Arrangements with the IAEA.

II.           Verifying Suspension(s), the Additional Protocol, and U.N. Resolutions

In addition to verifying compliance (or noncompliance, as the case may be) with Iran’s basic safeguards obligations, the IAEA has played a role in verifying the degree to which Iran has complied with its agreements or obligations to suspend various aspects of its nuclear program. This role in verifying suspension began with Iran’s agreement in November 2003 – as part of a deal Tehran made with officials from Britain, Germany, and France pursuant to which those three countries would act to block the referral to the U.N. Security Council sought by American diplomats – to suspend all enrichment and reprocessing activities.  (The specific scope of this expansion was expanded somewhat further in February 2004.)  The IAEA was asked to verify Iran’s claims in this regard.

In December 2003, Iran agreed to sign the IAEA’s Additional Protocol (AP), a safeguards standard established by the IAEA in the 1990s in reaction to the obvious inadequacies of the standard CSA model with regard to undeclared nuclear activities, and with which Iran had in October 2003 promised to act in compliance until the AP’s entry into force.  (Tehran said it would also provide “transparency” beyond the AP in order to assuage international concerns about its program.)  Iran submitted initial declarations pursuant to the requirements of the AP in May 2004, and provided supplemental information in mid‐2004, thus providing the IAEA with a starting point for additional verification visits.  Iran would thereafter repeatedly stress that its agreement to cooperate “in accordance with the Additional Protocol” until that protocol’s formal entry into force was purely “voluntary” decision intended to serve as a “confidence building measure,” but the IAEA did thereby acquire an additional role in this way, including greatly expanded facility access as described in Article 5.a.(i) of the AP.

Unfortunately for Iran, IAEA investigations carried out to this end verified that Iran had not in fact suspended all of its work in producing centrifuge components, and Tehran and the Agency also quarreled over whether these suspension promises included stopping UF6 production at Esfahan after all.  In June 2004, the supposed “suspension” dissolved entirely, and Iran announced that it would officially resume making and testing centrifuge components, as well as doing “hot tests” at the Esfahan uranium conversion facility for the production of UF6.  Full‐scale UCF activities were re‐started in August 2005, and all other supposedly “suspended” enrichment‐related research and development work resumed in January 2006.  In February 2006, Iran also announced that it was abandoning its previous commitment to comply with the Additional Protocol: it declared that from that point forward, it would only be bound by the obligations of its CSA.  That same month, Iran began feeding UF6 into its centrifuge cascades at Natanz.  (Thereafter, Iran even threatened to cease complying with the CSA itself, declaring in April 2006 that it was prepared to continue to permit IAEA inspections under the CSA provided that the IAEA did not issue a report to the U.N. Security Council.)

At this point, the Security Council itself became involved, for in February 2006 – as noted earlier – the IAEA Board of Governors had belatedly reported the Iranian situation to the Council pursuant to Article XII.C of the IAEA Statute.  Nothing prohibited Security Council involvement before an IAEA referral, of course, but the politics of the situation were such that it was unlikely that the Council would have addressed the matter in advance of a formal IAEA determination that matters could not be handled in Vienna alone.  By 2006, however, Iran’s ongoing refusal to cooperate fully with the IAEA, commencement of uranium enrichment, and repudiation even of the side agreements that had previously enticed some Board members to oppose a XII.C report had changed this political balance.

In July 2006, therefore, the Council passed Resolution 1696, calling upon Iran to suspend all enrichment‐related and reprocessing activities, including research and development, and asked that this be verified by the IAEA.  (The Council also called upon Iran to act in accordance with the AP, as well as to implement all transparency measures IAEA requests.)  Resolution 1696 was a resolution pursuant to Article 40 of Chapter VII of the U.N. Charter, and thus constituted a “provisional measure” rather than the actual imposition of a binding legal obligation.  This resolution, however, set the stage for the Council’s subsequent moves – under Article 41 of Chapter VII of the Charter – to toughen the legal regime governing Iran’s nuclear‐related behavior.

