Kill Chain: The Rise of the High-Tech Assassins (23 page)

Technology, whether in the form of signals intelligence or pictures, was always central to JSOC, whether in Afghanistan or Iraq. Artful jockeying of high-level connections back in Washington ensured McChrystal a disproportionate share of technical resources: at one point the entire non-JSOC U.S. force in Iraq had just one Predator drone for all purposes. The forceful Irishman, Michael Flynn, McChrystal’s intelligence chief, elevated this hoarding of resources to a matter of doctrine, claiming that “Intelligence, surveillance, and reconnaissance [predominantly drones] are most effective against low-contrast enemies (i.e. people) when
massed.
… It is not enough to have several eyes on a target—several eyes are needed on a target
for a long period
.” By these statements he meant that he needed three Predator drones watching a target 24 hours a day, 7 days a week. Given that 168 support staffers were required to keep one Predator 24-hour Combat Air Patrol in the air, this was clearly an expensive undertaking.

A high-ranking British visitor to JSOC’s Balad operation commented that it smacked of “industrial counterterrorism.” He did not mean it as a compliment, but many in the system took it as one. Some spoke approvingly of the “machine.” McChrystal himself could wax lyrical about his creation. Reminiscing years later about happy days at Balad, he described his impressions thus:

… as night fell, the operations center hummed with serious, focused activity. Soon, the rumble of helicopters and aircraft, some throaty, some a high whine, bounced across the darkened gravel and off the cement walls and barriers of our compound. The sound grew in layers, building like a chorus singing a round, as one set of rotors, propellers, or jet engines came alive, joined the cacophony, and then departed the airfield. Gradually, the chorus dissipated until silence returned to the darkened base.

The entire operation was very self-contained and secretive, with little news seeping into the outside world of what was going on apart from discreet references by privileged insiders. Even other components of the occupation regime were largely left in ignorance; McChrystal communicated with the regular forces only at the highest level. A “flimsy” (a printed message on a secure fax) would arrive each morning in the Baghdad military headquarters’ SCIF (Secure Compartmented Information Facility), the repository for especially secret material detailing the previous night’s JSOC raids. The paper had to be destroyed in the classified shredder by noon at the latest. “It was bad stuff,” said one former inmate of the SCIF who made a point of perusing these short-lived documents. “They were really running riot, shooting up rooms-full of people, massacring families, night after night after night.”

Such mayhem denoted what McChrystal later described as an artful shift in strategy. Despite the resources directed against him, Zarqawi had survived and expanded his operations, helping to kill hundreds of Shia in suicide bomb attacks and most dramatically blowing up the much-venerated Al-Askar Shi’ite shrine in Samarra in February 2006. The revised JSOC strategy, according to McChrystal, was to “disembowel the organization by targeting its midlevel commanders. They ran AQI day to day and retained the institutional wisdom for operations. By hollowing out its midsection, we believed we could get the organization to collapse on itself.”

Such an approach indicated the influence of social network analysis, a fast-growing discipline in the world of counterterrorism in which esoteric algorithms were deployed to probe the structure and dynamics of enemy organizations. A leading pioneer had been the mathematician and social scientist Valdis Krebs, who deployed such analysis on the 9/11 hijackers’ relationships with each other to demonstrate, by using elaborate diagrams, that their conspiracy was undetected because they adopted a low profile and kept to themselves. Central to this approach was the focus on what was called relational analysis—the links between different “nodes” rather than “attributes”—meaning who or what these nodes actually were (so heaven help a pizza-delivery store owner getting a lot of calls from a terrorist cell). Thanks to such studies, the business of assassination, or targeted killing, could move beyond a crude fixation with killing enemy leaders to more elaborate scenarios for “shaping” the enemy network by killing carefully selected individuals whose elimination would make the entire structure more fragile and thus easier to disrupt. This theoretical approach was becoming ever more fashionable, spreading into every nook and cranny of the national security apparatus. A classified study commissioned by the Pentagon’s Strategic Command in 2008 found that there were no less than “185 separate Attack the Network efforts across the military that are not consolidated, centralized, or coordinated.” The study’s authors referred to this structure as “ad-hocracy.”

Following such an operation, the social network charts, based on the intelligence monitoring of the network’s phone links, showed the disappearance of such links, indicating that the network had been disrupted. But the vanished links might have been equally likely indications that survivors had sensibly concluded that they should stay away from the phone and find some other way to get in touch. The network had not fragmented, even though it might have looked as if it had on intelligence diagrams of the network, which of course showed only those links known to intelligence. As Keith Dear, a Royal Air Force intelligence officer formerly serving in Iraq and Afghanistan, has acutely pointed out: “Targeted killing is often justified by the display of a social network chart before and after a targeted killing in order to explain how the group fragmented.” But, he explained, the charts ignored the fact that the group was probably using other ways to communicate. “The illusion that they fragment is based on the acceptance of the abstraction [of the chart] as reality.”

While the happy operation at Balad was doing its work, another campaign was under way to promote the notion that the United States could turn the tide of the war by adopting COIN, a doctrine of counterinsurgency that emphasized the cultivation of popular support as an essential tool. David Petraeus, the ambitious officer who parlayed COIN as a means to a rapid ascent through the ranks, succeeded in enshrining its precepts into an official U.S. Army Field Manual, FM3-24, published to rapturous public acclaim in December 2006. In the section devoted to “targeting,” intelligence analysts are required to identify “targets to isolate from the population, and targets to eliminate … the targeting board produces a prioritized list of targets and a recommended course of action appropriate with each.”

