Kill Chain: The Rise of the High-Tech Assassins

by Andrew Cockburn

  Since the IED networks were not controlled by some central authority, Rivolo wondered how it was that they all seemed to operate in much the same manner, sharing the same tactics and techniques. When new techniques appeared in one area, they appeared to spread, sometimes overnight, to other regions and even to groups on the other side of the sectarian divide. Eventually, he believed he found the answer in stigmergy, a theory originally developed by French biologist Jean Pierre Grasse to explain how it was that insects that were apparently working alone nevertheless worked—Grasse used termite nests as an example—in a manner that appeared to be coordinated.

  Stigmergic systems use simple environmental signals to coordinate actions of independent agents (each with their own decision-making process). Termites building a mound, for example, leave a chemical trace on each piece of mud they add to the mound that is attractive to other termites, who add their own attractive piece of mud on top of it. Rivolo thought that independent insurgent groups were somehow communicating in the same way, their efforts appearing coordinated, and therefore inducing the U.S. to believe it was facing some sort of unified operation with a leader who could be targeted.

  Perhaps unsurprisingly, this sophisticated theory failed to gain traction when Rivolo proposed it on his return to Washington. His observations on the benefits, or lack thereof, of high-value targeting, did receive a receptive hearing from IDA Director Dennis Blair, a blunt-speaking former four-star admiral. Blair found Rivolo’s arguments on this score “completely convincing” and tried to interest others in the military high command but without success. As he afterward opaquely explained to Rivolo, there were “too many layers of command” to effect any change in policy. Rivolo and Hickey did better with a simple suggestion that Iraqi cell-phone companies put a random delay of a few seconds on the time it took a call to go through. This would automatically end the use of phones as detonators, since the triggermen could not time the blast on a moving target. This idea, which was adopted, was a lot cheaper than $100,000 Warlocks, but the counter-IED business was already too valuable for alternative, and simpler, solutions. In January 2006, proclaiming the defeat of the homemade bomb to be a major national priority, Congress approved the creation of a whole new bureaucracy, the Joint IED Defeat Organization (JIEDDO). Adopting a mission statement of “Attack the Network, Defeat the Device, Train the Force,” the new bureaucracy rapidly swelled to more than 3,000 people and an annual budget that equally rapidly climbed to over $4 billion.

  Rivolo never thought much of JIEDDO, deriding it as a flaccid bureaucracy serving the interests of the contractors who feasted at its brimming trough, but in a roundabout way, it brought him back to Iraq. In December 2006, General Ray Odierno was appointed multinational corps commander–Iraq to command the day-to-day war. As such, he would inherit a large staff, including intelligence. But like many other leaders, military and civilian, he saw the need for a personal intelligence team headed by someone he could trust to be his “eyes and ears” on what was really going on. So he picked Jim Hickey, an old friend who had served under him in other commands. Hickey, in turn, recruited Rivolo as chief analyst.

  The essential point of such a team was that it should be outside the formal bureaucracy, operating as much as possible under the radar. Hickey therefore needed a cover, an excuse for his handpicked team of a hundred people to be in Odierno’s headquarters. So he approached the JIEDDO director, a grandiloquent retired general named Montgomery Meigs, and proposed that Meigs authorize and pay for an intelligence unit at the Baghdad headquarters that would ostensibly be working for JIEDDO. He stipulated that Meigs could not attempt to exercise any control or interfere with its operations, but could claim credit for any success it might enjoy. Meigs agreed.

  Installed at Camp Liberty, the vast U.S. military headquarters on the outskirts of Baghdad, Rivolo had total access to all information, classified or otherwise, in the Iraqi theater of operations, and he settled down to learn all he could about the IED war. There was plenty of information on hand. Every significant incident, bombing, firefight, accident—all inevitably jargonized to SIGACTS (the after-action reports that units were required to file on all incidents)—was assiduously noted and entered into the record. But as had been his reaction when introduced to the DEA, Rivolo was not impressed with the army’s staff work, especially when he discovered that their ignorance of basic statistics left them incapable of making the simple calculations necessary to account for “statistical fluctuations” in the reports of bomb attacks. “Odierno would get a chart with bullets that said ‘IED activity today is up 37 percent.’ Well, shit, that looks pretty bad. The following day they would say ‘oh we had a great day, IED activity was down 90 percent.’ True, but irrelevant.”