A.           Security Council Obligations

In December 2006 – in the fact of continued Iranian intransigence and further enrichment work – the Security Council duly acted under Article 41 of the U.N. Charter to impose a sweeping additional suite of legal obligations upon Iran, compliance with which was to be verified by the IAEA.  Resolution 1737 began by actually requiring Iran to take all of the steps that the IAEA’s Board had asked it to take in a February 2006 resolution.  Through this exercise of the Council’s Chapter VII authority, therefore, Iran became obliged to “re‐establish full and sustained suspension of all enrichment‐related and reprocessing activities, including research and development, to be verified by the Agency,” to “reconsider the construction of a research reactor moderated by heavy water,” to “ratify promptly and implement in full the Additional Protocol,” and pending ratification of the AP, to “act in accordance with the provisions of the Additional Protocol.”

On top of this, on the strength of the IAEA’s earlier finding that verification work in Iran required more investigative authority than even the AP provided, the Security Council thus also required Iran to

“implement transparency measures, as requested by the Director General [of the IAEA] ... which extend beyond the formal requirements of the Safeguards Agreement and Additional Protocol, and include such access to individuals, documentation relating to procurement, dual use equipment, certain military‐owned workshops and research and development as the Agency may request in support of its ongoing investigations.”

In addition to requiring Iranian compliance with specified requests by the IAEA Board, Resolution 1737 also mandated Iran’s suspension of all enrichment‐ related and reprocessing activities, as well as all heavy‐water‐related projects including the Arak/IR‐40 reactor. It was specified that all this, too, was to be verified by IAEA.  To this end, the Security Council declared – again under Article 41 – that Iran must provide such access and cooperation as the IAEA requested for the purpose of verifying the suspension and resolving all outstanding issues identified in IAEA reports.

This package of requirements in Resolution 1737 (2006) formed a core of significant new obligations imposed upon Iran in order to resolve concerns about the nature and extent of its nuclear ambitions, and the IAEA was thus greatly empowered by having Iran be made legally obliged to cooperate as needed for this purpose. These Resolution 1737 obligations have been repeatedly reaffirmed by subsequent Article 41/Chapter VII Security Council Resolutions such as Resolutions 1747 of March 2007, 1803 of March 2008, and 1929 of June 2010.  (The list of Board resolutions with the demands of which Iran is obliged to comply has also been lengthened.)  The IAEA Director‐General has also been instructed to report back to the Council on Iranian compliance with these various requirements, thus providing a direct feedback mechanism to the United Nations without any further Article XII.C reporting under the IAEA Statute.

Unfortunately, however, none of this has had any discernible effect upon Iranian behavior.  Though obliged to do so by virtue of its membership in the United Nations, Iran has refused to accept the jurisdiction and Chapter VII authority of the Security Council.  In fact, it has decried “interference of the United Nations Security Council” in its affairs,61 and has insisted on providing information and cooperation to the IAEA only pursuant to its own (faulty) interpretation of the CSA.  It has continued to pursue (and to accelerate) uranium enrichment, and rejects the United Nations’ imposition of additional obligations and requirements to cooperate with the IAEA.  On the occasions that Iran has provided some information or permitted access beyond what the CSA would require, it has done so only on an ad hoc basis and inconsistently, and amidst declarations on its part that such actions are purely “voluntary.”

B.           The Issue of Military Dimensions

One of the more dramatic issues that has arisen in the course of IAEA investigations into the Iranian nuclear program is the existence of evidence suggesting that Iran has been engaged in “weaponization” work – that is, in preparing for the conversion of enriched uranium or separated plutonium into an actual nuclear warhead.  Except insofar as such work may have involved nuclear material subject to safeguards, this is not a matter that would ordinarily have been subject to the IAEA’s jurisdiction pursuant to Iran’s CSA, but the Agency began asking questions about such matters at a time when Iran still promised full voluntary cooperation in the interest of “transparency” and “confidence‐building,” and – not surprisingly, in light of their clear and direct implications for international peace and security – the issues raised have become ones that the U.N. Security Council has explicitly required Iran to resolve.