McChrystal’s shift to targeting midlevel commanders would appear to have rested on this sort of carefully considered approach. However, as a former Pentagon analyst with an institutional memory stretching back to the days of the Phoenix program observed to me with some amusement, “You could suggest any set of targets and say their loss would collapse the organization—low level, middle level, top level, it can all be made to seem equally valid. In the end it always comes down to this: the poor sap with the most links gets iced!” Seeking to verify such a cynical conclusion, I asked a JSOC veteran who had worked closely with McChrystal in Iraq if there had indeed been a thought-out plan as to whom to target, with careful consideration of how that would affect the enemy network. “No,” he replied after pondering the matter for a few seconds, “it was all kind of ad hoc.” Fundamentally, McChrystal’s campaign was following the same trajectory of previous “critical node” campaigns stretching back to the strategic bombing of Germany in World War II.

Ultimately, Zarqawi was run to earth and killed, though largely thanks to old-fashioned human intelligence rather than elaborate technology. His isolated safe house, located thanks to a tip-off, was hit with two precision-guided 500-pound bombs, shortly after which he expired. At the subsequent press briefing the military displayed a twice life-size matte photo-portrait of the dead jihadi in a large gilt frame that reminded some who viewed it of a hunting trophy. President Bush, who had promoted McChrystal to three-star rank in February, called with congratulations. The
New York Post
headlined “Gotcha!” and
Newsweek
, in its cover story, speculated that Zarqawi’s demise might be a “turning point in the long, frustrating war on terror.”

“Things changed when we got Zarqawi,” the former Pentagon-based specialist in high-value targeting told me. “Morale was getting a little low, at least in the military, up to that point. There was a kind of fatigue setting in—I remember people were saying ‘It’s always failing, maybe it’s not worth it.’ After all, we’d had fifty HVTs on the Iraq blacklist in 2003 and hadn’t killed a single one of them. We hadn’t gotten Osama bin Laden, or Mullah Omar. So Zarqawi was the first really high-value guy we got, and we had several successes shortly after that. Zarqawi—that was when it changed.”

A week after their leader’s death, al-Qaeda in Iraq named his successor, an Egyptian with an impressive jihadi record named Abu Ayyub al-Masri.

Zarqawi had been dangerous. Al-Masri was worse.

Whereas Zarqawi, as a former associate told American interrogators, had the prime goal of fighting for the Sunnis in Iraq, al-Masri saw Iraq as only part of a wider war against the West. Al-Masri repaired relations with al-Qaeda’s distant senior leadership in Pakistan, boosted his group’s revenues from various criminal enterprises, and cracked down on careless cell-phone use. He insisted on truthful reports from subordinates and improved the group’s digital operations. Suicide bombers were put to work editing and uploading propaganda videos while they waited to carry out their terminal mission. IED attacks and American casualties went into a steep upward curve.

The U.S. military did not mount a propaganda operation to raise al-Masri’s profile.

9

KILLING EFFECTS

In the summer of 2005, a small group of Americans gathered in the Jordanian desert to test a revolutionary weapon. If it worked, they expected to change the shape of the ongoing war next door. Among the testing party was Rex Rivolo, still working as a senior analyst with the Institute for Defense Analysis.

The weapon was an airplane, a very small, cheap, simple plane designed to fly low and slow. Built by a Jordanian company, the $200,000 Seeker looked like a helicopter but with high wings and a pusher propeller behind the fuselage so that the view from the cockpit was wide and unobstructed. It could stay in the air for as long as 7 hours and 15 minutes, land on a narrow road or the open desert, and refuel with regular 87-octane gas from any gas pump. Its camera could transmit infrared and daylight color imagery direct to the SUV that served as its ground station. The pictures could be screened on an ordinary PC. It carried a pilot and an observer, and required a maintenance crew of just one. With an enhanced camera and other accoutrements the entire system still came to no more than $850,000, less than a third of the price of a Predator (which required a support staff of 168) and a tiny fraction of the bill for an air force jet fighter.

Underlying the simplicity was a subtle concept. As weapons systems became more complex and expensive, they were bought in progressively fewer numbers. Because such systems were costly and scarce, their control tended to be pushed ever higher up the military hierarchy. But of course at that level control is exercised through mechanisms that are themselves complex and expensive. Think General Franks and his link to the trailer in the CIA parking lot, which was in turn linked via satellite to the Predator, which he could fire only after asking his lawyer, the secretary of defense, and the president. Meanwhile, U.S. soldiers, especially road convoys, were being shredded by a proliferation of roadside bombs. Rivolo and his teammates, led by a former navy F-18 pilot named Dan Moore, believed that an aircraft that could be bought in quantity and efficiently watch over a convoy or whatever else required attention would make a profound difference.

Their experiment, in which Jordanian Special Forces “ambushed” road convoys protected by the Seeker, was a great success. Time and again, the attacks were thwarted by warnings easily communicated from the plane to the ground commander. The aircrew, as the test report noted afterward, could scan an entire area with the naked eye (using night-vision goggles when it was dark) and then use the sensor to focus on objects of particular interest. A drone, as the report noted, “would require systematic search of the same wide area through a small aperture (the soda straw), which would make the mission much more difficult or impossible to conduct from a UAV ground station.”

The exercise was a triumphant endorsement for putting human eyes and brains close to the battlefield rather than filtering information through layers of imperfect video, fallible communications, and command bureaucracies. Accordingly, the team concluded their report by recommending that the plane be put into service in Iraq and Afghanistan. They printed multiple copies of the report and prepared to distribute them around the Pentagon. The response was swift. Orders came down that all copies were to be collected by a senior official in IDA and destroyed immediately. Higher authority evidently did not want publicity for a successful demonstration of a cheap and effective counterinsurgency weapon.