The colonel was just back from Iraq when we met with him in the fall of 2005. He spoke in the blunt way of a soldier who had served 25 years in elite secret units. He had been in plenty of precarious situations and had the battle scars to show for it. The special operations unit he commanded knew how to surprise and kill unsuspecting enemies and had scored some successes, but the colonel was pessimistic about one thing: "No amount of training can prepare you to hunt down the bad guys after midnight when the intelligence you receive does not pinpoint their hideouts."

Other commanders we spoke with who had served in Iraq and Afghanistan had also experienced frustration in their mission--capturing or killing insurgent, terrorist, and militia leaders and key operatives, and taking out the factories where they make improvised explosive devices (IEDs). They echoed the colonel's message: "Actionable intelligence" was often absent from the U.S. war effort. A commander whose area of operations had been Tikrit told us, "Rather than intelligence on precise insurgent targets--'There is a unit of Jihadi fighters using the abandoned Bus Station on Tikriti Road as a base'--we would be told to 'Search for a Jihadi unit in a two-three block area around the Bus Station.'" The difference was between looking for a needle on a platter and a needle in a haystack; between a precise mission and an indefinite one requiring too much time on the ground in hostile territory.

The military men we talked to (all of whom, both active and former operatives, insisted on anonymity) all said the same thing: When we're spending $40 billion a year on intelligence and committing 150,000 men to the Iraqi front, why can't we create the actionable intelligence required to roll up the insurgents? As it worked out, this was exactly the question we'd been trying to answer over the previous year and a half. Our journey had taken us to three continents, where we met, sometimes more than once, with former intelligence, military, police, and domestic/internal security service leaders from democratic governments, and with former leaders of armed groups who were once deadly enemies. In cosmopolitan London, Washington, Tel Aviv, and other world capitals, and in more remote settings in Central America, Mexico, Africa, and Australia, we listened intently as these former practitioners discussed what to our surprise turned out to be a common set of measures--intelligence dominance was the summary term we came up with to describe the model--they had used to overcome bloody threats posed by armed groups. They convinced us that this is the wheel the United States must now reinvent if it is to win in Iraq and on other murky fronts of the war on terror.

Comprehensive Intelligence

Our meeting on March 22, 2004, with two former top-ranking Israeli intelligence officers had been scheduled for months. But it was not "business as usual" that day in Tel Aviv. Earlier in the morning, Sheikh Ahmed Yassin, the main leader of Hamas, had been killed by missiles fired from an Israeli helicopter hovering over the Gaza Strip, not that far from where we were sitting. The operation was one part of Israel's campaign to kill key active terrorist leaders to weaken the group so that it becomes less effective or negotiates an end to the armed struggle. Since the beginning of the Al-Aqsa intifada in 2000, the wheelchair-bound Hamas chief had ordered most of the suicide bombers who had killed Israelis in cafés, on buses, and in other public places. According to the Israelis, from the start of the second intifada through 2005, Palestinian terrorists killed 1,074 Israelis and wounded 7,520, astounding figures for a small country. The comparable losses for a country the size of the United States would be 50,000 dead and 300,000 wounded.

Right on time, the two men appeared in the lobby of the beachfront hotel where we were staying. After exchanging pleasantries at a quiet spot in the hotel's outdoor café, the soft-spoken ex-senior officer of the Shin Bet, Israel's renowned internal security agency, said adamantly, "To defeat terrorists you must know everything about them. Everything! Who are their leaders and how do they plan and carry out operations? How are they organized and what methods are used for recruitment? What are their weaknesses and vulnerabilities?" Drawing on 25 years of experience, including senior command positions, he insisted that without systematic profiles of the enemy, operations to neutralize such unconventional adversaries are usually futile.

