Clandestine HUMINT operational techniques

Basic agent recruiting

This section deals with the recruiting of human resources who do not work for a foreign intelligence service (FIS). For techniques of recruiting FIS personnel, see Counterintelligence.

In principle and best practice, all country B officers in country A report to an executive function in their home country. In CIA terms, this might be a head of a country desk or a regional desk. Russian practice was to refer to "Center".

Actual recruiting involves a direct approach by a case officer who has some existing access to the potential recruit, an indirect approach through an access agent or proprietary, or has reason to risk a "cold" approach. Before the direct recruitment, there may be a delicate period of development. For details, see Clandestine HUMINT asset recruiting.

Basic agent operations

This section deals with the general structure of running espionage operations. A subsequent section deals with Specialized Clandestine Functions, and another with Support Services for both basic and specialized operations

The agent may join, or even create, a new network. In the latter case, the agent may be called a lead agent or a principal agent. The latter term is also refers to access agents, who only help in recruiting.

Well-managed agent relationships can run for years and even decades; there are cases where family members, children at the time their parents were recruited, became full members of the network. Not all agents, however, operate in networks. A Western term for agents controlled as individuals is singleton. This term usually is reserved for the first or most sensitive recruitments, although specialized support personnel, such as radio operatives acting alone, are called singletons. In Soviet tradecraft, the equivalent of a singleton is a separated acting agent. Professional intelligence officers, such as Robert Hanssen, may insist on being singletons, and go even farther, as with Hanssen, refuse in-person meetings. Even as a singleton, the agent will use security measures such as secure communications.

Agents also may operate in networks, for which the classic security structure is the cell system.

The agent may join a proprietary, although that is more likely to be for access or support agents.

Training

Before the agent actually starts to carry out assignment, training in tradecraft may be necessary. For security reasons, this ideally will be done outside the agent's own country, but such may not always be possible. Increasingly less desirable alternatives might be to conduct the training away from the operational area, as in a safe house in a resort, and then a safe house inside the operational area.

Among the first things to be taught are communications tradecraft, beginning with recording the material of interest. Skills here can include the operation of cameras appropriate for espionage, methods of carrying out documents without detection, secret writing.

Once the information is captured, it must be transmitted. The transmission may be impersonal, as with dead drops or car tosses. It may involve carriers. It may be electronic. If there is a need for personal meetings, the agent must know how to request them, and also to alert the network leader or case officer that the agent may be under suspicion.

Teaching countersurveillance techniques to agents is a calculated risk. While it may be perfectly valid for an agent to abort a drop or other relatively innocent action, even at the cost of destroying valuable collected material, it is much more dangerous to teach the agent to elude active surveillance. The ability to elude professional counterintelligence personnel following the agent, for example, may confirm the counterintelligence organization's suspicion that they are dealing with a real agent.

Still, the agent may need to have an emergency escape procedure if he confirms he is under surveillance, or even if he is interrogated but released.

Continued testing during operations

Case officers should constantly test their agents for changes in motivation or possible counterintelligence compromise. While "name traces cannot be run on every person mentioned by the agent, do not be stingy with them on persons who have familial, emotional, or business ties with him" to detect any linkages to hostile counterintelligence. Until an agent is well established as reliable, meetings must always be done with care to avoid detection. "The prime emphasis is put on vigilance and checking-has he been planted by the local counterintelligence, are his motives in agreeing to collaborate sincere? The need for personal meetings with such an agent is increased, for they give the opportunity to assess him more completely."

An experienced US operations officer emphasized that field operations personnel should report status and progress often. Only with such reporting can a headquarters staff remain vigilant, looking globally for penetrations, and also aware of political implications. Reporting and headquarters advice is critical for joint operations (i.e., with the intelligence service of another country). Headquarters, aware of all joint operations with a given service, can give advice from a broader viewpoint without compromising the need for local initiative.

Operating the agent

Even with the most sensitive agents, occasional personal meetings are important in maintaining psychological control. Nevertheless, some agents, especially trained intelligence officers like Robert Hanssen, will almost never meet, but provide material good enough to prove their bona fides. A Soviet officer commented, whatever an agent's role in the intelligence net, personal contact should be made with him only when it is impossible to manage without it. The number of meetings should be kept as low as possible, especially with sources of valuable information.

Personal meetings may be held to give an agent his next assignment and instructions for carrying it out, to train him in tradecraft or the use of technical or communications equipment, to transmit documents, reports, technical equipment, money, or other items, or to fulfill several of these purposes. In actual practice several purposes are usually served by a meeting. In addition to its particular objectives more general needs can be filled. A meeting held for training purposes may be a means for clarifying biographic data on the agent or his views on various subjects. At every meeting with an agent one should study him and obtain new data on his potential and talents, thereby providing a better basis for judging his sincerity and deciding how much trust to place in him.

