Mark Hibbs on the Leaked IAEA Safeguards Report

So the nuclear nonproliferation blogosphere is lit up with the news of the leak of the 2012 IAEA Safeguards Implementation Report to Bloomberg and to Reuters last week. This also set the Twitterverse on fire, as Mark Hibbs reports in his piece over at Arms Control Wonk, which tries to dampen the hysteria and provide a pro-IAEA spin to the report’s revelations about how much of the IAEA’s budget and time have been spent on the Iran case.

I think in Mark’s apparent haste to explain some of the comments he made that were quoted in the report, he provides some rather superficial and ultimately erroneous analysis of the new report, blaming others in their analysis for “conflat[ing] the difference between compliance and performance evaluation.”

What I think Mark doesn’t see is the much more fundamental and underlying problem, which is that the concept of “compliance” with IAEA safeguards has been mangled and misused by the IAEA in its evaluations and reports for years.

The INFCIRC/153 CSA is very clear about how the system is supposed to work, and what standards the IAEA is supposed to use in their investigations and assessments. As readers of this blog will know, I have discussed the issue of the IAEA’s mandate extensively both here and in a Roundtable over at the Bulletin of the Atomic Scientists.

In brief, it is the IAEA’s now unfortunately institutionalized practice of exceeding its investigation and assessment mandate, that produces all of the problems of inconsistency, politicization, and error in the application of safeguards by the IAEA, including what Hibbs describes as “conflat[ing] the difference between compliance and performance evaluation.”

In the case of Iran in particular, the IAEA has gone beyond its mandate in deciding that it will not determine Iran is in “compliance” with its safeguards obligations, unless the IAEA can assess not only that no declared fissile materials have been diverted from peaceful to civilian use (which is the sole, correct CSA standard), but also that there are no undeclared fissile materials or related facilities in Iran and never have been, and that Iran has never performed any experiments of any kind the knowledge from which could conceivably be employed someday in the development of a nuclear weapon.

This standard is, of course, wildly incorrect, as I have taken pains to demonstrate.

Included in the mix of condemnable acts by Iran that has clearly factored into the IAEA’s unwillingness to declare Iran in compliance with its safeguards agreement according to this erroneous standard, are Iran’s failure to provide design information on the Qom facility, and the Arak facility, in what the IAEA considers a timely manner.

But wait, isn’t this a perfect example of what Hibbs means by conflating the difference between compliance and performance evaluation? As he explains:

In general it can be said with authority (and fully backed up by my historical files) that quite a number of states – and,  importantly, mostly states with significant nuclear programs – have at critical times resisted efforts of the IAEA to require that they provide more information and access . . . . Does this record show that states as a matter of course have resisted taking on additional obligations to the IAEA? Yes. But does this resistance imply that these same states were cheating on their current safeguards obligations? No.

But isn’t that exactly what the IAEA itself is doing, when it takes Iran’s alleged failure to meet a subsidiary arrangement standard on design disclosure into account in making its determination of Iran’s “compliance” with its safeguards obligations?

Instead of looking to the IAEA’s critics as the conflators, Hibbs should look to his good friends at the IAEA itself (the ones who leaked the new report to him), which has been the primary author of the mess that has been made of the concept of “compliance” with IAEA safeguards.

This is why when the people who sent the Tweets Mark refers to, read the excerpts from the new safeguards report and see, for example, its observation that a number of other states failed to make reports of design information of nuclear facilities in accordance with the modified Code 3.1 standard, they think: “Wait a minute, that’s exactly what the IAEA has been criticizing Iran for, and why they say Iran is still noncompliant.  Why aren’t there noncompliance reports on these other states, then?”

Again, it’s the IAEA that has conflated all sorts of erroneous things into their erroneous definition of “compliance,” and that’s what is producing this inconsistency.

And here is the problem of principle involved.  To the extent that the IAEA has gone beyond its clearly delineated and limited textual mandate for investigations and assessment, found in the CSA, and into broader and more subjective standards for investigation and assessment (e.g. like the standard of “no undeclared fissile materials,” and the concept of “cooperation with the Agency”) the Agency has become increasingly susceptible to being co-opted and used for political purposes by powerful states.

