Playing to the Edge: American Intelligence in the Age of Terror (33 page)

A very smart, really hardworking person I relied on to be my chief of staff was Larry Pfeiffer. He was career NSA on loan to the CIA front office when the intelligence reform act hit in 2005. With all the turmoil from the legislation, he was almost the IC equivalent of a stateless person when I grabbed him to be my ODNI chief of staff, and I happily kept him when I moved on to CIA. Smart move.

Another NSA veteran I enlisted was Bob Deitz, who had been my general counsel at Fort Meade. At Langley he was consigliere, a kind of gadfly in residence who stayed behind after a meeting to challenge everyone’s thinking.

None of us had any grandiose schemes for reinventing CIA. There was a war on. We thought our job was to right the ship and keep it sailing. After the 2008 presidential election, Congressman Mike Rogers of the HPSCI told his old Hill friend and incoming White House chief of staff Rahm Emanuel that our crew had gotten the trains to run on time. We took it as a real compliment.

Early in our tenure, Kappes, Morell, Pfeiffer, and I did have one short off-site to set out some goals. We decided one area of emphasis would be agency culture.

Driving down Virginia Route 123 and looking through the fence line, CIA appears to be a singular noun to a casual observer.
The
CIA. It is not. On a
very
good day, CIA might be a collective noun. Most days, it is decidedly plural.

There are four big directorates in the agency, each with a culture of its own. The self-confident case officers in the National Clandestine Service reminded me of air force fighter pilots. Tenured faculty was the thought that came to mind when talking to some DI analysts. The folks in S&T (science and technology) were true heirs to James Bond’s “Q.” And the Directorate of Support housed the blue-collar, just-make-it-happen kind of folks I grew up with in Pittsburgh. They made the agency
run—communications, logistics, security, and everything else we needed to operate.

All good. Each culture was geared to each directorate’s specific task. In a Bubble session I explained that I wanted to strengthen everyone’s CIA identity, but not at the expense of the directorates. I used my air force experience to elaborate. I told the audience, “America’s air force spends the better part of most days trying to convince its fighter pilots that they are actually part of a larger institution. But at the end of the day, there is one thing they want them to remember. They really are fighter pilots.”

And the agency wasn’t nearly as internally fenced off as it had been in the past; at one point in the agency’s history, case officers and analysts actually had separate cafeterias. Still, we thought that we could use a little more overarching “CIA-ness” in the mix.

One initiative was to set up a very small CIA subsidiary operation in Austin, Texas. Office space was cheap; Texas is on an independent power grid, so it reinforced continuity of operations; air service to DC was good enough; and there was a lot of great talent to recruit in the local area. The idea was—in the new location—to experiment with a structure that ignored the old directorate lines, a kind of small CIA-X where we could safely experiment with new approaches. We wanted to see what CIA would look like if we had a blank sheet and no legacy organizational structure, contractor base, IT system, or workforce. We never found out. Congress stopped us in our tracks. We were probably too late in the administration for them to countenance any new big ideas. But it also serves as something of a morality play when people complain about government inertia and inflexibility.

So we had to content ourselves with other steps. One technique was to communicate to everyone directly and simultaneously, the theory being that common knowledge would help build more common attitudes. My rate of e-mails to the workforce was comparable to the pace of DIRGRAMs I had set at NSA, about two per week.

We set up a special account called “Ask the Director” so that questions
and ideas (and gripes) could be aired internally, instead of being first heard from “anonymous sources” in the
Washington Post
. We got pretty heavy traffic for about ninety days, responded to all of it, and then saw the pace gradually taper off.

Many in the current workforce (and the press) often turn to former agency officers for their views, so we tried to rope in the alumni as well. We got pretty good crowds for alumni days in the Bubble, and although we couldn’t clear them for classified briefings, Steve and I leaned pretty far forward telling them what we were trying to do.