  Rivolo bluntly informed them that there is a standard statistical formula for eliminating fluctuation to reveal an underlying trend and that the erratic numbers they were solemnly briefing meant nothing. In truth, the underlying trend was bleak. Shorn of scatter, the charts of bomb attacks and the resulting killed and wounded followed an undulating but inexorably upward slope. There had indeed been a decline in casualties in the spring of 2005, the clear result of a belated decision to “up-armor” the Humvees that soldiers had previously tried to protect with scrap metal and sandbags encased in plywood. But the ominous trend soon resumed, ticking up in a saw-toothed pattern, month after month, peaking at Ramadan in the fall, then dropping sharply as the Iraqi weather turned unpleasant before starting another upward climb in the spring.

  “Apart from when we first put armor on the vehicles,” recalled Rivolo, “nothing else we were doing was making much difference.” This was certainly not for want of effort by the burgeoning anti-IED military-industrial complex, now lavishly nurtured by JIEDDO’s multibillion-dollar budget, as well as a host of other centers and task forces across the military bureaucracy. Multiple initiatives funded by the organization included the training of bees to detect homemade bombs (soldiers had to watch screens with magnified images to see if the insects stuck their tongues out when close to explosives), sniffer dogs, and “Fido,” a $25,000 “molecular sniffer” designed to duplicate canine detection capabilities.

  Apart from such eccentricities, more substantial efforts were being introduced in the form of vastly heavier armored vehicles known as MRAPs. In his memoir of his time as secretary of defense, Robert Gates writes movingly of the death and suffering inflicted on troops by inadequate armor, and his pride at having forced a crash MRAP program on a reluctant military. However, according to a former senior staff officer at army headquarters in Baghdad, the initial requirement that MRAPs be built to withstand an explosion of several hundred pounds was progressively and quietly scaled back by developers to a more modest thirty-five pounds. Since many insurgent bombs were far more powerful than that, these massive and costly vehicles in fact made little practical difference to the number of dead and wounded GIs coming home from the wars. An exhaustive analysis of the number of killed and wounded per successful attack on the commonest type of MRAP compared to those suffered by Humvees equipped with additional armor in 2007 revealed that while the Humvee suffered 2.4 killed and wounded, the equivalent number for the MRAP was 2.3. Nevertheless, no less than 27,000 of these massive vehicles would be built and fielded for use in Iraq and Afghanistan. The total cost came to $40 billion.

  Meanwhile, jostling for airspace above the bomb-pitted highways of Iraq were growing flocks of ISR, the acronym for “intelligence, surveillance, and reconnaissance” coined in the 1990s by Admiral William Owens, the apostle of net-centricity. Drones, U-2 spy planes (some operated by the air force, others by a specialized army unit), Task Force ODIN (“observe, detect, identify, and neutralize”), an operation dedicated to tracking down the deadly IEDs and the groups behind them—all were thrown into the mission. JIEDDO even embarked on an effort to deploy satellites on the fringes of space, once dedicated to seeking our Soviet ICBMs with thermonuclear warheads, to search for men digging a hole in the road.


  Occupying space, the surveillance industry was also moving into time. Adopting an initiative originally conceived to track loose nuclear weapons, the military poured money into “change detection.” The concept was simple: if you had constant surveillance of every conceivable location where a bomb might go off, then, following an explosion, you merely had to rewind the tape to see who buried the bomb and follow the tape back in time to retrace the offender’s steps to wherever he picked it up. The possibilities were potentially limitless; sufficiently extensive surveillance would reveal the bomb maker’s lair, the origins of his materials, the home base of his financiers, and ultimately the location of his leaders. In a story called “On Rigor in Science,” the Argentine writer Jorge Luis Borges conceived of an empire “in which the cartographers’ guilds struck a map of the empire that was the exact same size of the empire and that coincided point for point with it.” In effect, this was the goal that the proponents of change detection, or “ground movement target indicators” as the concept was also known, were seeking to achieve by watching and recording ever-greater areas in hopes of providing the answer to everything. Year after year, new programs to achieve this goal were unveiled, though on close examination they usually turned out to be the same concept but with a different name. The goal may have been impossible, but the attempts were very profitable.

