Wednesday, December 05, 2007

Sources and Methods

The growing imbroligo over the Iran NIE is raising new concerns about the intelligence sources and methods used to produce the document--and its controversial judgments.

Individuals familiar with the assessment suggest that "new" information led analysts to conclude that Tehran halted its nuclear weapons development program in 2003. That represents a sea change from the previous NIE (issued in 2005), which concluded that Iran was actively pursuing nuclear weapons.

While there are legitimate concerns about the personal biases and suspected political agendas of the study's primary authors, Americans should also worry about the quality and reliability of the information used in formulating the assessment. To paraphrase Michael Ledeen, believing the NIE's key judgments means that the evidence has to be awfully good. And evidence of that quality has been in notoriously short supply, both from our sources in the Middle East, and the spy agencies tasked with sorting it out.

This much we know: the full version of the NIE covers 150 pages, including appendices and other supporting documentation. The report's key judgments section, declassified earlier this week, runs only four pages, including a chart that highlights key changes between the latest assessment and the 2005 version. Without the declassification of some supporting data, we can only accept the conclusions of an intelligence community with a poor track record on WMD matters, particularly among rogue states.

As with any National Intelligence Estimate, we assume the new Iran assessment makes use of the full array of intel sources and methods--SIGINT, HUMINT, IMINT, MASINT and even open-source reporting. But we also recognize that information from these same sources led to a dramatically different conclusion just two years ago. Moreover, the volume and quality of collection from these platforms has not improved dramatically--as far as we can tell. Technological refinements in our intel systems are offset by the adversary's own advances, and their attempts at denial and deception.

Consider the example of signals intelligence, or SIGINT. The National Security Agency (NSA) remains the preeminent SIGINT organization in the world. But agency veterans will tell you that the SIGINT environment has become increasingly challenging, thanks to the proliferation of fiber-optic technology and low-cost encryption devices. Phone calls and other communications that once bounced between relay towers are now routed over fiber-optic cable; intercepting them means tapping into the line, a difficult proposition in places like Iran, Syria or North Korea.
The problem is further compounded by wide availability of personal encryption devices. Complex cyphers that were once the exclusive property of governments and intelligence services can now be downloaded from the internet. Increased use of these systems and devices slows the decrypt of adversary communication--and the flow of information to decision-makers.
What about cell phones, you ask? They operate on a tower-based line-of-sight system. True, but intercepting those transmissions (usually) means getting inside hostile territory, further complicating the collection task.

And SIGINT isn't the only intel discipline facing such challenges. In the internet era, there are scores of websites that offer information on the orbits and potential collection "windows" for spy satellites. Today, an effective satellite warning program is just a few keystrokes away--and there's little we can do about it.

HUMINT? The Robb-Silberman Commission Report (released in 2005) deplored the state of our HUMINT capabilities, noting the meager haul on Saddam's WMD programs in the run-up to the war, and the wholesale lack of reporting on Iraqi leadership intentions. Similar problems are said to exist with the Iranian "target," which presents similar challenges. Clearly, there have been no major "breakthroughs" in traditional collection methods over the past two years.

What changed (apparently) was the receipt of new information from well-placed sources that forced a revised assessment. We maintain that the most likely source for this information was General Ali Rez Asgari, who defected to the west earlier this year. As the long-time commander of the Iranian Revolutionary Guards Corps (and a Deputy Defense Minister), Asgari had access to a wide range of highly sensitive programs, including his country's nuclear efforts. Officials who spoke with Bill Gertz of the Washington Times have also hinted that Asgari was the source for the new information.

If General Asgari's reporting is the primary reason for the revised assessment, it's almost certain that he smuggled reams of information out of Iran. Having been burned by "single sources," before, it's unlikely that the intel community would assign such credence to his data--unless Asgari had documentation to support his claims. That raises obvious concerns about plants, deception and double agents, but (so far) General Asgari's claims have apparently withstood scrutiny.