We also tried things like lengthening the time in the on-boarding process where new arrivals got acculturated to the agency as a whole before heading off to their directorates.

We emphasized our common agency heritage and dotted most of 2007 with events marking CIA’s sixtieth anniversary.

We also gradually pulled corporate activities like the ops center, information technology, finance, and HR out of the directorates and up to the agency level—the better to see, manage, and unify these important functions.

I think I surprised everyone my first summer by demanding to review how they were burning their money before the fast-approaching end of the fiscal year. We still didn’t have auditable accounts (no one in the IC did) nor could we really tote up the real cost of manpower, but I at least made the point that I thought it was all
CIA
money, whereas in the past the wheelbarrow had dumped off funds to the directorates at the beginning of the fiscal year and that was the end of it.

We also took a bit of a run at contractors, the number of which had grown massively since 9/11. Unlike some, I wasn’t mad at contractors. I considered them part of the workforce and managed them as such. I reminded folks that during my tenure, I put contractor stars on the wall. There was no doubt that our use of them made us more effective.

That last point was often contested by members of Congress, who accused us of using contractors to deflect accountability for our actions. Often, when I briefed them on an operation, they would ask whether it
had been accomplished by a contractor or a government employee. I rarely knew and rarely went out of my way to find out. I usually took the question for the record, pointed out that the individuals involved were the best Americans available for the task, and then reminded them that I was as liable for contractor actions as I was for those of agency officers.

But if our use of contractors had been effective, no one could claim that it had been efficient. We often found that we were bidding against ourselves for contractor support, with different parts of the agency pursuing the same product or service from the same vendor. We were also a bit of a farm team for our contractors. Agency folks were leaving, with their experience and their clearances, for higher-paying contractor jobs.

I arbitrarily directed that we were going to reduce our reliance on contractors by 10 percent. There was no science behind that, but I knew there was at least that much slop in the system. When we met that goal in less than a year, I added an additional 5 percent cut.

I also said that CIA would not renew for one year the clearances of anyone who resigned from the agency to come back to work for us as a contractor. Retirees could—they had paid their dues—but not people who were just selling back to us what we had just taught them.

But that was on the dramatic end of my actions. I had promised the agency on my first day that no one had sent me up the parkway to blow anything up. I was true to my word. It was a way of telling people that I had confidence in them, that they weren’t terminally screwed up, as some of their critics were claiming.

In that introductory conversation with Leon about the quality of the staff, I added, “If you give them half a chance, they won’t let you fail, the way they wouldn’t let me fail.”

That actually applied to everyone at CIA, not just the senior staff, and Leon would learn soon enough why I was so confident in them.

Because my thirty minutes to brief President Bush on CIA’s covert actions and sensitive collection were on Thursday mornings, you can imagine what the headquarters looked like on Wednesday afternoons, as items suggested by the staff began to flow into my office. I would do one review
midafternoon and then another just before I left the office for the night. Then I would wake up very early on Thursday, spread the potential items out on my kitchen counter, and decide which were and were not ready for prime time. The important point is that easily 50 percent of the specific operational details I briefed on Thursday morning I learned about for the first time on Wednesday afternoon.

I felt fully in charge. Steve Kappes and I set the left- and right-hand boundaries of acceptable activities. We ensured that actions were purposeful and directed toward defined and legitimate objectives. But in the world in which we were operating, we had to depend a great deal on local autonomy and judgment. There was no other way. What we were asked to do couldn’t be done with tight stick and rudder control from a seventh-floor suite in Langley, Virginia.

If we were a military organization, this would all have been called “mission type orders”: broad direction and clear limits from headquarters. The rest is yours.

In a religious context, since we focused on removing impediments and enabling, we might have described ourselves as “servant-leaders.”

At CIA we were director and deputy, charged to lead, decide, discipline, and (one hopes) inspire.

