Playing to the Edge: American Intelligence in the Age of Terror (6 page)

We hurried to set up an alternative site, a physically separate complex that allowed us to replicate about 80 percent of NSA’s capability with about 20 percent of its capacity. That wasn’t a perfect world, but at least it preserved us against catastrophic failure. We decrypted our first message at the new facility in late September 2003. For us it felt a little like driving the golden spike at Promontory Summit in 1869. It let us breathe a little easier.

We had one final option to exercise for continuity of operations. As we approached one holiday season with a particularly high quotient of background terrorist chatter, I called David Pepper of GCHQ to tell him that in the event of catastrophic loss at Fort Meade, we would entrust the management of the US SIGINT system to him and to our senior
representative in London. The long pause on the other end of the secure line betrayed both the gravity of the threat and the enormous burden I was imposing on a friend.

 • • • 

W
HILE WE WERE WORKING
the present and preparing for the future, Congress wanted to question us about the past and how the 9/11 attack could have happened. The House and Senate Intelligence Committees organized themselves into something called the Joint Inquiry Commission (JIC) and began to hold hearings, both open and closed. This was the beginning of a wave of inquiries (the 9/11 Commission, the WMD Commission, and later, similar looks at CIA interrogations) that burned up big chunks of time for some of the intelligence community’s best people.

The big open hearing for the JIC was in October 2002, where George Tenet, FBI director Bob Mueller, and I were at the witness table together. It was an all-day event, and there was later a
Time
magazine cover photo of all three of us with our right hands in the air being sworn in.

It wasn’t as bad as it might have been. The chairmen, Congressman Porter Goss and Senator Pat Roberts, kept a pretty tight rein on things. In the hearing I admitted that, sadly, NSA had no prior knowledge of the 9/11 attack and that we were challenged by a global telecommunications revolution. We had competed successfully with the Soviets. “Now we had to keep pace with a global telecommunications revolution, probably the most dramatic revolution in human communications since Gutenberg’s invention of movable type.” And we weren’t doing very well.

I complained that our resources and people had been cut by about a third in the preceding decade, the same decade when “mobile cell phones increased from 16 million to 741 million, an increase of nearly fifty times. . . . Internet users went from about 4 million to 361 million, an increase of over ninety times. Half as many landlines were laid in the last six years of the 1990s as in the whole previous history of the world. In that same decade . . . international telephone traffic went from 38 billion minutes to over 100 billion. This year [2002], the world’s
population will spend over 180 billion minutes on the phone in international calls alone.”

I admitted that I had under-resourced the counterterrorism mission when it came to linguists and analysts but pointed out that “if these hearings were about a war that had broken out in Korea or a crisis in the Taiwan Straits, if we had been surprised by conflict in South Asia, if we had lost an aircraft over Iraq, or if American forces had suffered casualties in Bosnia or Kosovo—in any of these cases I would be here telling you that I had not put enough analysts or linguists against the problem.”

I ended by congratulating the committee for prompting a needed national dialogue on the balance between security and liberty. It was an oblique reference in open session to the challenges created by the Stellarwind program, a secret collection effort (see chapter 5) then under way but known to fewer than ten people in the crowded committee room. “I am not really helped by being reminded that I need more Arabic linguists or by someone second-guessing an obscure intercept sitting in our files that may make more sense today than it did two years ago. What I really need you to do is to talk to your constituents and find out where the American people want the line between security and liberty to be.”

There, in October 2002, I summed up the question of a decade later pretty well: “In the context of NSA’s mission, where do we draw the line between the government’s need for CT [counterterrorism] information about people in the United States and the privacy interests of people located in the United States?”

After Edward Snowden’s revelations in 2013 (chapter 21), NSA was accused of indifference to and wanton violations of American privacy. Neither accusation was true. By the way, my 2002 question to the Congress went unreported (and largely unanswered).

I was pretty defensive; after all, the subtext of the questioning was pretty clearly, How did you guys let this happen? But I needn’t have been as defensive as I was. They were really going after George Tenet.

They were hammering George hard over CIA’s losing lock on a terrorist pair who had attended a meeting in Kuala Lumpur and then
continued on to San Diego. The pair had lived in California in true name and ended up being two of the muscle guys on the American Airlines flight that hit the Pentagon. The record wasn’t clear about who knew what, or when and what CIA did or did not pass to the FBI. A few members were pressing George for the name of the analyst who had been most involved. George refused. At one point he feigned leaning over to reach into his briefcase, which was on the floor between us, and as his mouth passed by my ear, whispered, “I’m not giving her up. I’m not giving her up.”