Agents, to varying extents, need reinforcement. Salary is important and also gives a lever of compromise, although pressing it too hard can offend a truly ideologically motivated agent. Some agents benefit from recognition that they can never show, such as a uniform of your service, or decorations from it.

Agents will be more comfortable if they believe that they will have protection, preferably exfiltration, if compromised. Protecting their families may be even more important. When the agent operates in a country with a particularly brutal counterintelligence service, providing them with a "final friend", or means for suicide, can be comforting even if they never use it.

Agent communications

This section deals with skills required of individuals, either agents or support personnel. Most skills are concerned with communications.

Meeting places for personal meetings

A Soviet officer commented, perhaps counterintuitively, that it is harder to have longer meetings with agents when the case officer is under diplomatic cover. The reason is that local counterintelligence is aware of the case officer, where the existence of an illegal (i.e., nonofficial cover in US terms) officer may not be known to them. For the legal officer, "here it is best either to have reliable safehouses or to deliver the agent discreetly to the official residency building. The latter is a serious operational move. If neither is feasible, it is better to have Headquarters dispatch an officer to a third country, either legally or illegally, for the meeting."

Clandestine transfer operated by humans

It is a case-by-case decision whether the material exchanged should have safeguards against accessing it in other than a precise manner. One straightforward protection method is to have the material on exposed photographic film, in a container that does not suggest that it contains film and might be, innocently, opened in a lighted room. Self-destruct devices also are possibilities, but they confirm that the transfer involved sensitive material.

Brush pass and other physical exchange with couriers

Under the general term "brush pass" is a wide range of techniques in which one clandestine operative passes a physical item to another operative. "Brush" implies that the two people "brush" past one another, typically in a public place and preferably a crowd, where random people interfere with any visual surveillance. In a properly executed brush pass, the agents do not even stop walking; at most, they may appear to bump into one another.

During the brief contact, a common means of executing the exchange is for both to be carrying otherwise identical objects, such as a newspaper, briefcase, or magazine. The information being exchanged is in one of them. As the two people separate, they still appear to be holding the same object in the same hand.

More challenging versions are reminiscent of passing a baton in a relay race, and would be most commonly done with small objects such as a photographic film cartridge. In this more dangerous method, the transfer is from hand to hand, or from hand into a pocket. While this technique obviously takes better manual dexterity and is more prone to error, it has the countersurveillance advantage that the operatives are not carrying anything after the transfer, and can blend into a crowd even more easily.

A variation of the brush pass is the live letter drop, in which one agent follows a predefined route, on foot, with a prepared report hidden in a pocket. En route, a second agent unknown to the first agent picks his/her pocket and then passes the report on unread, either to a cut-out or to an intelligence officer. This technique presents opportunities both for plausible deniability and for penetration by hostile agents.

Dead drop

A dead drop is a container not easily found, such as a magnetized box attached to a metal rack in an out-of-sight alley. The box could be loosely buried. It should be possible to approach the container to fill or empty it, and not be easily observed from a street or window.

Typically, a clandestine collector will put espionage material, perhaps in encrypted form, into the box, and use some prearranged signal (i.e. signal site) to let a courier know that something needs to be taken out of the box and delivered to the next point on the route to the case officer. Such a route might have several dead drops. In some cases, the dead drop might be equipped with a device to destroy its contents unless it is opened properly.

Signals to tell a courier, or a case officer if there is no intermediate courier, that the dead drop needs service can be as simple as a piece of colored tape on a lamp post or perhaps a set of window shades raised and lowered in a specific pattern. While "wrong number" calls with a predefined apology can be used, they are more vulnerable to surveillance if the phone in question is tapped.

Car tosses

A car toss can take many forms, one of which can be considered a moving dead drop. An agent or courier can put a magnetized box inside a bumper on a parked car.

In some cases, if a car can drive slowly down a street or driveway not easily observed, a courier can toss a message container into an open window, making the transfer method intermediate between a brush pass and a dead drop.

Cars with diplomatic immunity have advantages and disadvantages for tosses. They cannot be searched if the toss is observed, but they also are followed more easily. Diplomatic cars usually have distinctive markings or license plates, and may be equipped with electronic tracking devices. Counterintelligence could wait until the car is out of sight following a toss, then apprehend and interrogate the courier, or simply keep the courier under surveillance to discover another link in the message route.

Methods of protecting message content

A message left in a dead drop, or dropped during an improperly executed brush pass, is quite incriminating if counterintelligence personnel can immediately see suspicious information written on it. The ideal material for transfer looks quite innocuous.