D
uring my time at CIA, the most discussed topic in the Oval Office was terrorism and the ongoing wars. Number two was Iran. It wasn’t all that close, but we talked about Iran a lot. There really wasn’t a number three. Sure, we addressed other topics, but we didn’t aggregate around any of them enough that they actually earned third place.

When asked by public groups what our intelligence priorities were, I would respond with a bit of Washington alphabet soup: CT-CP-ROW. Counterterrorism, counterproliferation, the rest of the world.

So counterproliferation was a big deal, and during the transition when the president-elect asked me how much of our CP effort was focused on Iran, I answered without hesitation, “Eighty percent.”

So
Iran
was a big deal, and President Bush was losing patience about it. He freely admitted the near impossibility of persistent penetration of a closed society like North Korea but wondered—with literally thousands of Iranian Americans able to routinely travel to their homeland, an economy that was largely integrated with the global economy, and an intelligentsia that was extensively Westernized—why we weren’t doing better with Iran.

I didn’t have a good answer. Iran was a tough target. Their counterintelligence services were numerous, large, and thorough. The MOIS (Ministry of Intelligence and Security) was the FBI’s counterintelligence equivalent and a whole lot more. Nominally under the president, in practice it reported directly to the Supreme Leader. It was augmented by a shadowy arrangement with the IRGC (Iranian Revolutionary Guard Corps), formed by Ayatollah Khomeini after the 1979 revolution to guard the Islamic Republic against internal and external threats. Thought to number more than 100,000, the IRGC was the fanatically loyal, direct-action arm of the Supreme Leader. Domestically, the IRGC controlled the million-man Basij Resistance Force, an army of street thugs that delighted in cracking the heads of dissidents, especially the children of the privileged from cosmopolitan, impious North Tehran.

The IRGC’s many external operations were under the control of the Quds Force and its dark, ruthless commander, Qasem Soleimani. And all of this was wrapped so powerfully into the rule of religious law (
velayat-e faqih
) rather than secular jurisprudence that it seemed that no action was off the table. Amazing what people will do when they think they’re acting on the will of God.

We had some successes, increasing the number of what we called operational acts—the things you need to do to conduct espionage—inside Iran, but we were more than a brick shy of a load.

George Bush rarely showed anger toward me, but one day in the Oval after an unsatisfying discussion on Iran, he intercepted me as I was leaving, and with his right index finger pointing toward my chest, said, “Mike, I don’t want to be left—I don’t want any president to be left—with only two choices here.”

The two choices were bombing or the bomb: going “kinetic” against the Iranian nuclear program or acquiescing in a nuclear-capable Iran.

The president was making a classic plea to his intelligence community. If knowledge is power, he seemed to be saying, work your magic to give me knowledge so I can develop options beyond being forced to accept the unacceptable.

Some very smart Americans have actually made it clear that
they
could accept one of the options. Brent Scowcroft and Zbigniew Brzezinski, both former national security advisors, believe that a nuclear Iran could be contained or at least that the alternative of bombing Iran would be worse. But that was not the policy of either President Bush or President Obama and, not that it mattered much, I agreed with the presidents. Now, what to do about it?

It was complicated.

On the margins of an NSC meeting late in the Bush administration, a senior official cornered me and asked what I thought Iranian nuclear doctrine might be. I replied, accurately enough, “I have no idea and I doubt that they do either.”

My questioner persisted. “How many weapons might they get?”

“Probably just a few.”

“And how many do
we
have?”

“Oh, all told, thousands.”

“So why don’t we just deter them?”

At which point the tumblers clicked, and I began to better understand the conversation. “Oh, this isn’t about deterring them,” I objected. “This is about deterring us. Look at their behavior—with Hamas, with Hezbollah, in Iraq, in Afghanistan. Hell, we judge that it is the policy of the Iranian government—approved at the highest levels of that government—to facilitate the killing of young Americans and other allies in Iraq. And this is before they have one of these things in the garage. Imagine what they might do with this as a trump card.”