This is why, when those same people read in the new safeguards report about how much of the IAEA’s safeguards budget is being spent on Iran, and what proportion of man-hours are being spent on Iran, they naturally think: “Ya know, the IAEA has found no evidence that Iran is doing anything illegal, and yet it keeps chasing ghosts and endless speculations, provided by Western national intelligence agencies, and is never satisfied, even after all this time and all this money being spent.  And now it seems that there are a lot of other countries who are doing the same things the IAEA say are so condemnable about Iran. Something doesn’t add up here.”

And so they start to pay closer attention to reports about how much of the IAEA’s budget is provided by the US and its allies; how closely aligned with the US DG Amano was revealed to be in the WikiLeaks cables; and how much technical assistance the IAEA gets from the US in running its safeguards program. And they think: “Maybe this doesn’t add up because it’s not really about holding Iran to the same legal standards as other countries at all. Maybe this is really just one more way for the US to pursue its political agenda against Iran – whom it considers to be a grave threat to its chief Middle East ally Israel, and to itself – by using its influence over the IAEA to get the Agency to act as its proxy in criticizing Iran’s nuclear program. And maybe the IAEA is never going to be satisfied that Iran is in fact in compliance with its safeguards obligations, because the US doesn’t want it to ever be satisfied, and thereby recognize Iran’s legal justification for its peaceful nuclear program.”

And then they think: “Yeah. Now it does all add up. That really sucks. People should know about this.”


16 Comments on “Mark Hibbs on the Leaked IAEA Safeguards Report”

  1. yousaf says:

    The problem is that the IAEA is run in an ad hoc manner.

    I think you had captured the point here:


    Thus, in conducting these investigations into technologies other than fissile materials and producing this report, the IAEA is acting wholly outside of its authority pursuant to its safeguards agreement with Iran.

    This is why the report’s findings are so indeterminate. Since the IAEA is acting outside of its legal authority in this section of the report, it does not have a legal standard to apply to its conclusions regarding possible nuclear weapons related activities not involving fissile material. Throughout the report, the Director General expresses “concern” about the information being presented, and requests “clarification” from Iran in order to address these concerns. However, since there is no treaty language in Iran’s, or any other state’s, safeguards agreement that deals with non-fissile-material activities related to nuclear weapons, there is no prohibitive or regulatory standard that the Director General can point to against which to make a conclusion of compliance or non-compliance. In short, as the ancient legal maxim states, there can be no illegality where there is no law. The IAEA is simply “concerned.”

    ==================END QUOTE===========

    “Concerns” don’t amount to legal issues, at least not always.

    As for impartiality, it turns out the Agency cannot conclude that all nuclear material in 50 other nations (besides Iran) is in purely peaceful activities either:

    “(b) For 51 of these States, the Secretariat found no indication of the diversion of declared nuclear material from peaceful nuclear activities. *Evaluations regarding the absence of undeclared nuclear material and activities for each of these States remained ongoing.* On this basis, the Secretariat concluded that, for these States, declared nuclear material remained in peaceful activities.”

    So whatever the mainstream media like to say about why they (wrongly) think Iran is considered in non-compliance because the IAEA cannot prove that Iran’s nuclear program is purely peaceful — the IAEA cannot prove that about 50 other states either.

    Actually, a more recent report makes clear that there are 54 nations for which the IAEA cannot prove their nuclear programs are purely peaceful.

    • Dan Joyner says:

      Yes. And you are right to say that this shows inconsistency in applying the standard that the IAEA has been erroneously applying to Iran – i.e. if Iran is in noncompliance because the IAEA can’t confirm there are no undeclared materials in Iran, then why aren’t all these other countries in noncompliance for the same reason?

      BTW, I just commented on Mark’s post over at ACW, linking to this piece and inviting his comments. We’ll see if he allows my comment to stand over there. Mark used to be good about allowing comments from people who don’t agree with him. We’ll see if he is now.

      • yousaf says:

        This is also relevant, By Peter Jenkins, former UK Ambassador to the IAEA and UN (Vienna):


        “Addressing the Board of Governors of the International Atomic Energy Agency (IAEA) on 3 June, the director general devoted more than half of his statement on Iran to a continuing absence of clarity in relation to certain unresolved issues.

        He complained that, despite ten rounds of talks since January 2012, the IAEA and Iran were still well short of agreement on a document (a so-called “structured approach”) that lists the issues on which the IAEA believes greater clarity is needed. “To be frank, for some time now we have been going round in circles,” said Director General Yukiya Amano.

        What he did not say, and what IAEA spokesmen have rarely, if ever, said over the last 18 months is that the “structured approach” document contains a mixture of apples and pears. The document describes issues that fall within the scope of Iran’s Comprehensive Safeguards Agreement (CSA) and issues that are without the scope of that agreement…..

