The NIE on Iran's Nuclear Intentions and Capabilities

The Office of the Director of National Intelligence has released its 2007 summary of Iran's Nuclear Intentions and Capabilities with regard to weapons. I have previously reviewed NIEs on terrorism here and here.

I will draw on the intro of that second to help orient the reader on what an NIE is and what it does:

First and foremost it is a bland cross-agency summary of where the INTEL Community thinks things were as of MAR 2006. And as it is a set of Judgments on that intelligence information, it is summary in nature and an overview of the material.

Secondly, this is a high-level summary for an Executive and so must be short and easy to read. The President has daily IC briefings with his Cabinet officers and can get the daily work and updates as necessary. Every so often, probably quarterly, the IC does a summation of their work and then tells what is going on. Thus there will be little to NO backing evidence in the actual summation, but an overview of what the actual information hold.

Third, while it tries to be predictive, it is done so cautiously and must gain adherence from the Agencies involved. Further, Federal Agencies have a large amount of CYA involved, and so try not to 'make waves' or put forth partisan views, but offer the best analysis on the facts.

Fourth and finally, INTEL in all of its forms (HUMINT, ELINT, SIGINT, IMINT, MASINT, etc.) is never, ever certain and has a probability of being correct attached to it or likelihood of certainty. Thus any analysis, even one that is apparently 'slam dunk' is only in the 85-90% certainty range and is seen as conclusive as nothing else fits the overviews when pieced together. The whole of the information shows up much more than the pieces by themselves do. This is 'connecting the dots' to form the larger picture, so that a bunch of unconnected dots have inter-relationships and form a coherent whole.

This does mean that things can get cast into a view that is NOT as it appears due to unexpected connectedness, lack of connectivity or causality that is not there, but apparently *is* there. The art of INTEL analysis is to cast worst and best case scenarios and see what the outgrowths of those are and how well each fits with new and incoming data. These models of how relationships form, work and reach outwards are then reviewed and adjusted to take new facts into account. Often entire structures need to be removed and rethought because facts fully contradictory to the existing models and with a very high degree of reliability come to light which invalidates previous worst and best case scenarios.

Before going on to the actual analysis, there also needs to be something stated about the DNI and the Intelligence Community (IC), which I have looked at here in regards to the problems of the IC, and here with regard to the turf wars, which I will draw on for this understanding of bureaucracy in government:

We do not see this in politics today as it is harsh, results-oriented, extremely basic and makes good sense. Plus it endangers pet projects by Congress, who are loathe to give up same. Until such time as that can be done, trying to reform the system from the *inside* is a non-starter and any added 'oversight' organization immediately looks to create its own, sinecured fiefdom, put in further blocks to actually producing work products, and increase the overall overhead of the entire affair in time, money and wasted man hours. Process reform from bureaucracies and those empowering them do not work, as the Federal Government has demonstrated for decades. If it did then the Dept. of Education could tout 100% reading for all Citizens and that overall reading comprehension rate is above that of the 5th grade. Unfortunately those exact same problems remain just where they were in 1958 when Johnny couldn't read. Putting in a Director of National Intelligence hasn't seemed to have made the flow of INTEL any faster and, in point of fact, is allowing Agencies to further propagate their *own* outlook on things contrary to that of the President. More oversight is leading to less accountability. The Dept. of Agriculture, originally set up to research better ways to actually manage crops and to have a goodly sized fund to support farmers in case of emergency, now has that portion of its budget as the smallest part of it and over 20 times that amount going to crop subsidies and paying farmers not to farm. Somehow I can imagine better uses for those funds and skills in Afghanistan than in paying for crop subsidies here.

Thus the DNI will increase regularity, increase oversight and decrease accountability across the IC. This NIE is a product of that regularization process and is most soothing to the reader, as it is regularized, bland, and with large amounts of CYA by the IC and its component organizations. In no way is it a 'fraud', but folks do need to understand what is going on to render such documents and to not dress up a pig to be Cinderella.

The opening paragraph is a statement of the DNI and what it does and that it is, indeed, following the mandates of Congress. The ODNI does no INTEL work on its own, and is a coordinating office amongst the IC components and our International partners in the INTEL field. That is the 'depth' it brings: coordination, not original product.