Among the earliest questions to come up in this respect concerned the IAEA’s discovery in late 2003 that Iran had irradiated samples of bismuth metal. This was not a reportable event under the CSA, but bismuth irradiation produces an isotope of Polonium, Po‐210, that can be used in the “neutron trigger” that helps form the core of an implosion‐type nuclear weapon.  Thereafter, in 2005, IAEA investigators came across a 15‐page document – acquired by Iran from Pakistan via the A.Q. Khan proliferation network – providing directions for how to machine uranium metal into hemispheres to form the core (or “pit”) of an implosion‐type weapon.

Since then, the IAEA has collected a sizeable dossier of information – mostly provided by member states about Iran rather than by the Iranians themselves, who continue to deny having done any weaponization work – relating to such topics. This information, the Agency has said, comes from more than 10 separate governments, and has been the subject of intensive IAEA verification analysis. The Agency finds this information credible.  (When shown this information by the IAEA, moreover – which in most instances the collecting governments eventually permitted the Agency to do – the Iranians have given incomplete and sometimes contradictory explanations.)  The IAEA’s “military dimensions” dossier on apparent Iranian weaponization work is highly provocative, and the questions it raises remain unresolved.

Among the items in this dossier are documents relating to the process of testing specialized high‐explosives and high‐speed synchronized detonators suitable for the core of an implosion‐type nuclear weapon, as well as designs for digging an apparent nuclear testing pit 400 meters into the ground (complete with diagnostic instrumentation wiring).  The IAEA also possesses what appears to be Iranian engineering documentation (in electronic form) on preparing a ballistic missile re‐entry vehicle for a spherical payload to be detonated at an altitude of 600 meters.  Especially when coupled with evidence of a visit to Iran by a former Soviet weapons scientist who specialized in the preparation of spherical implosion explosive arrays – and evidence of the existence at the Parchin military base of a large high explosive containment vessel for such hydrodynamic experiments – these elements add up to a disturbing picture of weaponization research and development work.

(Interestingly, some of the IAEA’s documentation related to apparent Iranian weaponization work dates from 2004 and 2005, and the Agency even claims to have information suggesting that Iranian experts were doing mathematical calculations related to the implosion core of a nuclear device as recently as 2008‐09.  This suggests that the U.S. Intelligence Community got things rather wrong in its 2007 National Intelligence Estimate [NIE] on Iran, in claiming that Iran had suspended weaponization work in the autumn of 2003. The NIE, in fact, was even wrong to claim that Iran stopped its “covert uranium conversion‐related and uranium enrichment‐related work” in 2003.  As noted earlier, construction of the Fordow enrichment plant may have begun as early as 2002, and even Iran admits that it began building that facility in 2007, before the issuance of the U.S. NIE.  The public performance of U.S. intelligence analysts and their political masters vis‐à‐vis Iran has not been impressive.)

As noted, acting under Article 41 of Chapter VII of the U.N. Charter, the Security Council has required Iran to resolve questions about these matters and to cooperate as needed with the IAEA to this end.  This, however, has yet to occur even as Iran accumulates an ever‐larger reserve of HEU capable of being easily and quickly enriched up to bomb‐grade levels.

C.           Undeclared Activities

The most recent critique to emerge of the IAEA’s work in Iran – apart, that is, from Iran’s repeated misreadings of its own CSA and the associated Subsidiary Arrangements, its fallacious reinvention of Article IV of the NPT, and its effort to will Chapter VII of the U.N. Charter out of existence – concerns the IAEA’s objective of verifying the absence of undeclared activities and verifying that nuclear material in Iran is being used for exclusively peaceful purposes.  Legally speaking, this criticism is no stronger than Iran’s other complaints, but, like the Code 3.1 issue, it has been given special attention in the West by virtue of having been championed by Daniel Joyner of the University of Alabama. For this reason, rather than on its merits, it deserves mention here.

This critique takes Iran’s side in arguing that the IAEA still possesses only the authority vis‐à‐vis Iran that was originally given it by Iran’s CSA, and that therefore essentially all of its efforts to investigate Iran’s nuclear program have been ultra vires and illegitimate.  The argument relies upon a close reading of some obscure documents (e.g., the minutes of one particular IAEA Board meeting) but a strangely shallow perusal of other obvious and important ones (e.g., Iran’s CSA itself, and multiple Security Council resolutions).