  The army’s initiative, Constant Hawk, billed $84 million for 2007, while the air force’s offering, Angel Fire, garnered $55 million. Meanwhile the air force research laboratory in Dayton, Ohio, was hard at work on developing the Gotcha radar system. Mounted on a drone that resembled a miniature B-2 bomber, this was billed as capable of remaining aloft for 24 hours or more, beaming back a constant stream of radar, video, and infrared images of an area 20 kilometers wide to a specially designed $2.2 million supercomputer, all in hopes of catching the bomb layer at work. “The capability to scan an entire city from the air in any weather and detect any movement of any object as small as a cockroach is a goal of the Gotcha radar program at the Air Force Research Laboratory,” ran an air force promotional release.

  Further inquiry reveals that Gotcha was in fact a variation on our old friend JSTARS, once conceived to observe and target advancing armored Soviet hordes. The dream of seeing everything, targeting everything, hitting anything had, with “rewinding the tapes,” now extended into four dimensions.

  There was much in this enormous and expensive effort to impress outsiders. Defense Secretary Gates, for example, was beguiled by Task Force ODIN’s videos. “It was amazing,” he recalled in his memoir, “to watch a video in real time of an insurgent planting an IED, or view a video analysis tracing an insurgent pickup truck from the bomb-making site to the site of an attack. It was even more amazing—and gratifying—to watch the IED bomber and the pickup truck be quickly destroyed as a result of this unprecedented integration of sensors and shooters.”

  Reality was less gratifying. In May 2007, some months after arriving in Baghdad, Hickey enlisted Odierno to sponsor a meeting at which the units that were deploying various technologies could present their results and make their case. When it came to their turn, Task Force ODIN displayed Constant Hawk videos of the type that so impressed Gates. Then Rivolo asked an awkward question: How many bomb planters had they actually tracked back to their point of origin? The task force representative came clean: one. Other vaunted technologies emerged with even less credit. Compass Call Nova, for example, a major air force contribution to the anti-IED effort, was cruising the skies at $34,000 per hour, year in and year out. Rivolo had made careful study of its performance by correlating specific missions flown with what happened on the ground below. The results were painfully clear: in the period he examined, between October 2006 and May 2007, hundreds of bombs had gone off or remained quiescent completely regardless of the $100 million planes overhead pulsing out their electronic commands. As he tersely informed Odierno, Compass Call Nova had “no detectable effect.” However, when queried at the meeting as to why they continued to fly these pointless and expensive ($100 million per year) missions, the air force representative replied, “[T]he field commanders want it.” (Rivolo later discovered that an inquisitive researcher at the Center for Naval Analysis had done a similar study a year before and had come to the same conclusions, which were duly ignored by high authority.)

  There was, however, no mistaking who controlled the principal strategy for defeating the IED. It was evident even in the furniture. Most of the equipment around the base was “pretty beat up,” according to one of the tens of thousands of Americans who passed through its heavily guarded gates during the years of the occupation. The office equipment, furniture, and electronics were worn, much of it old, having been hurriedly shipped in from other bases around the world as the occupation army settled in for the long haul. One single-story building close by Odierno’s office was different, though. “You walked in there, and everything was brand new,” remembers a visitor. “They had the most up-to-date computers, big servers, nice big plasma screens in full working order on the walls. You could tell it had priority.”

  This was the home of the High-Value Targeting Cell, a counterpart to the cell in the Pentagon that had tracked Saddam and, before him, Milošević. Here, analysts tracked individuals on NSA “manhunt” lists while others combed through reports from units around Iraq for sightings of the wanted men or others to be added to the Joint Priorities Effects List, the JPEL, a constantly refreshed master list of targets slated for elimination. All around Iraq and Afghanistan, every unit maintained its own list of high-value targets, also known as high-value individuals, each death or capture dutifully recorded in the SIGACTS. For example, a May 5, 2006, report on the shooting of Allah Harboni, spotted from a U.S. army observation post and shot as he tried to escape by car, was summarized as: “AT 0810C, A 3-187 OP ENGAGED AND KILLED ONE OF THEIR BN [Battalion] HVI’S IN THE SALAH AD DIN PROVINCE IN SAMARRA VIC 38SLC9652784715.”