The information for such profiles, added his associate, himself a retired general with 30 years in Aman, the intelligence division of the Israeli Defense Forces, requires local intelligence that is collected block by block, village by village. He had relied on this kind of intelligence throughout his career. The two counterterrorism experts then went on to describe the way the local intelligence units had come to be organized and staffed to identify armed group strengths, weaknesses, and vulnerabilities and to turn this information into operational opportunities. Over several hours, they laid out the details of how in a particular locality--a section of a city or rural area--the intelligence unit assigned to it attains comprehensive knowledge of the armed groups in that location. They explained how a profile of each group's ideological and organizational characteristics is produced and exploited.

The Israelis and others have learned over decades that intelligence dominance involves a major commitment of time, money, men, and patience. If you do it right, they explained, you will obtain the intelligence that enables you to control the territory vital to your security.

Some democracies--notably the United Kingdom and Israel--have mastered this approach through bloody trial and error, in the course of meeting the challenges posed by armed groups. Though each country tailored its techniques to the specifics of its geopolitical situation, the techniques they came up with are similar from country to country.

"We had to be in command of the local situation on the ground," said the soft-spoken local practitioner who met us at the airport in Belfast and drove us into the countryside a few weeks after our visit to Israel. He had served at the operational level his whole career, turning down every proffered promotion to a management position because he knew that the war against the IRA and other paramilitary groups had to be waged on the streets and in the alleyways. "We had to know what the IRA boys were doing, keep them on the defensive, always causing them to worry about our next move. You must collect comprehensive intelligence--complete block by block coverage--of each location out of which the terrorists operate."

This intelligence operator described an approach similar to the one the Israelis had told us about. It was perfectly logical; not to adopt such a strategy would, in fact, have been counterintuitive. But it was not, and is not, part of the official lexicon of the U.S. intelligence community, whose operations we have both studied for decades. (Perhaps there is one exception: This sort of approach briefly became part of the U.S. tool kit during the late 1960s in Vietnam, but it became a lesson lost when the North Vietnamese invaded and their tanks swept into Saigon in 1975.)

In talking with the intelligence practitioners on the island of Ireland, along the Mediterranean, and elsewhere, we tried to explain the challenge of changing the U.S. intelligence community's conception of its mission, as well as its approaches to collection, analysis, and counterintelligence, all of which are deeply rooted in World War II and the Cold War experience. This enduring organizational culture, we said, with its focus on threats posed by states, heavy reliance on technology, and relative dearth of case officers who do local intelligence work, does not provide the kinds of capabilities they employed to deal with the armed group challenges. "Academic nonsense!" said one of these former senior officers during our discussion. "The United States needs to get serious with what you call 'intelligence dominance' in Iraq, or suffer the strategic consequences."

Organizing for Intelligence Dominance

Putting intelligence dominance into practice to gain control of territory plagued by armed groups means utilizing all the tools in the intelligence toolbox--integrating collection, analysis, covert action, and counterintelligence instruments--to maximize effectiveness against targets. The operations that flow from intelligence dominance may involve targeted killings of terrorists, as they sometimes have for the Israelis, or the interdiction of arms and money and the denial of safe houses and the occupation of territory, as they did more often for the British fighting the IRA. But there is one common denominator: Those intelligence services of friendly governments that have become dominant started at the local level, working through various types of local intelligence units.

Penetrating an armed group's apparatus to learn of its capabilities and intentions requires human intelligence ("HUMINT," in the trade). Electronic surveillance alone won't do. It can help, but only local HUMINT will tell you what the enemy plans to do, and where and when he intends to do it. This was a mantra we heard repeatedly from those who had had success fighting on the front lines of this war of shadows in these and other democracies.

"Developing a comprehensive profile of your enemy is a building-block process," explained one former chief who was intimately involved in doing exactly this when he was coming up through the ranks of his service. "We did it street by street, village by village, beginning with basic intelligence collected at the local level."

Basic intelligence? It took some explaining before we fully grasped what this entailed. A former Shin Bet specialist who had worked with local units for more than two decades described it as "the big picture of daily life in your area of responsibility." Basic intelligence is information on all key political, social, and religious activists and leaders in a specific geographical locale. It spells out how they communicate and interact with one another and the surrounding population. It pinpoints all the major financial, political, social, and religious networks the armed group uses.