As described above, the Security Council has clearly and repeatedly both imposed new legal obligations upon Iran under its Chapter VII authority and directed Iran to provide whatever cooperation the IAEA needs as it works, at the Council’s request, to verify compliance with these obligations. The Security Council has also required Iran to comply with the Additional Protocol, pursuant to which Iran must permit the IAEA a broad range of investigative authorities beyond what the CSA provides, and to comply with the requests of the IAEA Board – which have explicitly demanded more transparency than would be the case under either the CSA or the AP.

Nor, of course, is there anything in the IAEA’s Statute that would prohibit it from stepping up and doing such work at the request of the Security Council. Pursuant to Article 48 of the U.N. Charter, decisions of the Council must be carried out by members of the United Nations, not merely on their own but by acting through appropriate international agencies such as the IAEA.  Furthermore, the standing agreement between the U.N. and the IAEA obliges the Agency to consider any resolution relating to the IAEA that should happen to be adopted by a United Nations body.

It is a curious reading of these documents indeed that denies that the Council has made it possible for the IAEA to investigate beyond the confines of its CSA authorities. In a sense, Iran’s position is actually more intellectually coherent than Joyner’s – though no less wrong – inasmuch as officials in Tehran regard the Security Council itself as acting ultra vires, or simply as lacking authority to impose obligations upon countries in this fashion in the first place, and thus dismiss everything it has said vis‐à‐vis Iran. (It is less clear how Joyner can apparently accept the authority of the Council and yet ignore what it has done pursuant to that authority.)

Per Iran’s CSA, moreover, the IAEA has both “the right and the obligation” to ensure that safeguards are applied to “all” relevant nuclear material in peaceful activities in Iran in accordance with the terms of that agreement.  This necessarily implies the ability to make inquiries about the possibility that some material or nuclear activities have not been declared.  (The CSA certainly does not say, for instance, that the IAEA shall apply safeguards “only to such nuclear material as the Government of Iran has declared to it.”)  The Agency clearly has authority to ask what it feels it needs to ask in order to acquire sufficient confidence that there are no undeclared activities, and to ascertain the correctness and completeness of a government’s declarations.

To be sure, the Agency has not always had adequate tools for fulfilling these responsibilities – particularly with regard to requiring the host government to provide information and access beyond what is specifically covered in a CSA and the associated Subsidiary Arrangements. (The Agency, in other words, was assigned responsibility for doing more than it was given the ability to accomplish: it could ask whatever it thinks appropriate, but Iran was not always obliged to let IAEA inspectors do what they wish in order to check the veracity of Iran’s answers.)  The Additional Protocol was developed to help fill this gap: its role is to provide investigative authorities more suited to the IAEA’s ongoing role in verifying the absence of undeclared activities and the correctness and completely of declarations.  (The AP refers to the IAEA’s responsibility in this regard,80 but it does not create this responsibility, which predated the Protocol. The AP is grounded in the recognition that this responsibility cannot be properly fulfilled without more investigative tools.)  As we have seen, in cases of systematic and sustained deception such as that of Iran, even the AP may not be enough.  The Protocol is a vital addition to the safeguards system nonetheless.

(The Agency’s standard of proof for reporting safeguards issues to the Security Council, it should be remembered, is actually fairly low. Pursuant to Iran’s CSA, for instance, the IAEA Board may report Iran to the Security Council if it

“finds that the Agency is not able to verify that there has been no diversion of nuclear material required to be safeguarded under this Agreement, to nuclear weapons or other nuclear explosive devices.”

The IAEA does not need to show that there has been a diversion – e.g., to nuclear weapons – but rather needs merely to have concluded that it is insufficiently sure that one hasn’t occurred.  Without the expanded tools of the AP, therefore, the recognized inadequacy of the CSA alone might under many circumstances all but compel the IAEA to refer a country to the U.N. Security Council – not necessarily because there was any actual evidence of diversion, but simply because without the ability to investigate further, the Agency would be unable to verify non‐diversion. The Additional Protocol, therefore, actually serves the interests of host governments – as well as the IAEA itself – by permitting the Agency to satisfy most of its investigative responsibilities with regard to potentially undeclared materials and activities without having to involve higher authority.)