  Meanwhile, on the expansive grounds of Camp Slayer, yet another of the sprawling bases in Baghdad housing the military occupation regime, stood an ornate, domed building known as the Perfume Palace. Originally constructed for the pleasure of the notorious Uday, elder son of Saddam, it had escaped the “shock and awe” bombing and was now occupied by U.S. military intelligence. Jerry-built, with questionable drains, it was not a popular place to work: the chief of military intelligence in Iraq at one point decreed a weekly Be Happy Day as an initiative to raise morale. Although many of the inmates were servicemen and women, civilian contract employees occupied several floors. They were there thanks to JIEDDO, which was sluicing money into Attack the Network along with the rest of its mission statement. Most of them were retired U.S. policemen, retained by defense contractor SAIC to analyze intelligence dossiers in search of high-value targets. All commanded high salaries, as much as $300,000 a year, though their employers charged JIEDDO far more. Target hunting had become a profitable business. (Asked if their nominees ever found their way onto an actual target list, a JSOC officer bluntly answered, “No.”)

  For the HVT industry, the benefits of hunting down leaders of the “IED networks” appeared self-evident, as had assassinating Hitler, Patrice Lumumba, or Pablo Escobar in years gone by. Since the elimination of formerly critical nodes such as Saddam and Zarqawi had paid little dividend, the target list was expanding, as such lists always do. Toward the end of 2007, Rivolo, already dubious about the presumptions behind the strategy, began to look for data that could reveal whether or not the strategy worked. He found it in the SIGACTS.

  With full access to the SIGACT database, Rivolo extracted the records of 200 cases in which high-value targets had been killed or captured between June and October 2007. Then he went through the records again to see what happened in the neighborhood where each leader had operated. This was the crucial question. Had his elimination made a difference in the fight against the insurgents? Rivolo counted the number of IED attacks against Americans in the 30 days following e
ach high-value target death or arrest within a given distance from the event and compared it to the number in the 30 days before the death or arrest as a percentage of change. Repeating this procedure for different distances, Rivolo plotted the results on one axis of a graph and the distance on the other. When complete, the graph delivered a simple, unequivocal message: the strategy was indeed making a difference but not the one intended. Hitting HVIs did not reduce attacks and save American lives. It increased them. Each killing had quickly prompted mayhem. Within 3 kilometers of the target’s base of operation, attacks over the following 30 days shot up by 40 percent. Within a radius of 5 kilometers, a typical area of operations for an insurgent cell, they were still up 20 percent. Summarizing his findings for Odierno, Rivolo added an emphatic punch line: “Conclusion: HVI Strategy, our principal strategy in Iraq, is counter-productive and needs to be re-evaluated.”

  How could the removal from the scene of ringleaders of attacks on Americans generate such a counterintuitive result? Just as the field officers had told Hickey and Rivolo during their 2005 trip, dead leaders were invariably replaced quickly, “usually in twenty-four hours, always in forty-eight,” recalls Rivolo. For a variety of reasons, new commanders were almost always eager to press the fight harder. Often, they would be relatives of the dead man and hot for revenge. In addition, having just succeeded to the command, they would feel the need to prove themselves, especially if the late leader’s martial energies had been faltering due to battle fatigue or other interests, highlighting the need for a new broom. Always, they were more deadly.

  A week or so after submitting his findings, Rivolo asked Odierno if he had read the study. “Yeah,” replied the powerful commander shortly, “there’s a limit to what I can do.” Bureaucratic politics, it seemed, superseded empirical truth. Odierno’s reliance on Hickey’s operation to tell him what was going on rather than the elaborately staffed formal apparatus was ruffling feathers. “Hickey was going directly to Odierno every day and Odierno was just ignoring the other people and they knew that and they weren’t happy about that.” Returning to Washington in February 2008, Rivolo presented his conclusions on the strategy to his superiors at IDA. Unfortunately, IDA Director Dennis Blair was gone, having been replaced by former air force chief of staff Larry Welch, who appeared disinclined to challenge established doctrine; there were by now too many vested interests involved in targeted killing. Some thought it didn’t matter anyway. “When you mow the grass,” the senior counterterrorism official who had drawn my attention to Israeli influence on the strategy remarked offhandedly to me, “you don’t expect the grass not to grow again.”