      • Dan Joyner says:

        Yes, that is a great piece by Peter that I highly recommend.

      • Bibi Jon says:

        According Peter Jenkins:

        The IAEA Safeguards Glossary suggests that the objective of safeguards in states that have entered into CSA’s is two-fold:

        – to verify that all nuclear material in a state’s possession is in peaceful use

        – and to detect the existence of any undeclared nuclear material or activities


        Is Amb. Jenkins correct?

      • yousaf says:

        I am not sure but this may be a case where the IAEA may have a mandate, but not the legal authority to carry out the mandate fully.

        e.g. There are two cases about the intrusiveness of inspections legally authorized, with an AP and without:


        What verification measures are used?

        Safeguards are based on assessments of the correctness and completeness of a State’s declared nuclear material and nuclear-related activities. Verification measures include on-site inspections, visits, and ongoing monitoring and evaluation. Basically, *****two***** sets of measures are carried out in accordance with the type of safeguards agreements in force with a State.

        One set relates to verifying State reports of ******declared****** nuclear material and activities. These measures – authorized under NPT-type comprehensive safeguards agreements – largely are based on nuclear material accountancy, complemented by containment and surveillance techniques, such as tamper-proof seals and cameras that the IAEA installs at facilities.

        Another set adds measures to strengthen the IAEA’s inspection capabilities. They include those incorporated in what is known as an “Additional Protocol” – this is a legal document complementing comprehensive safeguards agreements. The measures enable the IAEA not only to verify the non-diversion of declared nuclear material but also to provide assurances as to the ******absence of undeclared********* nuclear material and activities in a State.


        So absent an AP, IAEA cannot do much regarding the possibility of undeclared material: they have a mandate but not the legal authority to look into it.

        For 54 countries this situation applies.

        It is (in my simple un-lawyerly view) similar to the Police: they have a mandate to stop crime, but not the legal authority to waltz into your bedroom at 3am.

        Countries delimit the IAEA’s authority on purpose to preserve a measure of sovereignty.

        Especially to guard against a politicized Agency, this would seem to be a wise move.

        As Pierre Goldschmidt correctly summed up, “The Department of Safeguards doesn’t have the legal authority it needs to fulfill its mandate and to provide the assurances the international community is expecting.”


  2. Cyrus says:

    I recommend Amb Peter Jenkins’ latest contribution to LobeLog on the point that the IAEA reports on Iran mixes up apples (Iran’s legal obligations under the NPT and its CSA) and pears (demands on Iran by the UNSC) at

    “This obfuscation is not accidental. Western governments are using Iran’s alleged non-compliance to create an impression that Iran remains a grave non-proliferation threat, despite US national intelligence estimates that Iran has never engaged in the manufacture of nuclear weapons and has not decided to do so, and also to justify the pressure tactics that have so far signally failed to bring about capitulation to the Western demand that Iran renounce dual-use nuclear fuel cycle technologies”

    • Cyrus says:

      Oops, nevermind. Yousaf beat me to it.

      • Dan Joyner says:

        Thats ok. Peter’s piece deserves to be mentioned twice.

      • yousaf says:

        That’s a great quote from his piece also!

        Jenkin’s piece got stepped on by the PRISM story — should be disseminated.

      • Johnboy says:

        I have to say that Jenkin’s suggestion that the IAEA compile TWO documents (one detailing the issues outstanding under Iran’s CSA, and the other detailing allegations that the UN Security Council tasks the IAEA to investigate) is just about the sanest commentary I have read in a long, long time.

        But do I think that the IAEA would give that suggestion even a moment’s thought?

        No, I don’t, not for a second.

        And the reason why is pretty obvious: the first document would be no longer than a paragraph, maybe two.

        The second document would contain everything else, and Amano has no comeback if the Iranians tell him to shove THAT document where the sun don’t shine.

        Best to continue conflating the two – this ensures that the standoff continues indefinitely, but it prevents Amano from being exposed as Not Wearing Any Clothes.

      • Dan Joyner says:

        On that reasoning, I think we should all be in favor of it 🙂
        Just kidding.

  3. Nick says:


    Sorry, but I disagree with your comment on Fordo and Arak. Modified SA Code 3.1 does not apply to either one. And as Soltanieh said last week in Vienna, Arak DIQ was provided to the Agency many years ago.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s