In any event, it is specious to contend that the IAEA is somehow acting ultra vires in seeking to ascertain the absence of undeclared activities in Iran, to verify the correctness and completeness of Iranian nuclear declarations, and to verify Iranian compliance with the nonproliferation requirements imposed by the U.N. Security Council on top of Iran’s ongoing obligations under its CSA and Subsidiary Arrangements.

III.           NPT Compliance

It is not the IAEA’s job either to assess or to enforce compliance with the NPT itself, but some countries clearly have – either expressly or implicitly – reached the conclusion that Iran is in violation of that treaty. U.S. officials had publicly assessed as early as 1993 that Iran had

“a continuing interest in nuclear weapons and related technology that caused the United States to assess that Iran was in the early stages of developing a nuclear weapons program, with an emphasis on developing centrifuge technology.”

It took a discreditably long time for the United States to reach the logical conclusion therefrom that Iran was in violation of the NPT, but U.S. diplomats declared at the April 2004 NPT Preparatory Committee meeting that

“Iran’s oil rich environment, grudging cooperation with the IAEA, its deception, and its 18‐year record of clandestine activity leads us to the inevitable conclusion that Iran is lying and that its goal is to develop a nuclear weapon in violation of its Article II commitments.”

Just what precisely the standard is, or should be, for reaching such a noncompliance conclusion, however, is worth further attention. Article II of the NPT obliges Iran – and indeed all non‐nuclear weapons states (NNWS) party to the treaty – not to

“receive the transfer from any transferor whatsoever of nuclear weapons or other nuclear explosive devices or of control over such weapons or explosive devices directly, or indirectly; not to manufacture or otherwise acquire nuclear weapons or other nuclear explosive devices; and not to seek or receive any assistance in the manufacture of nuclear weapons or other nuclear explosive devices.”

Article III, furthermore, requires NNWS

“to accept safeguards, as set forth in an agreement to be negotiated and concluded with the International Atomic Energy Agency in accordance with the Statute of the International Atomic Energy Agency and the Agency’s safeguards system”

in order to verify

“the fulfillment of its obligations assumed under this Treaty with a view to preventing diversion of nuclear energy from peaceful uses to nuclear weapons or other nuclear explosive devices.”

Strangely, apart from the United States, no country has explicitly set forth the standards they apply when interpreting Articles II and III of the NPT.  The Americans, however, carefully spelled this out for the first time in 2005.  (They also offered a discussion of their understanding of compliance analysis vis‐à‐vis IAEA safeguards, though neither the IAEA Secretariat nor its Board is of course bound by this U.S. interpretation.) As first and only existing explication of these matters ever offered by any government, the U.S. standards are worth recounting here. As explained on behalf of the United States in a report authored by the U.S. State Department bureau responsible for such matters,

“In assessing a NNWS Party’s compliance with its Article II obligations – not to manufacture, or to seek or receive any assistance in the manufacture of, nuclear weapons – no simple, clear, ‘bright‐line’ rule exists. In explaining the term ‘manufacture’ to the U.S. Senate in connection with the NPT ratification process, Arms Control and Disarmament Agency Director William Foster stated that it was not possible to ‘formulate a comprehensive definition or interpretation,’ and he doubted the efficacy of such efforts ‘unrelated to specific fact situations.’ Accordingly, compliance assessments are highly contextual, and no single, comprehensive definition, unrelated to specific factual situations, would be useful. However, the United States has explicitly stated that the prohibition against the ‘manufacture’ of a nuclear weapon, as well as against seeking or receiving any assistance in this regard, reaches more than simply the final assembly of such a device. In addition, Director Foster advised the Senate that ‘facts indicating that the purpose of a particular activity was the acquisition of a nuclear explosive device would tend to show noncompliance.’  Thus ... an important factor in Article II compliance analysis is the purpose of a particular activity.