From there the ODNI clearly states that it is added above the National Intelligence Council, which had been the previous body to help coordinate cross-IC NIEs. The ODNI is, thus, a coordinater and manager of the NIC and getting the rest of the IC to play together to form cross-IC documents and products. ODNI is an added procedural office, then, with its own goals and outlooks separate from the IC components, although given authority over their budgets and line items. Notice that in no way does ODNI have any responsibility for removing governmental pork: it is only an oversight and management organization and extra procedural step for vetting cross-IC needs, not put in charge of removing time wasting Congressional projects or re-purposing them to better fit National Intelligence needs.

To regularize the NIE process and familiarize the reader, the second page goes over that. It is of note that the procedure is one that adds bias into the NIE at the start of the process, this taken from the second paragraph:

The NIEs are typically requested by senior civilian and military policymakers, Congressional leaders and at times are initiated by the National Intelligence Council (NIC). Before a NIE is drafted, the relevant NIO is responsible for producing a concept paper or terms of reference (TOR) and circulates it throughout the Intelligence Community for comment. The TOR defines the key estimative questions, determines drafting responsibilities, and sets the drafting and publication schedule. One or more IC analysts are usually assigned to produce the initial text.

Stepping an outsider through this on a high level, we see the following on page 3:

1) Request from proper source for INTEL product, in this case the NIE, which goes through the NIC for approval - There is minimal bias in this step, as it is an attempt to look at things cross-IC. The only bias is in the *originating* request, which can be damned biased for the originator, but the NIC examines the request and then tries to remove the initial bias as best it can.

2) The request is handed to a component of the IC (National Intelligence Organization) - Notice that all other acronyms in this paper are spelled out, but NIO is not. An NIO is something we call an 'Agency' or 'Department' with CIA, NSA, DIA, NGA and others being NIOs. At this point, although there is specialization to INTEL collection type generally assigned to these NIOs, each has its own bureaucracy and outlook. An NIE gains bias at two points here: first, in selecting which NIO(s) get to draft the structure of the NIE and the internal NIO bias of each component organization. The ODNI is set up to try and minimize this, but its function has inherent bias in its set-up: if the ODNI wants a confrontational document, one NIO may be preferred over another, if a more soothing document is wished, a less confrontative and more bureaucratic organization is chosen.

3) The NIO creates the Terms of Reference (TOR) - The TOR is an outline of what the stated area of INTEL is and how it will be looked at based on the TOR language. TOR language typically has a 'shelf life' depending on the 'product' and its 'importance', so that stolid forms of INTEL will rarely change basic TOR terminology and meaning, while those that are fluid will often have single-document specific TOR language. Which NIO and individuals within NIOs that choose this scope of mission language is a source of bias, also, this reaching down to the mid-executive or high civil service staff level.

4) Cross-IC gathering of information is then performed by the guiding group for the document involved - This becomes a major area of deconfliction for assessment and is part of the 'blandifying process'. Different organizations with different specialties have their own way of viewing such things as 'risk' and level of confidence on sources. The most troubling from this is the final vetting of source and *who* gets to do that, as seen here at the end of paragraph 2 on page 3:

Working with their Agencies, reps also assign the level of confidence they have in each key judgment. IC reps discuss the quality of sources with collectors, and the National Clandestine Service vets the sources used to ensure the draft does not include any that have been recalled or otherwise seriously questioned.

This is, because it is handed to the clandestine side of the house, directed at CIA and parts of DIA (especially the field Military Intelligence staff). These are the 'keys to the Kingdom' as an NIO that has multi-source reliable information may have it invalidated due to a single source seen as unreliable, or have the reliability so pushed down as to not be considered as part of the NIE. In theory this is made to remove questionable materials and increase reliability, but in practice this can be utilized as a major way to remove information that the clandestine services, particularly CIA, find to be counter to anything it is looking at.

If one remembers back to the 1980's, it was the clandestine services that confidently predicted the continuation of the USSR up to 2030. INTEL on terrorist groups and Saddam's operations have both demonstrated that the clandestine services have decade long problems, actually going well past two decades if the USSR outlooks are included, plus not realizing how close India and Pakistan were to gaining nuclear power status. These organizations are now the 'gatekeepers' to getting views into the NIE. Instead of an all-source reliability assessment, including open source, this reliability and authenticity gatekeeping position has been, basically, handed to the CIA.