"Collecting it is labor intensive," this veteran added. "Up to 40 percent of your collection capabilities gather basic intelligence. And most of your sources are not recruited agents in the classical sense. You establish networks of local people you meet and interact with frequently." He was describing a situation in which an operative functions somewhat like the policeman on the beat--constantly talking to, interacting with, and keeping tabs on the people in his neighborhood and, most of all, keeping his eyes open for slight changes or new developments in the local scene.

The men we met who had battled the paramilitary groups in Northern Ireland emphasized that infrastructure intelligence was the next step after basic intelligence had been established. A former chief of intelligence for the second largest police Special Branch in the United Kingdom told us, "You begin with personnel. Who are the leaders, their strengths and weaknesses and, importantly, political and other differences that affect their cooperation and interaction? Who makes up the cell's rank and file, what skills do they possess, and how committed are they to the cause? What are the cell's physical capabilities--safe houses, weapons caches, bomb-making facilities, and so on? Once we had it, the big picture, we could really begin to think about what to do with the IRA groups operating in our territory."

Infrastructure intelligence is mainly collected through recruited agents, run by case officers whom the former Special Branch intelligence chief called "hybrids." When collecting basic intelligence, they followed standard operating procedures akin to those of law enforcement. But when the focus narrowed to the armed group, their procedures corresponded to those of standard intelligence tradecraft.

When human collectors are in place and the composition and nature of the armed group has been defined, our interlocutor from the Special Branch said, "You're now ready to go get them, to put their leaders out of commission and shut down their safe houses and their bomb factories--in today's term, IED factories." To do this, he said, you use target intelligence to pinpoint the movements and activities of key leaders and personnel in armed groups well before the point of attack. This information from recruited agents is supplemented by intercepts of various electronic communications, from cell phones to the Internet, and by photographic imagery. The result is continual live coverage of selected targets, augmented by basic and infrastructure intelligence.

One former operator gave us an example: "Back in the mid-1980s I was running a local intelligence unit covering a particularly dangerous part of Belfast. I learned from information collected in another part of the city that several IRA subcommanders were going to hold a big planning meeting for a bombing campaign at a safe house in my grid. We put all the ones we knew about under close observation but came up dry. None of them gave any indication that they were planning for a major activity.

"Then the break came from a small piece of local intelligence. One of the boys in my unit had gotten pretty close to a local grocery store owner. He would see the owner two, three times a week. At one of those get-togethers the store owner happened to mention that a Mrs. McCoy had come in several times over the last week with grocery lists many times larger than the size of her normal purchases. 'You'd think she was planning to feed a small army,' the grocery man said. We put her house under surveillance, tapped the phones, all the normal stuff. Turns out it was the IRA safe house we were looking for. You boys can fill in the details of the eventual outcome."

Opportunity Maximizers

As we talked with these professionals, we realized that intelligence and security services that achieve intelligence dominance do so by seamlessly joining analysts and operators at the local level. The U.S. approach, on the other hand, generally tends to maintain definite boundaries, often sharp divides, between operational and analytical divisions of the intelligence services.

Here is how one highly experienced analyst we interviewed at length--an Israeli woman--described this alternative arrangement. "Almost every piece of analysis I produced included an operational recommendation. I had to identify an operational objective, a target; it was part of my job description." She was trained to deduce from the intelligence data she handled operational opportunities that could subsequently be exploited by military, intelligence, and police forces.

We explained to her that this was very different from the way many U.S. intelligence analysts approached their job, which in the main built walls between analysts and operators in order to preserve the "objectivity" of the former. "Doesn't your approach taint the analyst?" we asked, "by drawing him in too close to operations?"

Shaking her head vigorously, she replied, "Intelligence analysts do not live in some pristine world." The kind of operationally focused assessments she produced, assessments that identified specific armed group targets, were generated by working at the local level--hand in hand, day in, day out--with human collectors and operators. Her role was to identify local collection gaps and task the local case officers to fill the gaps from their sources.