Iran was already an incredibly destabilizing force in the region, especially in Iraq. The Badr Corps, one of the most important Shia militias in Iraq, was an arm of Iranian policy. Tehran had also enlisted the aid of its Lebanese proxy Hezbollah and its murderous operations chief, Imad Mughniyah, to create a Hezbollah clone in Iraq, an effort supported by several training camps in Iran. The IRGC Quds Force was also aggressive on the ground. It helped plan a deadly raid—complete with faked
identity cards and American-style uniforms—against a US checkpoint near Karbala in January 2007 that resulted in five soldiers killed (four of whom had been kidnapped).

The Quds Force was also manufacturing and shipping EFPs—explosively formed projectiles—to Shia militias. These were ingeniously designed shaped-charge devices that could penetrate even the thickest American armor. EFPs were the biggest killers of Americans in Iraq, and we knew, with high confidence, that this was the Iranian government’s intent.

And they weren’t bashful about any of this. Soleimani famously sent a text message to Iraqi president Talabani for the US commander: “General Petraeus, you should know that I, Qasem Soleimani, control policy for Iran with respect to Iraq, Lebanon, Gaza, and Afghanistan. The ambassador in Baghdad is a Quds Force member. The individual who’s going to replace him is a Quds Force member.” Such remarkable confidence, even a touch of messianic triumphalism, and this
without
a nuclear weapon.

But what to do? The first issue was to determine what
they
were doing. Acquiring a usable nuclear weapon requires three things: fissile material (i.e., the uranium or plutonium that creates the detonation); delivery systems, like ballistic missiles, that can get a weapon to a target; and finally, the warhead itself, sufficiently hardened and miniaturized that it can fit on a missile, survive reentry, and be detonated with confidence.

The toughest task of the three is manufacturing the fissile material, and blessedly that requires an industrial process complex and extensive enough that it’s hard to keep completely hidden. So we followed it as best we could. How many centrifuges? How much enriched uranium? To what levels? On what timelines? At which sites?

We weren’t bad at this. We were on to the unannounced uranium enrichment facility near the holy city of Qom well before it approached operational status and well before the Iranians knew that we knew. When the Iranians finally suspected that we were on to them, they sent a vaguely
worded letter to the UN’s nuclear inspectors in Vienna prompting President Obama and his British and French counterparts to out the illicit nuclear site during a September 2009 G20 summit in Pittsburgh.

The Qom facility was especially troubling. Built in secret, on an IRGC base, under a granite mountain, it could accommodate about three thousand centrifuges. That was too small to support the enrichment needed for a nuclear energy program. It was too big for some sort of pilot project. But it was just right for producing enough enriched uranium for a modest weapons effort.

With a boost from International Atomic Energy Agency (IAEA) reporting, we worked to get a good handle on the production of enriched uranium at Natanz, Iran’s main facility. We needed to know how much had been enriched to 3 to 4 percent, the level needed for a civilian energy program, and how quickly that could be converted to 90 percent purity, the level needed for a weapon.

In 2010 the demand for information increased even more when Iran began to enrich to 20 percent, ostensibly for a research reactor in Tehran, still well short of the 90 percent needed for a weapon, but—because of the peculiar physics involved in enrichment—about nine-tenths of the way to weapons-grade material. We relied on technical intelligence to vet and supplement what they were learning on the ground.

There was a second set of questions. If we were going to influence Iranian decision making, we would have to have a pretty good idea of how they made decisions. President Bush was consistent on this theme. He wanted to know what buttons he could push, but Iranian decision making remained very opaque to us. I always preferred questions on the nuclear program itself.

Not having an embassy in Tehran didn’t help. The Iranians would have labeled it a nest of spies, but I would have settled for some smart political officers able to openly absorb information as they walked through the bazaar. We were relegated to asking countries that did have legations in Tehran what their diplomats were hearing and seeing.