At one time, invisible ink, a subset of steganography, was popular in espionage communications, because it was not visible to the naked eye without development by heat or chemicals. While computer-based steganographic techniques still are viable, modern counterintelligence laboratories have chemical and photographic techniques that detect the disturbance of paper fibers by the act of writing, so the invisible ink will not resist systematic forensic analysis. Still, if its existence is not suspected, the analysis may not be done.

Microphotography

Another technique, for hiding content that will resist casual examination, is to reduce the message to a photographic transparency or negative, perhaps the size of the dot over the letter "i" in this article. Such a technique needs both a laboratory and considerable technical skill, and is prone to damage and to accidentally falling off the paper. Still, it does have a countersurveillance value.

Encryption

Encryption, especially using a theoretically secure method, when properly executed, such as the one-time pad, is highly secure, but a counterintelligence agent seeing nonsense characters will immediately become suspicious of the message that has been captured. The very knowledge that a dead drop exists can cause it to be trapped or put under surveillance, and the member of a brush pass that carries it will be hard-pressed to explain it.

One-time pad encryption has the absolute requirement that the cryptographic key is used only once. Failure to follow this rule caused a serious penetration into Soviet espionage communications, through the Venona project analysis.

It is extremely difficult for a nonprofessional to develop a cryptosystem, especially without computer support, that is impervious to the attack by a professional cryptanalyst, working for an agency with government resources, such as the US NSA or Russian Spetssvyaz. Still, when the message is very short, the key is random or nearly random, some methods, like the Nihilist Straddling checkerboard may offer some resistance. Improvised methods are most useful when they only have to protect the information for a very short time, such as changing the location or time of an agent meeting scheduled in the same day.

Plain language code

Less suspicious when examined, although very limited in its ability to transfer more than simple content, is plain language code. For example, the final attack order for the Battle of Pearl Harbor came in a radio broadcast of the Japanese phrase, "Climb Mount Niitaka". Subsequent espionage communications referred to ships as different types of dolls at a doll repair shop.

Plain language code is most effective when used to trigger a preplanned operation, rather than transfer any significant amount of information.

Steganography, covert channels, and spread spectrum

Steganography, in the broadest sense of the word, is a technique of hiding information "in plain sight" within a larger message or messaging context. It is hard to detect because the secret message is a very small component of the larger amount, such as a few words hidden in a Web graphic.

Even more sophisticated computer-dependent methods can protect information. The information may or may not be encrypted. In spread-spectrum communications, the information is sent, in parallel, at very low level through a set of frequencies. Only when the receiver knows the frequencies, the time relationship on when a given frequency or other communications channel will carry content, and how to extract the content, can information be recovered. Basic spread spectrum uses a fixed set of frequencies, but the signal strength in any one frequency is too low to detect without correlation to other frequencies.

Frequency-hopping spread spectrum is a related technique, which can use the parallel transmission of true spread spectrum, not using any one frequency long enough for plausible interception. The pattern of variation among channels may be generated and received using cryptographic methods.

Methods of protecting against electronic detection of the fact of messaging

Avoiding detection of radio signals means minimizing the clandestine transmitter's exposure to hostile direction-finding. Modern techniques generally combine several methods:

• Burst transmission or otherwise minimizing
• High-gain antenna and/or directional antenna
• Receiver or relay away from detectors, as, for example, satellites.

Exploring agent information often meant a good deal of interaction, in which the home service would clarify what the agent reported, give new orders, etc. One approach used in World War II was the Joan-Eleanor system, which put the case officer into an aircraft at high altitude. From that altitude, there could be fast interaction in voice, so that they get to the key issues faster than with many separately encrypted and transmitted messages. The modern equivalent is a small, low probability of intercept radio transceiver, using a directional antenna aimed at an orbiting satellite communications relay. Avoiding detection of radio communications involves all the principles of transmission and reception security.

Termination

For any number of reasons, a human source operation may need to be suspended for an indefinite time, or definitively terminated. This need rarely eliminates the need for protecting the fact of espionage, the support services, and the tradecraft and tools provided.

One of the most difficult challenges is ending an emotional relationship between the case officer and agent, which can exist in both directions. Sometimes, an agent is unstable, and this is a major complication; perhaps even requiring the evacuation of the agent. More stable agents may be happy with termination bonuses, and perhaps a future emigration opportunity, that do not draw attention to their own side's counterintelligence. In some instances, an intelligence agency may issue a "burn notice", indicating to other such agencies that an individual is an unreliable source of information.

Especially in the case of non-national organizations, termination can be very literal, ranging from having a trusted operative kill the problematic agent, or, when culturally appropriate, sending the agent on a suicide mission.

When the clandestine phase is preparation for a DA mission such as the 9/11 attacks, or the assassination attacks, using suicide bombers, by the Liberation Tigers of Tamil Eelam, termination of the operational cells is rather obvious. If there are support cells in the operational area, they may be vulnerable, but it would be good tradecraft to withdraw them shortly before the attack.

