Deep Cover

Wednesday, October 30, 2002

Individuals who can penetrate terrorist cells, set up clandestine logistics networks, and work in dangerous and hostile environments in the secret wars that lie ahead.

In the new wars against terrorist organizations and rogue states, the United States often needs effective “cover”—the ability to send military forces and intelligence officers overseas without our adversaries knowing who they are. Unfortunately, this has rarely been a particular strength for the United States, and the approaches we have used in the past are poorly suited for today.

Until we develop better cover, we will be unprepared for the threats that lie ahead. U.S. military forces need a clandestine infrastructure to support strikes, raids, and hostage rescue operations. Intelligence organizations need effective cover to penetrate terrorist cells and international criminal organizations and—even more important—to gain access to foreign communications.

The Traditional Approach

Traditionally, U.S. military and intelligence organizations have relied mainly on “official” cover. That is, most personnel under cover openly acknowledged they were U.S. government employees. The only thing that was secret was their link to the U.S. armed forces or intelligence.

This practice of relying mainly on official cover has strong historic roots in the intelligence community. For example the spiritual father of the CIA’s clandestine services, Allen Dulles, ran the U.S. station in Bern for the Office of Strategic Services under official cover during World War II.

Official cover was effective during much of the twentieth century because most intelligence targets were in places where you could make a plausible case for having U.S. officials present. In the case of Allen Dulles, for example, Switzerland was neutral and lay astride the diplomatic and financial crossroads of Europe. The record shows that Dulles usually did not have to seek out the German generals, Italian colonels, or Austrian businessmen who provided him information. They came to him. Most of Dulles’s agents were walk-ins—people who knew (or suspected) who Dulles really worked for and volunteered their services.

Official cover worked for U.S. intelligence during the Cold War, too. Its main targets were Soviet bloc officials or officials of Third World countries that were players in the Cold War. The would-be recruits were diplomats, military officers, or members of foreign intelligence services. Such officials tended to work in the same cities and move in the same social circles as their American counterparts.

The military has also used official cover to collect intelligence before going into battle and to prepare logistic support for U.S. forces before they arrive. Sometimes the military uses the CIA; sometimes it uses its own personnel.

In either case, official cover is much less effective today. The main intelligence targets are less likely to be foreign military officials, intelligence officers, or diplomats and are more likely to be religious leaders and students at foreign mosques and temples, organized crime figures, or ideological fanatics. The most valuable targets may be nerdy computer jockeys who run foreign Internet service providers. Why would an official from the U.S. government be meeting with these people? You begin to see the problem. If you want to recruit foreign agents, you need to be able to mix in.

Similarly, clandestine U.S. military operations are likely to be required in countries where the United States has little or no official representation—places like Iraq or Libya, for example. Even if there is an official U.S. presence, providing clandestine support to a military operation could create major complications.

Even worse, if official cover fooled anyone in the Cold War—even then, many experts were skeptical—it is utterly unlikely to fool anyone today. There is so much information available on the Internet and other public databases that it is easy to spot the profile of a case officer working under official cover. A few months ago I spent an afternoon with Margot Williams, director of computer-assisted reporting at the Washington Post. Williams is a wiz with a computer and the Internet, and I wanted to see firsthand how thin official cover is.

For the sake of illustration, I gave Williams a name to look up. Call him Ron; he and I had been at a dinner together a few nights before. I asked her what she could tell me about him. She typed and moved the mouse around. Within 30 seconds, she was able to tell me that Ron was an official at the CIA. Indeed, Ron—who spent a good portion of his career under cover—is now one of the agency’s top officials responsible for clandestine operations. He’s overt now, so we weren’t compromising any secrets, but you’ll see the problem in a minute.

It took just five minutes for Williams to show me how anyone could find out where Ron had lived during the past 15 years; the names of his wife and children and his ex-wife; all court proceedings that he might have been in; all properties that he owned. Williams also listed all of Ron’s previous places of employment, where he went to high school and college, his place of birth, his birthday, and his Social Security number (these last two items, incidentally, are what are used to pass security clearances).

Once Williams figured out that Ron was a CIA official, she then pulled all of the files in the Washington Post’s own archives that mentioned him. By this time we had dozens of records and documents about an official whose relationship to the CIA was at one time officially secret.

And that’s the problem. In the CIA, like most government agencies where people work under cover, officials eventually rise to a level in the organization where they no longer conceal their links to the U.S. government. In the past this might not have mattered as much, but today the Internet can draw up decades of one’s life in seconds—compromising any agent that might be linked to you 10 or 15 years ago. It is harder to recruit agents if they know they might be exposed a decade later. In the future, the problem is just going to get worse as biometrics makes it possible to catalog an individual’s essential facial bone features, iris patterns, or DNA. When that happens, not only will it be hard to fabricate a life history, but it will be hard even to wear any kind of disguise.

After pulling together dozens of pieces of data, Williams launched a link analysis tool, a software package that can take a huge collection of documents, phone listings, schedules, and address books; select names from each; and show the links between each name. So when Williams clicked an icon, the program filled the screen with a complete social network of people and places—all linked to Ron in the middle.