We asked for examples of how this worked on a daily basis with case officers. She talked about situations in which she had listened in on meetings taking place in the field between case officers in her unit and their recruited agents in "real time." "I could say to the case officer, 'Ask him about this . . . ' or 'Get him to clarify that . . . ' By being able to do so, I could get the specific details I needed to guide the police and military commanders."

She could request that a specific suspect be detained if, based on other sources, she knew he was potentially a valuable source for a critical piece of information. She could request the interrogation take place quickly, especially if the intelligence sought was perishable. She could provide a local interrogator with specific questions and detailed knowledge that he could use in interviewing the suspect.

Finally, she had access to local signals intelligence (SIGINT) that was collected by specialists she worked with. As with case officers and interrogators, she could point the SIGINT specialists to specific targets within the armed groups in her sector. She also could pass on requests to have national level SIGINT and imagery platforms focus on time-sensitive armed group targets.

All of this information was fused and rapidly assessed with one principal purpose in mind--identify specific targets in her sector to attack or co-opt. The action could be executed by her service or some other arm of the government. But the objective was always the same: to "seek out opportunities to hit the terrorists."

Collectively, the men and women sitting around the table at another meeting in Tel Aviv had vast experience targeting enemies in unfriendly occupied territory. They admitted that they had not always had intelligence dominance, and that they had had it and lost it before reestablishing it. After establishing intelligence dominance in the territories occupied after the 1967 war and in parts of Lebanon in the 1990s, as a result of the Oslo Accords and the withdrawal from Lebanon, they had to give up this advantage and withdraw from areas in the West Bank and Gaza. Then, when the second intifada erupted in late 2000 and Israeli casualties mounted, they were tasked with reestablishing their dominance. It was not easy to do so, but they did, and their intelligence successes contributed to the Palestinian Authority's gradual deemphasis of terrorist activity against the Israelis in favor of political and diplomatic initiatives, and even led Hamas to engage in a temporary cease-fire that held until recently.

For hours, they described the organizational structure required, and listening to them we realized that the intelligence-led struggle they described had relevance for the U.S. effort in Iraq and elsewhere. The first step was to divide the targeted territory--neighborhood, sector, even individual street--into grids. The next step was to assign to each grid an intelligence unit with responsibility for collecting basic, infrastructure, and target intelligence and turning it into operational assessments that could be used to weaken and undermine all armed groups active in that locale. One veteran described these local units as "the brains of the entire intelligence and security system."

There were several factors that contributed to a unit's success. First, each member received considerable professional training before going to the field. "You don't recruit kids to the service," said the one former intelligence chief, who retired at a rank equivalent to major general in the army. "We looked for men and women with successful work experience, who were in their late twenties." They either had already mastered spoken Arabic or passed a language proficiency test demonstrating their capacity to do so. Once selected, all recruits went through months of rigorous training. Only then were they considered ready for operations.

The same was true for the units' interrogators. Standards for selection were demanding. Knowledge of the adversary's language and culture was a given. Then, after a nine-month course in the methods of interrogation, stressing how to use knowledge and skill rather than violence, they went to a field unit to observe and learn from practitioners in action.

The same kind of professional preparation was mandated for the other members of the local intelligence unit. But the lines of communication and coordination within the units were as important as the professionalism of the members, these intelligence veterans repeatedly stressed. The commanders of the local units were taught that communications among them must be rapid and seamless. That was a second secret of success--short lines of communication between case officers, analysts, local SIGINT, and other members of the unit. The centrality of analysts in the local units was a third factor. "They identify the targets and focus the collectors to acquire the intelligence necessary to develop operational recommendations for how and when to move against them," said one former officer, who for over a decade coordinated several of these local intelligence units.

Another factor contributing to the success of local intelligence units was the length of time personnel assigned to them served. In most cases there was a commitment for several years. And not infrequently, it was extended. Moreover, one's next tour was often in a regional intelligence center with responsibility for coordinating the operational activities of several local units, including the one in which he or she had previously served.