We should note, however, that the defector may not be the primary source for the new NIE and its startling judgment. In his latest dispatch for Newsmax, Kenneth Timmerman offers this disturbing revelation:

[The NIE's] most dramatic conclusion — that Iran shut down its nuclear weapons program in 2003 in response to international pressure — is based on a single, unvetted source who provided information to a foreign intelligence service and has not been interviewed directly by the United States.

Newsmax sources in Tehran believe that Washington has fallen for “a deliberate disinformation campaign” cooked up by the Revolutionary Guards, who laundered fake information and fed it to the United States through Revolutionary Guards intelligence officers posing as senior diplomats in Europe.

And whatever the source, the information prompted a rapid change in the intel community's views on Iran's nuclear program. As Thomas Joscelyn reports at the Weekly Standard (H/T: Ed Morrissey), one of the NIE's key authors offered a completely different take on Tehran's nuclear efforts just five months ago. Testifying before Congress on 11 July of this year, Thomas Fingar of the NIC stated that Iran was still pursuing nuclear weapons:

Iran and North Korea are the states of most concern to us. The United States’ concerns about Iran are shared by many nations, including many of Iran’s neighbors. Iran is continuing to pursue uranium enrichment and has shown more interest in protracting negotiations and working to delay and diminish the impact of UNSC sanctions than in reaching an acceptable diplomatic solution. We assess that Tehran is determined to develop nuclear weapons--despite its international obligations and international pressure. This is a grave concern to the other countries in the region whose security would be threatened should Iran acquire nuclear weapons.

Compare that to key judgments from the NIE that Mr. Fingar helped prepare:

"We judge with high confidence that in the fall of 2003, Iran halted its nuclear weapons program."

"We judge with high confidence that the halt lasted at least several years."


"We assess with moderate confidence that Tehran had not restarted its nuclear weapons program as of mid-2007, but we do not know whether it currently plans to develop nuclear weapons.

Obviously, any intelligence estimate is only as good as the information it's based on. Political agendas and personal biases aside, it's clear that the bottom-line assessment of the new NIE raises questions about the quality and reliability of its source data. No one can reasonably expect the intel community to reveal all sources and methods that were used in generating the report. However, it is not unreasonable for lawmakers--and the public--to demand a more detailed explanation as to how intelligence analysts arrived at their astounding conclusion, and the data they used to support that assessment.

Maybe it's time for another exercise in competitive analysis. Intel vets will groan about the inaccuracies of the CIA's "Team B" in the 1970s--and they certainly have a point. But it's also worth remembering that Team B was right on more than a few counts. Simply stated, we cannot afford to be wrong on the Iranian nuclear issue. Maybe it's time for another set of eyes to take a look at that revelatory information that prompted the new assessment.


section9 said...

Well, I'd feel a little more comfortable if you had sources other than NewsMax. I assume they got the report off the wire.

OMG, if Fingar depended on an unvetted source, this thing will fall apart, but we won't know it until it is too late.

Ken Prescott said...

[The NIE's] most dramatic conclusion — that Iran shut down its nuclear weapons program in 2003 in response to international pressure — is based on a single, unvetted source who provided information to a foreign intelligence service and has not been interviewed directly by the United States.

Welcome to Mister Rogers' National Intelligence Estimate!

It's a beautiful day in the intelligence community . . .

Hello, boys and girls! Can you say "Curveball?" I knew you could!

GeorgeH said...

Tehran halted its nuclear weapons development program in 2003.


previous NIE (issued in 2005), which concluded that Iran was actively pursuing nuclear weapons.

Not necessarily a conflict.
If Iran managed to buy working plans (Pakistan, North Korea)in 2003 then they would have stopped research on building a bomb and concentrated on separating U-235, which is what they seem to have done.

Ken Prescott said...

Even the New York Slimes is throwing a BS flag on this play. Says it all.

davod said...

Michael Tanji (A former intel proffessional)at Danger Room has a different take on this:
Intel Insider: Iran Report Ain't Political

Let the battle commence.

My feeling is that the three wise men who controlled the summary may have massaged it so the headline read what they wanted.