5) The draft NIE is sent to the National Intelligence Board for review, rework, re-draft, until the document can be finalized. That is a long process.

6) The summation is to demonstrate, again, what the ODNI has done so that it can justify its next budget.

That is *not* a humorous jab at ODNI, but an actual reason and rationale to put such language into such documents: if the ODNI were TRANSPARENT to the process, it might find its budget getting squeezed in the next budget cycle. By putting in such verbiage throughout the document, ODNI justifies its staff, funding and any and all increases it is looking for to Congress and the Executive.

That is a 'Turf War' game.

Page 4 hands us the TOR scope of the document. Funny its not called that.

That scope statement is part CYA, part process justification and some of it is 'real', meaning that at some point in the staff process (not the analyst process, but the high level NIC/NIB/ODNI staff) a decision is made to say: if we remove all our previous NIEs and positions from our vies, what does this stuff say to us now?

Good and critical analysis of product evaluation (product being the reports from analysts as filtered through Agency staff and then filtered through, again, to executive staff) is vital as well as trying to divorce oneself from past documents and give a fresh review on things. This has been a problem for previous NIEs as NIOs are more wedded to previous views than is the executive staff. And the NIOs also have a better handle on their own source material than the executive staff, which can be troublesome if another NIO, CIA say, invalidates some or all of a view from its perspective and does not take into account the originator's suite of sources for a view. At the highest level of the NIE the ability of analysts to actually get critical information through the multiple levels of 'oversight' is diminished to near nothing.

With all of that said, the proviso of not examining Iranian 'intention' but its 'capability' becomes clear: is Iran capable of processing nuclear material through to a finished device? I will address that problem later, after the examination of the conclusions, such as they are.

Page 5 is one of the very, very few times I have ever seen the IC come forward with its direct view on 'level of confidence' (LOC)! Frame it as a first of its kind. It is given as the 'estimates of likelihood' but is a good shading of LOC and follows the same concept for the same reason: they both measure uncertainty. Estimates of likelihood are based on original analysis which has an LOC and then combined with other products, which change the estimate and then the overall thing is given a final, very fuzzy LOC.

One of the huge sticking points inside the IC is varying LOC based on INTEL type. SIGINT, on the one hand, is very direct: the signal from a given point was interecepted, decrypted and reviewed. That has a high LOC to it. Interpreting who sent it and to whom has its own LOC which starts to move it from 'certain' into shades of gray. Add in other possible sources for sending such messages and LOC goes down, yet again. Throw in analyst knowledge of source type, use, and analyst background and the LOC gets moderated, but usually downward. This works by the concept of multiplication of percentages: two items that relate to each other at 90% LOC yield a sub 90% likelihood (although not exactly 81%) and that, combined with other documents with other likelihood estimates will then moderate downwards from the 81% (if it is the highest) until a final and synthetic LOC is formed.

Now put a 'gatekeeper' on vetting other NIO's LOCs and you get an idea of how the CIA can bolster or reduce the overall LOC for any given view, without, itself, being a primary shareholder in an INTEL area.

That brings us to page 6 and the actual assessments!

What are they saying, given ALL of the above?

1) Iran had been persuing nuclear weapons technology up to the fall of 2003.

2) A generalized halt in the nuclear weapons program started in Iran and lasted for many years.

3) Iran had not restarted its nuclear weapons program as of mid-2007 and we cannot tell its intentions past that point.

4) Iran does not now have a nuclear weapon (as opposed to radiological).

5) The assessment is that Iran has stopped its work directly on devices and nuclear weapons since mid-2005 and attributes this to international pressure.

Those are the 'moderate to moderately high' estimates of likelihood, save the last which is bureaucratic interpretation not linked to NIO analysis as it has NO estimate on it. Yes, all of point 5 is internal view *only* backed by high level summation from multiple organizations, but lacking an estimate of likelihood it becomes something else.

A guess.

The few 'low to moderate' points are telling about what we are missing in our IC:

We continue to assess with low confidence that Iran probably has imported at least some weapons-usable fissile material, but still judge with moderate-to-high confidence it has not obtained enough for a nuclear weapon.

We have no idea if Iran has been importing processed fissionable materials from other countries, like North Korea, say, via clandestine means.