Ron is such a high-ranking official that he is reasonably well known, so the test might not have been fair. So we did a quick run on another CIA officer from the Directorate of Operations—call him Jose—as a more representative case. Jose hadn’t worked under cover for years, but Williams showed me how she could go back to identify some of the places where the official databases had him listed as a U.S. official posted abroad. She also showed how you could cross-check other databases to find people he might have met—in other words, people who might be suspected by their own governments of being spies working for the United States.

Also, by comparing Ron and Jose’s publicly acknowledged employment records, Williams showed how we could cross-reference several officers who had previously worked under cover and put together a profile of what current CIA officers working under cover would look like.

Every intelligence service in the world that’s worth anything knows all about these tools and tricks. The point is, creating effective cover is becoming a huge problem for the United States today, and we need to take it seriously.

In short, we are fooling no one—except possibly ourselves—by thinking our adversaries do not know who our undercover officers are. To be sure, the use of case officers working under official cover still has its uses—secret liaisons with foreign intelligence services, for example. But this traditional approach is bumping up against its limits in the war against the new threat. Because it lacks truly effective cover, U.S. intelligence lacks the ability to attack many of the targets we need to penetrate.

Obstacles to Change

If government officials and outside experts are aware of these problems, why have the intelligence community and armed services been so slow to change?

In fact, the intelligence community does have people working under nonofficial cover (NOC). The problem is that the traditional model of the case officer working under official cover is deeply ingrained in the CIA’s organizational structure, personnel system, and tradecraft. The organization has developed incentives and procedures that make it hard to adopt a different mode of operation.

Suppose you are an American patriot who happens to be dark-skinned, fluent in Farsi, and eager to serve your country. The current system gives you two options.

Choose Option A—working as an NOC—and you have two jobs—as, say, an executive for a company doing business overseas nominally earning $125,000 a year and as a U.S. government employee spying for $55,000 a year. Except that you only get the lower salary. You will have to spend most of the time abroad. You won’t be able to meet with most of your real colleagues in the CIA.

In Option B—working under official cover—you get to live in McLean, Virginia, and drive 15 minutes to work. You get to see your buddies in the cafeteria during lunch, and you can send your kids to Fairfax County schools. You can network like crazy, get steady promotions, and every couple of years go abroad and pose as a U.S. official.

With this kind of choice of career paths, it’s a wonder we can get anyone to work under deep cover at all. But that’s exactly what we need today.

The problem parallels a principle in economic theory—Gresham’s Law, named after Sir Thomas Gresham, a Scottish banker who popularized the idea. Its simplest form is “Bad money drives out good money.” For example, if some one-dollar coins are made of silver and others are made of tin, soon only the tin coins will remain in circulation. People will keep the silver coins because the silver they contain is harder to come by and because the tin ones have the same official value. The same thing happens in organizations that try to operate officers under deep cover and official cover. Soon, almost all the officers are under official cover. Like tin, official cover is easier to come by, and, if the organization rewards all its officers equally, soon no one will want to operate under deep cover.

Truly clandestine intelligence and military operations must be based within organizations whose true functions are entirely secret. Truly clandestine operations must be completely segregated from overt organizations—even organizations that have people under weak or official cover. They need their own career development path and management structure.

In fact, today weak cover can be positively dangerous to anyone trying to work under deep cover. Because it is so easy to spot intelligence officers working under official cover, they are the last people anyone who is serious about hiding their links to the U.S. government wants to meet. Such an encounter would beg the question of why someone who is supposedly a normal, unassuming private party is meeting with a person known to be working undercover for the U.S. government.

One CIA official (who understands the problems of official cover) observes, “Cover is like soap; once you rinse it off, you can’t put it back on.” Effective cover is only going to get tougher in an era of biometrics. Once you identify a clandestine operator, that person is marked for life. Even worse, anyone they contact will become suspect. Cover has to be a way of life.

Alternative Approaches

We need new ways to provide cover—and we need to be clear that whatever we are keeping secret is worth the time, expense, and risk. Fortunately, we have some experience in programs that suggest how we can create more effective cover.

For example, in 1960 the Naval Research Laboratory developed a new scientific satellite called SOLRAD (SOLar RADiation experiment), a 20-inch, 250-pound aluminum ball that measured variations in solar energy. SOLRAD was a real scientific experiment that produced real scientific data. But the satellite also contained a radio receiver that secretly mapped air defense radar transmitters each time it passed over the Soviet Union. The classified name for the project was GRAB (Galactic Radiation and Background). SOLRAD was actually the cover for the first U.S. reconnaissance satellite.

The Navy was so successful in keeping GRAB under cover that, when the CIA published a history of its own early reconnaissance satellite effort in 1995, it inadvertently referred to it as America’s “first successful reconnaissance satellite”—even though GRAB was operational five months earlier. The program was still secret, even to many government analysts.

GRAB stayed secret (unlike other classified satellites developed at the same time) because the Navy was able to embed the project in a much larger, ongoing activity that had a plausible—in fact, a true—reason for operating a satellite. The Naval Research Lab had carried out upper-atmosphere experiments for decades; SOLRAD was the next logical step. The lesson is that you are better off finding cover than creating it.