“U.S. officials have publicly outlined some of the ‘warning signs’ that may indicate a prohibited nuclear weapons purpose, and thus suggest that a country’s ostensibly ‘peaceful’ nuclear program might have violated Article II and should be closely scrutinized. Such indicia can include: (a) the presence of undeclared nuclear facilities; (b) procurement patterns inconsistent with a civil nuclear program (e.g., clandestine procurement networks, possibly including the use of front companies, false end‐use information, and fraudulent documentation); (c) security measures beyond what would be appropriate for peaceful, civil nuclear installations; (d) a pattern of Article III safeguards violations suggestive not of mere mistake or incompetence, but of willful violation and/or systematic deception and denial efforts aimed at concealing nuclear activities from the International Atomic Energy Agency (IAEA); (e) a nuclear program with little (or no) coherence for peaceful purposes, but great coherence for weapons purposes (e.g., heavy water production in a country the civil nuclear facilities of which use only light water as a moderator, or pursuit of enrichment facilities when other, cheaper energy‐producing resources or an outside source of enriched uranium are available, or the pursuit of a full fuel cycle for a civil reactor program too small to provide economic justification for such an effort).  As cited by Director Foster in his testimony to the Senate as relevant to a finding of ‘manufacture,’ activities related to the acquisition or testing of the non‐nuclear components of the nuclear explosion are an example of the type of activities that would provide a more direct indicator of a weapons program. Informed by the analysis of such factors, judgments as to the purpose of a Party’s nuclear activities therefore lie at the core of Article II compliance assessments.

“In sum, Article II assessments must look at the totality of the facts, including judgments as to the NNWS Party’s purpose in undertaking the nuclear activities in question, to determine whether the Party has engaged in efforts to manufacture or otherwise acquire a nuclear weapon or other nuclear explosive device, or has sought or received any assistance in such manufacture.”

It is not clear whether, or the degree to which, other countries that believe Iran to have a nuclear weapons program have followed this analytical model.  Few, if any, have made formal Article II compliance assessments, and none seem to have explained the analytical standards they apply.  To the extent that the international community does take a stand against Iran over its nuclear program, however, these NPT Article II questions are highly significant, and governments may be asked or expected to explain their reasoning. So far, only the Americans have done so.

-- Christopher Ford

About Dr. Ford

Dr. Christopher Ford presently serves as Republican Chief Counsel to the U.S. Senate Committee on Appropriations. From 2008-13, Dr. Ford was a Senior Fellow at the Hudson Institute in Washington, D.C. Before that, he served as U.S. Special Representative for Nuclear Nonproliferation, Principal Deputy Assistant Secretary of State, Minority Counsel and then General Counsel to the U.S. Senate Select Committee on Intelligence, and Staff Director of the Senate's Permanent Subcommittee on Investigations. A graduate of Harvard College (summa cum laude), Oxford University (as a Rhodes Scholar), and Yale Law School, Dr. Ford was also ordained by Roshi Joan Halifax of the Upaya Zen Center as a lay chaplain in a lineage Soto Zen Buddhism. He is also a student of Grandmaster Dong Jin Kim in Japanese jujutsu (black belt 2nd Dan) and Hapkido (also 2nd Dan), and served from 1994 until 2011 as an intelligence officer in the U.S. Navy Reserve. The author of the books "The Mind of Empire: China's History and Modern Foreign Relations" (2010) and "The Admirals' Advantage: U.S. Navy Operational Intelligence in World War II and the Cold War" (2005), as well as co-editor of "Rethinking the Law of Armed Conflict in an Age of Terrorism" (2012), Dr. Ford has written dozens of articles and essays in international and national security affairs. For a list of his publications, see The views he expresses here are entirely his own, and do not necessarily reflect the perspectives of anyone else in the U.S. Government.
Comments (0) Trackbacks (0)

Sorry, the comment form is closed at this time.

Trackbacks are disabled.

NPF Pages

Recent Additions to NPF

NPF Discussion Pages

Calendar of NPF Postings

April 2014
« Mar    

NPF Archives (by month)


To submit a comment to NPF, please send your feedback via e-mail to (In order to combat on-line "spam," this site now only accepts comments submitted via e-mail.) Thanks for your understanding!