It was a daylong hearing. No breaks. For us, anyway. Members came and went. Some got sandwiches from the cafeteria, brought them back, and ate them at the dais. At two o’clock, George turned to his legislative liaison chief and complained, “Are they ever going to feed us?”

They weren’t, so the chief sent out for egg salad sandwiches from the Senate cafeteria. When he delivered them to the three of us still at the witness table and still testifying, it seemed that every cameraman in Washington suddenly appeared in the well between our table and the dais ready to snap a potentially classic shot of witnesses chowing down. The sandwiches remained unopened and we remained unfed.

On the way out from the long day, I was intercepted by a family member of a 9/11 victim. She was still visibly grieving, and the day’s events surely gave her no comfort. I chatted with her for a short while, saying something like we were all doing our best. It was all I had to offer.

The closed session with the JIC was a bit bizarre. It was held in the secure hearing room on the House side with a lot of senators denied space on the dais and instead sitting at small tables and on small chairs in the well. They looked like parents sitting in their children’s grade school classroom.

In this session Senator Richard Shelby from Alabama led the charge against NSA for two intercepts that were made on September 10 but not processed, translated, and reported until September 12. One cut contained “The match is about to begin,” while the other observed, “Tomorrow is zero hour.”

The communicants were affiliated with al-Qaeda, but were not
senior-level folks. The phrases were oblique and embedded in lengthy conversations about a variety of things.

They certainly did not predict that an attack would take place and did not contain any details on the time, place, or nature of what might happen. No New York. No Washington. No airplanes as weapons. The longer we studied them, the more they looked like they might have been about the aftermath of the assassination of Ahmad Shah Massoud, head of the Northern Alliance in Afghanistan, who had been killed on September 9.

In any event, throughout the summer of 2001 we had had more than two dozen warnings like this that something was imminent. We dutifully reported them, yet none of them subsequently correlated with attacks.

Sometimes the absence of an attack was because of what we had reported. In July we tapped a communication between Afghanistan and Saudi Arabia. It was a discussion of a soccer match, and when asked where the match would take place, the Afghan side said, “Somewhere in your region.” We, of course, alerted Americans in the region; some DOD units took appropriate precautions. Later, we intercepted the same communicants, and when asked if the match had taken place, the reply was no, because the spectators were not there.

I tried to explain to the committees how SIGINT works, that “thousands of times a day, our front-line employees have to answer tough questions like: Who are the communicants? Do they seem knowledgeable? Where in the conversation do key words or phrases appear? What is the reaction to these words? What world and cultural events may shape these words?”

I explained that NSA rarely listens to a conversation while it is taking place. Intercepts are collected, stored, and sorted, and then a linguist works his or her way through the queue. That’s what happened with the September 10 intercepts. And the work plan to handle the existing backlog would surely have been affected by shifting priorities as the attacks of 9/11 unfolded.

Now, you might think that the intelligence oversight committees
would already have a pretty good idea about all this. But you would be wrong. They weren’t even buying the explanation now.

Pretty much exasperated, I actually pulled out the transcripts (not the reports, the transcripts) of the conversations we had intercepted. I read them in their entirety, including the small talk, greetings, and a lengthy discussion about needing batteries. “Real batteries,” I said. “These aren’t code words for anything.”

When I was done, I put the paper down and looked at the dais. “So you tell me. Strategic warning or two Bubbas pumping gas at 7-Eleven?”

We actually took a short break for lunch in the closed hearing, and we hurriedly downed some sandwiches as the members scattered. When we resumed in the afternoon, we started to pick up the “tomorrow is zero hour” theme again. But now George and I were getting notes from our staffs that details of the morning’s (closed) session were playing on CNN, including the ominous September 10 phrases.

As luck would have it, Congresswoman Jane Harman was lamenting the leaking of classified information in general when George and I interrupted our testimony to tell Porter Goss, who was chairing the session, what was happening. He was genuinely irate, and said so.