Late in the administration, Secretary Rice occasionally floated the
concept of a US-staffed interest section in Tehran, much like we had in Havana. I always quickly seconded the motion even as the vice president just as quickly counseled against it as unnecessarily legitimating the regime.

The two strains—where are they in their program and how do they make decisions about it?—came to a head in a National Intelligence Estimate being prepared in the summer of 2007. It was meant to replace the previous estimate, which confidently assessed that “Iran currently is determined to develop nuclear weapons despite its international obligations and international pressure.” A new draft had been prepared along similar lines, although the confidence level had eroded from high to medium, not because of any contrary evidence, but simply because the evidence we did have was aging off.

The original target for publication was spring, but we were unable to meet that as we re-scrubbed all our data, sources, assumptions, and judgments. We were also, at presidential direction, putting a lot more effort into collection, and that began to pay off. New data began to suggest that in 2003 Iran had suspended its work developing a weapon, perhaps in response to much of its nuclear program being outed the year before, or perhaps because there were two large American land armies to the east and west of Iran in Afghanistan and Iraq.

In any event, for whatever reason, the Iranians appeared to have stopped work on developing a warhead. They were still chugging away creating fissile material and developing missile delivery systems, but not the weapon itself. And this judgment was based not on the absence of evidence that such work was ongoing, but rather on evidence that it was not.

None of us thought that this ended our Iranian nuclear dilemma, but it did muddle the previous NIE’s key judgment about Iranian intentions.

In the new intelligence structure, NIEs are the province of the DNI, not CIA. But the grunt work on this one was being done by CIA analysts, so my deputy, Steve Kappes, and I spent two afternoons grilling our analysts about their work. This was going to be a big deal. We had to be right.

We hit them from every direction we could think of. How many sources? What kind? Assumptions? Alternative explanations? They had mastered their facts and their brief. Like any analysts, they could have been wrong. But they clearly knew their stuff.

As one final check we threw their conclusions to the CIA Counterintelligence Center and asked if all of this could be Iranian deception. These people were paid to be suspicious. And they were. But the worst they could come up with, based largely on a generalized suspicion rather than on any specific facts, was a recommendation to give the judgment on halting weaponization no more than a medium confidence level. But that’s not what the evidence said. We stuck with high confidence.

It was going to be tough to take this to the Oval. DNI Mike McConnell had prepped the battle space by giving the president and vice president a preview of what was coming. But this was a very unwelcome message, one bound to undercut the administration’s efforts to build an international coalition to punish and isolate the Iranians for their nuclear ambitions.

To his credit, though, and at some political cost, the president directed that we declassify and publish as much of the key judgments from the NIE as we could. The administration had used the previous key conclusion,
Iran is determined to develop
 . . ., as an anchor of its policy, so it felt duty-bound to publicly say that that judgment was now less certain. There was also the practical consideration that this was sure to leak anyway. It would be better to get ahead of the story.

It was a simple matter to redact the 15 percent or so of the key judgments that said too much about sources and methods. We let the other 85 percent stand as written. That was the right thing to do. Who could stand the story that the estimate was edited for reasons beyond classification before it was publicly released?

But we were releasing only two pages of key judgments from a 140-page NIE. A lot of important detail was going to be left on the floor. Nuclear issues are also especially complicated; they are difficult to summarize. And no NIEs are written for the general population or even
intelligence-beat press reporters. They’re written for a few hundred select government officials, cognoscenti who know and appreciate their arcane style, language, rhythm, context, and history. Now we were about to throw a much-truncated text out to the general public. All I could think of were bones being thrown before a circle of necromancers, with each free to make his own interpretation.

The administration belatedly figured this out. Mike McConnell was on a foreign trip, so as I was leaving the Shakespeare Theatre in downtown Washington on a rainy Friday night in late November, my security detail chief leaned into me and said, “Mr. Hadley wants to speak with you.”