That is extremely painful to read and points out the paucity of HUMINT and reliable cross-INT analysis on the third party arrangements that would indicate an AQ Khan similar (only in type, not personnel or outlook) network for fissionable material. In doing research for an article on the Red Mafia, the number of hints given to organized crime having access to such material in Russia is worrying, particularly as many firms are owned by state permission, invested in by organized crime and white industrial organizations and then have a portion of their output filter into underground networks. It is highly damning to the IC to admit this failure: we cannot even figure out if *anyone* has been *capable* of supplying such material in a clandestine way.

This is not good.

The next point on Iran, internally, getting enriched Uranium to the amounts necessary for a device cannot happen before 2009.

After that Iran is seen as working on generalized high level nuclear capacity that can be 'dual-purpose' in scope. Intention unknown. That intention unknown also goes to all internal workings of the IRGC in the nuclear area, although I would say with so many captures of 'Secret Cells' leaders tied to the IRGC, that we are getting a good look at the overall organizational structure of the IRGC.

Point F on the last page examines that Iran doesn't have enough internally produced HEU to produce a Uranium bomb, and that, in point G, a Plutonium device is off in the 2015 timeframe.

Their point about Iran being able to produce a nuclear device with time and commitment is well taken, but, as they haven't bothered to address intentions, that leaves us a bit out of luck on figuring out what, exactly, they are up to and *why*.


What has been left out?

Pretty obvious, really: Syria.

Iran is not alone in this endeavor and, as I have documented before, Syria has expertise on basic enrichment and processing, plus long term skill in working with scientific groups. As reported in the Jerusalem Post, the site bombed by Israel was most likely *not* a nuclear reactor site but a nuclear device construction site or research site into how to make nuclear weapons. The first report of this site was given to us by Ray Robison last year, before the bombing, with a report in a Kuwaiti newspaper that has been generally reliable on reports from Syria:

A Syrian nuclear program managed by Iraqi and Iranian scientists in the Al Haska area.

Brussels- From Hamid Geriafi

European intelligence sources based at the headquarters of the European Union in Brussels released information yesterday about the “presence of an active Syrian nuclear program in a secret location Northeast of the country, being supervised by nuclear scientists from Iraq, Iran and some scientists of the previous Muslim Soviet Republics, and it seems that the program has reached the stage of medium activity”.

The information, which Al seyassah got from British security sources in Brussels, revealed that “the brother of the Syrian president, Major Maher Al Assad -in charge of the Republican Guard brigade- along with his maternal cousin of the Makhlouf clan are supervising this program since the end of 2004 in the Northeastern district of Al Haskah situated next to the Iraqi and Turkish borders and which has a Kurdish majority”.

The information - of which the British security sources hinted that it might be coming from the German intelligence which is active in some Middle Eastern countries- revealed that “the Syrian nuclear program is relying on equipment and materials that the sons of the deposed Iraqi leader, Udai and Qusai supervised their transfer to Syria by using dozens of civilian trucks and trains, before and after the US-British invasion in March 2003”. Therefore according to those sources, the international inspection teams or the British and US intelligence could not find one “nuclear needle” in Iraq even though everybody knew about the existence of a huge program since the end of the seventies, before the Israeli planes struck and destroyed its main reactor in the Tuwaitha area near Baghdad in 1981.

The British security sources in Brussels assure Al Seyassah that “Iranian nuclear scientists are cooperating with their expertise, equipment and materials in addition to approximately 60 Iraqi nuclear scientists” that had found refuge in Syria after the onset of the war in Iraq and around 20 scientists that had moved from the former Soviet republics at the beginning of the nineties after the collapse of the Soviet Union and after the first government of Gorbatchev pulled out its nuclear weapons from those new republics”.

The sources said that “the Iranians are supporting the Syrian nuclear program, which was originally built on the remains of the Iraqi program after it was wholly transferred to Syria, with materials, equipment and expertise which are more advanced than it originally was. The Iranian could have brought up an advanced plant to enrich Uranium, of which the West is totally unaware.”