Special clandestine services

Agents of influence

An agent of influence, being witting or unwitting of the goals of a foreign power B, can influence the policy of Country A to be consistent with the goals of Country B.

In Soviet theory, influencing policy was one aspect of what they termed active measures (aktivnyye meropriyatiya). Active measures have a different connotation than the Western concept of direct action (DA), although Soviet active measures could include wet affairs (mokrie dela) conducted by Department V of the KGB, "wet" referring to the spilling of blood.

Strategic deception

Intelligence organizations occasionally use live, or even dead, persons to deceive the enemy about their intentions. One of the best-known such operations was the British Operation Mincemeat, in which a dead body, bearing carefully misleading documents, was put in British uniform, and floated onto a Spanish beach. In World War II, Spanish security services, while officially neutral, often passed information to the Germans, which, in this case, is exactly what the British wanted done. This operation was under the control of the Twenty Committee, part of the British strategic deception organization, the London Controlling Section. A related British operation in World War I was run by a controversial military officer, Richard Meinertzhagen, who prepared a knapsack containing false military plans, which the Ottoman allies of the Germans were allowed to capture. The plans related to false British strategy for the Sinai and Palestine Campaign, setting up a successful surprise attack in the Battle of Beersheba and the Third Battle of Gaza.

Active measures, however, reflected a national effort to influence other countries to act in concert with Soviet goals. These measures could involve state organizations up to and including the Politburo, much as the World War II British organization for strategic deception, the London Controlling Section, and its US counterpart, Joint Security Control, could get direct support from the head of government. Much of the Soviet responsibilities for active measures was focused in the KGB. Its "First Main Directorate uses active measures such as agents of influence, propaganda, and disinformation to promote Soviet goals."

In the present political context of Western democracies, the sensitivity, and separation, of clandestine and open contacts do not lend themselves to the process of building agents of influence.

"Active measures is not exclusively an intelligence activity, and in this sense it differs from the similar American concept of covert action. There are many differences between active measures and covert action. One is the Soviet ability to mesh overt and covert influence activities through centralized coordination of party, government, and ostensibly private organizations dealing with foreigners. Despite interagency coordination mechanisms, the United States is too pluralistic to achieve full coordination between all the overt and covert means of exercising influence abroad. Other major differences are in scope, intensity, and importance attributed to active measures and covert action, and in immunity from legal and political constraints."

While deception and influence operations could involve the highest levels of Allied governments in World War II, it is worth noting that while the West generally speaks of military deception, strategic deception operates at a higher level. A Soviet, and presumably Russian, term of art, maskirovka or "denial and deception", is much broader than the current Western doctrine of deception being run by lower-level staff groups.

In the military, responsibility for maskirovka easily can be at the level of a deputy chief of the General Staff, who can call upon all levels of government.

Returning to KGB doctrine, presumably still present in the SVR, "Influence operations integrate Soviet views into foreign leadership groups. Propaganda operations take the form of disinformation articles placed in the foreign press. Disinformation operations are false documents designed to incite enmity toward the United States."

The Second Main Directorate of the KGB, whose responsibilities are now primarily in the Russian FSB, is responsible for the recruitment of agents among foreigners stationed in the Soviet Union. The KGB influences these people unwittingly, as most regard themselves too sophisticated to be manipulated.

"The second deception program is counterintelligence, which aims to neutralize the efforts of foreign intelligence services. It achieves this through the use of non-Soviet double agents and Soviet double agents. Non-Soviet double agents are foreign nationals who have been 'turned'. A Soviet double agent is a Soviet with access to classified information. These officials may be used as false defectors".

"Influence operations integrate Soviet views into leadership groups. The agent of influence may be a well- placed, 'trusted contact' who consciously serves Soviet interests on some matters while retaining his integrity on others, or an unwitting contact who is manipulated to take actions that advance Soviet interests on specific issues of common concern."

Direct action services

There is no consensus on whether it is, or is not, advisable to intermingle espionage and direct action organizations, even at the headquarters level. See Clandestine HUMINT and Covert Action for more history and detail. A terminology point: current US terminology, ignoring an occasional euphemism, has now consolidated espionage into the National Clandestine Services. These are part of the CIA Directorate of Operations, which has some responsibility for direct action (DA) and unconventional warfare (UW), although the latter two, when of any appreciable size, are the responsibility of the military.

There is much more argument for doing so at headquarters, possibly not as one unit but with regular consultation. Certain services, such as name checks, communications, cover identities, and technical support may reasonably be combined, although the requirements of a particular field network should be held on a need-to-know basis.

Other countries might have the functions under the same organization, but run them in completely different networks. The only commonality they might have is emergency use of diplomatic facilities.