During the hearing we were inaccessible to the press, of course, but back at our headquarters the phones were ringing off the hook with journalists looking for comment—which, of course, could not be provided, since this was still top secret stuff. It was never officially determined where the leak originated. But one correspondent who called NSA did submit a curious question: “What does Shelby [the senator who had interrogated us on the intercepts in the morning] have against you guys?”

 • • • 

E
VEN WITH THE
JIC
DISTRACTION,
we were sharpening our game against al-Qaeda and, frankly, were getting pretty good at it. In March 2002 CIA and Pakistan’s Inter-Services Intelligence agency (ISI) rolled up Abu Zubaida, the first of many al-Qaeda senior leaders we would capture, and later George Tenet was kind enough to call me at home to
both alert me and thank me as Khalid Sheikh Mohammed met a similar fate.

By this time, though, Iraq was also on the front burner. I must admit the occasional dark moment when I asked myself why we were doing that. Saddam was dangerous, but he wasn’t going anywhere. I wasn’t a policy maker, though. I focused on getting any war there over quickly.

S
o, as we were sorting out the aftermath of 9/11 with Congress, we were running up to the war in Iraq.

For better and for worse, NSA had a powerful role in that as well. Clearly there were those in the administration who believed that 9/11 justified an assault on Saddam’s regime. To buttress their case they often suggested a connection between the Iraqi government and al-Qaeda and would sometimes call for more detailed analysis of specific pieces of signals intelligence that they felt supported their case.

It happened often enough that I feared some of our stuff might be taken out of context, so I directed our Iraq folks, whenever they were asked to do that, to put an overall caveat in their reporting along the lines of “Taken in its entirety signals intelligence neither proves nor disproves an operational relationship between the government of Iraq and al-Qaeda.”

Intelligence services, especially those like Saddam Hussein’s, have contacts with a lot of people. Those contacts aren’t a prima facie case of collaboration, cooperation, or subordination. I’m not saying it was unfair or somehow unethical to recommend the hypothesis or to look for evidence to prove it. In the end, though, there just wasn’t a case for it.

But we all thought that there was a case for weapons of mass destruction. I was in George Tenet’s conference room when we voted on the now-infamous National Intelligence Estimate (NIE). I voted yes—on all counts. I was comfortable with the vote, then. I had earlier told Condi Rice, the national security advisor, in a private conversation, that I had a roomful of evidence that Saddam had a WMD program. “A roomful of evidence,” I confided, “all of it circumstantial.”

That was often the nature of SIGINT, indeed the nature of a lot of intelligence. Years later Michael Morell, then deputy director of CIA, was briefing President Obama on the likelihood that Osama bin Laden was at that Abbottabad compound. It turns out that some in CIA viewed the odds at no better than fifty-fifty; others handicapped it at nine out of ten. When pressed by the president as to why the large spread, Michael replied that those who had been involved in the Iraq NIE were the most conservative, while those whose only work was counterterrorism and al-Qaeda were most optimistic. Michael added that in both cases the evidence was circumstantial—and that we may have had more of it for the Iraqi WMD question than we did for Abbottabad.

The intelligence guy is actually most at risk when he is telling the policy maker things he wants to hear. And there is no doubt that Saddam’s linkage to weapons of mass destruction gave the Bush administration its clearest public argument in articulating the case for war. Paul Wolfowitz, the deputy secretary of defense at the time, has said as much. It was the case most easily articulated. It was just wrong.

Years later, discussing this very issue, NSA’s historian surprised me by linking it to the great Civil War battle at Antietam, still the bloodiest day in American history. Union forces there were led by General George McClellan, an incredibly cautious man who always believed that he was outnumbered by Lee. In truth, in this fight he outnumbered Lee two to one and had the Army of Northern Virginia pinned with its back to the Potomac and only one usable ford behind it.

But McClellan was wary and his intelligence chief, Allan Pinkerton (of detective fame), fed his caution with reports of imaginary legions that
Lee had brought with him across the river. Neither man was dishonest, but their tendencies so strongly reinforced each other’s that it is still unclear who was really misleading whom. In any event, McClellan fought scared, committed only a part of his army and then only piecemeal, and allowed Lee to survive to fight another day.

The NSA historian’s clear lesson: when the intelligence is making a policy maker too happy, he ought to challenge it, and when he doesn’t, the intelligence briefer needs to launch a red team against his own conclusions to see if they can stand their ground.