The final condition for success was collaboration of the intelligence unit's commander with local police and military (combat) commanders. The two sides--intelligence and operations--had to develop a symbiotic relationship. The local unit had to develop close--very close--relationships with the local combat forces that had the capability and responsibility to go out and kill or capture the adversary if they knew where he was. The military and security forces had to receive and act on precise information in a timely way, usually within hours.

And, in turn, the combat commander had to use operations and boots on the ground to pass on quickly to the local intelligence units whatever information he came across in his unit's daily street patrols on foot. And this type of mutual support would continue up the line at the regional and national levels. This was another key to intelligence dominance.

The Plan

At the end of the 18-month tutorial we received from former senior intelligence and security officers, we thought we saw what the United States could do in places like Iraq. It was necessary to create an architecture of security based on intelligence dominance at the ground level. To do this would involve establishing physical control of terri tory and introducing intelligence operatives into areas within this territory who knew the language and culture and who were ready to stay on the ground for a prolonged period of time. Intelligence dominance would not "win" the war against the insurgents, but it could provide the means to win.

We presented the lessons we had learned, or the model, to senior officials in the Office of the Secretary of Defense, the Joint Chiefs of Staff, the National Security Council, and the intelligence community, as well as to the staff of the Senate Select Committee on Intelligence. Some then asked us to determine how intelligence dominance could specifically help achieve U.S. objectives in Iraq.

This past fall, we turned those details into an operational plan for establishing intelligence dominance in Iraq. A second round of briefings ensued. To put the plan into action, we proposed two things: first, a command decision to undertake a major pilot project to test this intelligence model in two regions; second, recognition that results will not come overnight. It will take months to select and train a hundred or more Americans in these techniques, and several more months for them to pass these methods on to Iraqis. Once that has been done, joint U.S.-Iraqi local intelligence units can be established in selected areas. While U.S. personnel can help direct the units, operations in the field will be executed by Iraqis, who can fit into the local setting.

After more than two years of protracted fighting against dispersed insurgent groups, it had become clear to at least some leaders that we had no alternative except to see if we could adapt the model for Iraq. And they arranged for other key officials to learn about it in an effort to forge a consensus. But innovation in Washington comes hard. A number of objections--showstoppers--rolled in.

Among them: the "not invented here" objection. This is a self-inflicted wound. It prevents the United States from drawing on the lessons and knowledge that friends and allies are standing at the ready, eager to share.

Next, we were told that the successful practices of others were just not relevant or adaptable to the situation in Iraq. In Iraq, Americans look different from our enemy, don't live next door, and don't know the language or culture. Contrary to myth, however, few British, Australian, or Israeli intelligence professionals looked like their adversaries or spoke their dialects or knew their tribal culture before training. Those who employed intelligence dominance did so by adapting it to diverse situations. There is little reason to believe the model cannot be adapted to Iraq, where Iraqis would implement it on the ground.

Then we heard that it would take too long to implement. "We are in the middle of a war and do not have the luxury of time and experimentation." This showstopper is equally illogical. While time may be running out, there is no available substitute. We know where more of the same will lead.

Finally, we were asked, Where would we get Americans or Iraqis who were willing to face the dangers inherent in local intelligence units? But every day, many Americans, Iraqis, and others are already out in the Sunni Triangle taking extraordinary risks. As we listened to these and other doubts, the words of highly successful foreign practitioners kept reverberating in our heads: "The United States needs to get serious in Iraq . . . or suffer the strategic consequences." They are exactly right. And until we do so, the dark alleyways of Ramadi, Falluja, Tikrit, and elsewhere will belong to the insurgents, and we will not prevail in this intelligence-led struggle.

Richard H. Shultz Jr. is director of international security studies at the Fletcher School of Law and Diplomacy, Tufts University, and director of research at the Consortium for the Study of Intelligence (CSI). Roy Godson is a professor of government at Georgetown University and president of the National Strategy Information Center. The research for this article was drawn from CSI's Armed Groups Project.

Next Page