That was how the Navy hid another secret program 20 years later. The search for the sunken liner Titanic was, in effect, a cover for a highly classified Navy program to inspect the wrecks of two lost Navy submarines: the Thresher, which sank 200 miles off the coast of Cape Cod in 1963, and the Scorpion, which was lost in the mid-Atlantic in 1967.

The Navy was concerned about the condition of the two subs’ nuclear reactors, as well as the nuclear-armed torpedoes that the Scorpion was carrying. However, the Navy did not want to draw attention to the wrecks, lest others—such as the Soviet Union—get the idea to grab parts of the submarines for themselves. So any survey had to be secret.

Robert Ballard worked for Woods Hole Oceanographic Institute, an independent research center in Massachusetts founded in 1930. Today Woods Hole employs about 600 people who work in fields ranging from ocean physics, biology, and engineering to policy studies. Most of its work is purely scientific and totally open. Ballard had taken a year off from Woods Hole and spent some time at Stanford University just as the Information Revolution was taking off. There Ballard got the idea of applying the new technology to create a new field: virtual exploration of the ocean using robotics, just the technology the Navy needed for finding subs deep in the ocean.

Ballard also had a dream: finding the Titanic. People knew where other lost ocean liners—Andrea Doria, Lusitania, Britannic—had gone down, and divers had explored most of them. But no one knew the exact location of the wreck of the Titanic, the grand prize of ocean explorers.

Ballard proposed a deal to the Office of Naval Research: fund most of the development for a deep-sea robot (Ballard would raise the remaining money). Ballard would first use the robot to look for the Thresher and Scorpion and survey the wreck sites. Then he and his team would use whatever time he had remaining to look for the Titanic.

This robot became the Argo/Jason system. Argo served as a docking station for Jason, a smaller sub-robot that carried the lights and camera. Jason could reach tight quarters, like the interior of the Titanic or the torpedo room of a submarine that had been lost years before.

For the Navy, using Ballard’s research as a cover offered many advantages—not the least being Ballard himself. Ballard was an officer in the Naval Reserve. So whenever the Navy needed him to perform a classified mission, it simply reactivated him. Robert D. Ballard, Ph.D., became Commander R. D. Ballard, USNR. He was a bona fide employee of the U.S. government—but still looked like an oceanographer, which he was. For keeping cover, the arrangement was perfect. By publicizing the search for the Titanic—which was bound to get the spotlight anyway—and being coy about how much time Ballard was actually at sea, the Navy succeeded in making sure no one brought up the subject of the Thresher and Scorpion. The secret parts of the mission remained secret until Navy officials decided to reveal the survey operation in 1996 to reassure the public that the reactors posed no safety hazards.

The basic technique was the same as in GRAB: Find a large research institute with enough routine activities so that a secret project can get lost in the noise; find or create a project with a bona fide scientific function. Blend the sensitive operations into the larger project. Publicize the hell out of the public operations, or just be nonchalant when the press follows its own instincts for a good story. Then simply say nothing about the sensitive part.

This, incidentally, was exactly the opposite of what the CIA did in its own deep-sea recovery project. The CIA planned to raise an entire Soviet submarine using the Glomar Explorer, a 600-foot ship built in 1973 especially for the mission. The CIA tried to pass off the Glomar Explorer as an ocean mining platform operated by a newly created company owned by billionaire Howard Hughes. The cover story was as implausible as it sounds, and the project ran for just over a year before its true purpose was blown (at a cost of tens of millions of dollars) by the Los Angeles Times in 1975.

Lessons Learned

Instead of hiring case officers, NOCs, or clandestine military personnel and expecting them to develop access to targets, we will likely have more success if we find people who already have access.

For example, we could find the 40-year-old American working abroad who was in the Marines or State Department back in the 1980s and wouldn’t mind working part-time on a military or intelligence operation today. The government could take advantage of his natural comings and goings in the country where he works. If the man or woman were a retired officer, we could simply reactivate him or her for the mission. Or we could hire the person part time. These people would then report through a truly clandestine program office.

The businessperson, artist, or scientist who is a part-time intelligence officer or clandestine operator for the military services is a new idea that is bound to raise every issue imaginable about contracting, liability, and conflicts of interest. That’s why we have lawyers. We can work out the issues.

The real problem is culture and imagination. The intelligence community currently uses American private citizens who travel abroad and are willing to share information. The military uses contractors. But both treat these people as outsiders—they are not really members of the team (let alone leaders), which usually makes them ineffective.

We need these people possibly even more than traditional light-cover case officers because they are more likely to have the access we need to hostile foreign countries. They may be our only means to penetrate terrorist cells, set up clandestine logistics networks, and perform the other chores of the secret wars that lie ahead.

The alternative is to use the old model, with case officers and military personnel working under official cover—an approach that calls to mind the rationale of the drunk looking for his keys under the streetlight. It’s not where he dropped them; it’s just that the light is better over there.