And the intelligence officer with the unhappy message needs to insist, really insist, the way the State Department’s Carl Ford (unsuccessfully) did when we were arguing over Saddam’s aluminum tubes and nuclear program in late 2002. Carl was stubborn to the point of being damn near obnoxious. All he got for it was a footnote that State Department’s intelligence shop disagreed with some of the NIE’s conclusions.

We compounded our error by stating our conclusions in the NIE in language that was far too categorical. No reader of the final product could conclude other than that we were firm in our judgments, despite some thin sourcing of our human intelligence and signals intelligence that (as noted) was circumstantial, at best.

The urban legend has it that we were pressured by the White House, and especially by Vice President Cheney, to write a case for war. I never experienced such pressure, and when I got to CIA and talked to those more directly involved, they reported that they felt no such pressure either. We just got it wrong.

As I was about to leave government, I mentioned to Leon Panetta, who was coming in at CIA, that I had read some of his commentary on Iraq while he was in private life. “Leon,” I said, “this wasn’t the White House. This was us. We just got it wrong. It was a clean swing and a miss.” The IC had lowballed its estimate of Iraqi WMD before the first Gulf War. Now we overcorrected.

NSA also had a role in Secretary Powell’s ill-fated call to arms in front
of the UN in February 2003. Three NSA intercepts related to Iraqi WMD were played during the speech. I had to clear their release (a formality) and make damn sure they were right and accurately translated (they were).

There was a fourth intercept that was in the stack of possible entries for the speech. This one was in some ways more explicit and detailed than the others. Playing it in New York would have been incredibly effective. But it would also have been incredibly misleading. Although it was highly suggestive of WMD use and while it was indeed from Iraq, it was not from the part of Iraq under government control and we saw no evidence of linkage to Baghdad.

I sent our best linguist to Langley to make the point. He did. There was no pushback. The cut was removed.

Later, as 2003 turned to 2004, it was very clear that we had gotten the WMD wrong. There were no weapons of mass destruction. We did a lot of soul searching. What do you do when you are wrong? Obviously you look to your tradecraft. We did.

But there was an even larger dimension. Shortly before Christmas 2003 an NSA historian developed a case study about another national decision that had been based on flawed intelligence: the Gulf of Tonkin Resolution, which launched a major US escalation in Southeast Asia. The case seemed to fit our current circumstance. Perhaps too well.

The historian e-mailed me a concern: although this was an “extraordinarily powerful and richly textured” episode, he began, “because of possible accusations of a ‘cover-up’ of intelligence errors, and what some might see as rough similarities to the current controversy over intelligence reporting related to the issue of WMD in Iraq, we want to make sure you believe it appropriate to use this particular case study this time.”

I responded the next day. “I don’t see a problem—sounds like a great way to use this historically rich material to address concerns that still exist today.” And so we did.

On August 2, 1964, North Vietnamese torpedo boats had attacked
the US destroyer
Maddox
in international waters as that ship was supporting raids on the North Vietnamese coast. SIGINT had warned of an attack, and the torpedo boats were repulsed. Score one for SIGINT.

Then, two nights later, in the midst of Johnson administration warnings to North Vietnam about further action, SIGINT misread North Vietnamese reporting on their continuing recovery operations from the first night as a second attack and issued a CRITIC (a kind of global warning). Subsequent US Navy evasive action and firing at some spurious radar hits were duly noted in North Vietnamese shore-based communications, which were in turn picked up by NSA and errantly catalogued as further evidence that a second attack was under way. President Johnson ordered air attacks against the torpedo boat bases, and Congress delivered the Gulf of Tonkin Resolution, authorizing the president to take “all necessary steps” to halt Communist aggression in Southeast Asia.

In the aftermath of the resolution, NSA stuck to its story that a second attack had occurred. It’s unclear if the agency’s subsequent investigation was careless, misguided, or just consciously ignored evidence. But it is clear that the August 4 reporting was wrong.

Tonkin and Iraq’s WMD were sobering lessons.

 • • • 

S
URPRISINGLY, AS DIRECTOR OF
NSA
,
I received no formal guidance to prepare for war with Iraq. But it didn’t take a rocket scientist to know that we were going to war, so we moved around what we later estimated to be about $400 million to get ready for the conflict. The deputy head of our SIGINT Directorate, a Gulf War veteran, directed the team we set up to prepare for war to read the agency’s history of the first Gulf War. He then conducted what the army calls a “rock drill” to synchronize what various tactical and national SIGINT units would do to support one another.