Add that to the missing Mitutoyo separators from the AQ Khan network, that I found via TMC net (although the link is probably dead by now):

31 AUG 2006 - Mitutoyo exported 10,000 devices since 1995, most of them illegally+

(Japan Economic Newswire Via Thomson Dialog NewsEdge) TOKYO, Aug. 31_(Kyodo) _ Mitutoyo Corp., a precision instrument maker at the center of an export scam linked to weapons of mass destruction, has exported some 10,000 precision measuring devices, most of them illegally, since around 1995, investigative sources said Thursday.

The Metropolitan Police Department's Public Safety Division is investigating the possibility that some of these instruments were exported to North Korea and other nations suspected of developing nuclear weapons via a nuclear black market formerly run by Pakistani nuclear physicist Abdul Qadeer Khan, the sources said.

In addition, the police are investigating suspicions that a Mitutoyo-made precision measuring machine of a different type from the firm's three-dimensional measuring machine that went to Libya via Khan's smuggling network was exported to an Iranian firm suspected of links to Iran's nuclear development program, they said.

On Aug. 25, the police arrested Mitutoyo Vice Chairman Norio Takatsuji, President Kazusaku Tezuka, and three other executives on suspicion of illegally exporting two high-tech measuring devices convertible for use in the manufacture of nuclear weapons to Malaysia in 2001.

One of the two 3-D measuring machines was found in a nuclear research facility in Libya by International Atomic Energy Agency inspectors during their 2003-2004 checks.

The machines can be used to make centrifuge machines to enrich uranium for nuclear weapons by determining their dimensions and minimizing shape distortions with high accuracy. Their export is subject to restrictions under the Foreign Exchange and Foreign Trade Law and the Export Trade Control Ordinance.

The police found that Mitutoyo's two other precision measuring instruments -- a form tracer for measuring the roughness of the surface of products and a roundness tester -- were handed over to Scomi Precision Engineering Sdn. Bhd. of Malaysia, suspected of being at the core of Khan's network, the sources said.

The police are investigating fresh suspicions that the two devices were exported to Libya as a "three-item set" with one of Mitutoyo's two 3-D measuring machines, which had been found to have been shipped via Dubai to Libya on an Iranian-registered vessel, they said.

The sources knowledgeable about investigations at the Public Safety Division said the Mitutoyo management decided to expand exports in the first half of the 1990s when its sales dived.

Then its project team developed computer software to make its precision measuring machines appear less accurate than they are to bypass Japan's export regulations on high-tech products convertible for the manufacture of weapons of mass destruction, the sources said.

Although some Mitutoyo employees alerted the management to the illegality of exporting high-performance machines using the software, referred to with an in-house code name of "COCOM," Takatsuji and others silenced the critics, telling them such exports have been decided as "a company policy," they said.

In addition to disguising the precision machines as devices with lower capabilities than they actually have, Mitutoyo filed export permit applications with customs, in which they falsely said their overseas arms were the final destinations of the machines to bypass the regulations on exports to countries and firms suspected of developing weapons of mass destruction, they said.

Mitutoyo, based in Kawasaki, Kanagawa Prefecture, is a leading maker of high-tech precision measuring machines and runs a network of subsidiaries, research institutes and factories in more than 20 countries, including the United States, Europe and Asia.

Takatsuji and Tezuka acknowledged the allegations related to the exports to Malaysia for which they had been arrested, informed sources said, adding the two executives have denied that the management ordered employees to bypass the export regulations.

But a senior official at the Public Safety Division said they believe the management issued such orders, adding Mitutoyo "has placed priority on just boosting sales and I cannot find any sense of ethics in them as a company."

Copyright 2006 Kyodo News International, Inc.

And that Iran only has 3,000 separators and, by all reports, Libya having less than 1,500, and we have a major problem of missing high tech nuclear separators roaming around the planet.

Yes, indeed, Iran, internally has sufficient capacity to do this work, but, no, we don't think they are.

And we have no good handle on clandestine networks on the criminal side going into and out of Iran.

We have no mention of Syria in the NIE, yet Israel was sufficiently worried to have Syria come out as an up and coming threat that needed to be addressed with a bombing raid.

It is always fascinating what the NIE does say... even more fascinating in how the bias of what is chosen to be looked at, how it is looked at and the blandness of the conclusions seem to arrive at odd times.

Iran, alone, is not acting in a manner to demonstrate its ability to create nuclear devices.

Iran and Syria?

Well this paper doesn't address that, now, does it?