Linguists were a perennial challenge, even before the demands of this war. A year earlier, in 2000, with al-Qaeda strengthening and with
information on their plans sparse, CIA had outfitted an early model of the Predator with electro-optical and infrared cameras to hunt al-Qaeda in Afghanistan. The agency was looking for bin Laden, his close associates, training camps, and any evidence of weapons of mass destruction. It was tough getting this under way; a whole infrastructure of people and equipment had to be built to handle the streaming video from the UAV (unmanned aerial vehicle).

Those early missions did not carry SIGINT packages, but that didn’t let NSA entirely off the hook. As weak as Afghan air defenses were, we still didn’t want the political embarrassment of losing a drone, so we could not simply ignore the Taliban’s limited radar coverage and its handful of MiG-21s. Any Afghan detection of the Predator was deemed sufficient to scrub the mission. So we couldn’t fly without monitoring the air defense network, and that required linguists. Actually, we were so limited in Pashto speakers, it required
the
linguist, who would be dedicated to this task whenever the Predator was in Afghan airspace.

The Afghans finally did attempt an intercept, but the Predator was so small and so slow that the unlucky MiG-21 pilot reported that he could not find it—a conversation duly reported by NSA. In truth, the actual intercept was a little oblique. Something along the lines of the “birds spotted a pigeon but lost it.” A good object lesson for anyone who thinks it’s OK to “talk around” classified data on the phone.

Great work, but we were going to have to get a whole lot better to support full-scale conflict or rather, as it turned out, several conflicts: Afghanistan, al-Qaeda, soon-to-be Iraq. We began to pull Arab linguists off mission in order to sharpen them in the Iraqi dialect. All accounts, except for CT, suffered.

There was no magic lever to pull here. Good language analysts are a rare and special breed, since they have to be masters of so many skills.

First and foremost, they need to know the language and the culture. They have to be target smart: Who is related to whom? What is the overall context? What code words do they use? Who lies to whom on the phone? Finally, they’ve got to know the SIGINT system. Intercepts
don’t come to you on a platter. You have to be a hunter far more than a gatherer.

In ways not widely understood—taking a communication not in your language, between two individuals not of your culture and perhaps not quite able to distinguish the world as it is from the world as they would like it to be (common in messianic groups like al-Qaeda)—taking all of that and turning it into something actionable is high art and science.

In one instance, analysts were puzzled by the continued references to a specific household article from a known terrorist group. One linguist decided to stay at work for a weekend to listen intently to all the available cuts. Analysts usually scan cuts to get to the important ones; otherwise they would never get through the queue. But by giving up a weekend, this analyst listened to everything and gathered references to this household article and a variety of colored accessories. By Sunday night he had it. The household article symbolized passports. The accessories were visas. Their colors signified specific countries. Breakthrough.

The last time NSA intercepted bin Laden was in 1997, but that wasn’t the last time it heard him. In 2000, one AQ intimate of bin Laden’s was calling another and using a harsh tone, almost barking orders, in a way that seemed very inappropriate to the language analyst. The cut was puzzling. Why the odd tone? When the analyst finally turned up the volume to a very high level, there, identifiable in the background, was the voice of bin Laden himself telling the first communicant what to say.

Only a few days after 9/11, Russia passed us a warning about al-Qaeda possibly using “the big one” based on an intercept of a communication between two terrorists. We immediately thought nuclear and hunted throughout our system to find a copy of the cut. We succeeded, and luckily it was of very high quality. With George Tenet demanding answers fast, I put our very best linguist on it. Within thirty minutes he reported that the targets were certainly players in the terrorist network, but the question was whether or not they were talking about a nuclear attack on us or whether they were talking about our taking dramatic action against
them. It was indeed speculation about our response to them. False alarm, this time.

So language skills really mattered. I upped the required standard for our military linguists. In an earlier world, where we were intercepting heavily formatted military messages, basic fluency would suffice. A ground controller directing a fighter to turn left or turn right or to ascend to a certain altitude was pretty easy to follow. Al-Qaeda didn’t talk that way. They were elliptical, metaphorical, indirect, nuanced . . . and clever. We needed better skills.