APPROVED FOR RELEASE 1994
CIA HISTORICAL REVIEW PROGRAM
18 SEPT 95
Guideposts for the dim, replicate half-world where counterintelligence raises deception to the second and third powers.
F. M. Begoum
The double agent operation is one of the most demanding and complex counterintelligence activities in which an intelligence service can engage. Directing even one double agent is a time-consuming and tricky undertaking that should be attempted only by a service having both competence and sophistication. Competence may suffice for a service that can place legal controls upon its doubles, but services functioning abroad-and particularly those operating in areas where the police powers are in neutral or hostile hands–need professional subtlety as well.
Other requisites are that the case officer directing a double agent have a thorough knowledge of the area and language, a high order of ability in complex analytic reasoning, a thorough grounding in local, laws governing espionage, enough time from other duties to run the operation well and report it well, a detailed understanding of the adversary service or services (and of any liaison service that may be involved), adequate control of the agent’s communications, including those with the adversary, a full knowledge of his past (and especially of any prior intelligence associations), a solid grasp of his behavior pattern (both as an individual and as a member of a national grouping), and rapport in the relationship with him.
Like all other intelligence operations, double agent cases are run to protect and enhance the national security. They serve this purpose principally by providing current counterintelligence about hostile intelligence and security services and about clandestine subversive activities. The service and officer considering a double agent possibility must weigh net national advantage thoughtfully, never forgetting that a double agent is, in effect, a condoned channel of communication with the enemy.
Some Western services have become highly skilled through long experience with double agent cases and other counterespionage operations. Of the Communist Bloc services, the Soviets manifest patience and a conceptual pattern both intricate and inherently consistent; to create or enhance confidence in an important double agent they are willing to sacrifice through him information of sufficient value to mislead the reacting service into accepting his bona fides. They make extensive use of provocateurs to establish double agents, especially among emigres. Not much is known about Chinese Communist capabilities in this specialty; available indications suggest mediocrity. The remainder of the Bloc is spotty: the North Koreans are amateurish, the Hungarians and Czechs have demonstrated competence, and the Poles, maintaining an old tradition, show a level of skill (but not of resources) approaching that of the Soviets. We Americans have acquired a broad range of experience since our entry into World War II, but twenty years is not enough time for mastering such an art. We are especially unversed in active and passive provocation.
His Nature and Origins
A double agent is a person who engages in clandestine activity for two intelligence or security services (or more in joint operations), who provides information about one or about each to the other, and who wittingly withholds significant information from one on the instructions of the other or is unwittingly manipulated by one so that significant facts are withheld from the adversary. Peddlers, fabricators, and others who work for themselves rather than a service are not double agents because they are not agents. The fact that doubles have an agent relationship with both sides distinguishes them from penetrations, who normally are placed with the target service in a staff or officer capacity.
The unwitting double agent is an extremely rare bird. The manipulative skill required to deceive an agent into thinking that he is serving the adversary when in fact he is damaging its interests is plainly of the highest order. The way a double agent case starts deeply affects the operation throughout its life. Almost all of them begin in one of the three ways following:
The Walk-In or Talk-In. This agent appears in person, sends an intermediary, makes a telephone call, writes a letter, or even establishes radio contact to declare that he works for a hostile service and to make an offer to turn against it. Although the danger of provocation is always present, some walk-ins and talk-ins have proved not only reliable but also very valuable.
The Agent Detected and Doubled. A service discovering an adversary agent may offer him employment as a double. His agreement, obtained under open or implied duress, is unlikely, however, to be accompanied by a genuine switch of loyalties. The so-called redoubled agent one whose duplicity in doubling for another service has been detected by his original sponsor and who has been persuaded to reverse his affections again -also belongs to this dubious class. Many detected and doubled agents degenerate into what are sometimes called “piston agents” or “mailmen,” who change their attitudes with their visas as they shunt from side to side. Operations based on them are little more than unauthorized liaison with the enemy, and usually time-wasting exercises in futility. A notable exception is the detected and unwillingly doubled agent who is relieved to be found out in his enforced service to the adversary.
The Provocation Agent. The active provocateur is sent by Service A to Service B to tell B that he works for A but wants to switch sides. Or he may be a talk-in rather than a walk-in. In any event, the significant information that he is withholding, in compliance with A’s orders, is the fact that his offer is being made at A’s instigation. He is also very likely to conceal one channel of communication with A-for example, a second secret writing system. Such “side-commo” enables A to keep in full touch while sending through the divulged communications channel only messages meant for adversary eyes. The provocateur may also conceal his true sponsor, claiming for example (and truthfully) to represent a Satellite military service whereas his actual control is the KGB-a fact which the Soviets conceal from the Satellite as carefully as from us.
The passive provocation, or “stake-out,” is a subtler member of the tribe. In Country C Service A surveys the intelligence terrain through the eyes of Service B (a species of mirror-reading) and selects those citizens whose access to sources and other qualifications make them most attractive to B. Service A then recruits from these and waits for B to follow suit. The stake-out has a far better chance of success in areas like Africa, where intelligence exploitation of local resources is far less intensive, than in Europe, where persons with valuable access are likely to have been approached repeatedly by recruiting services during the postwar years.
Sometimes a double agent operation is turned over by a liaison service to a U.S. service or by one U.S. service to another. When such a transfer is to be made, the inheriting service ought to delve into the true origins of the case and acquire as much information as possible about its earlier history.
For predictive purposes the most important clue embedded in the origins of an operation is the agent’s original or primary affiliation, whether it was formed voluntarily or not, the length of its duration, and its intensity. In extreme cases the agent may have volunteered or willingly agreed to work for a hostile service before the U.S. case officer who is now weighing the merits of doubling him was even born. The effects of years of clandestine association with the adversary are deep and subtle; the American case officer working with a double agent of Russian origin against, say, the KGB should never forget that the agent and his Soviet case officer share deep bonds of language and culture, even if the agent is profoundly anti-Communist.
Another result of lengthy prior clandestine service is that the agent may be hard to control in most operations the case officer’s superior training and experience give him so decided an edge over the agent that recognition of this superiority makes the agent more tractable. But add to the fact that the experienced double agent may have been in the business longer than his U.S. control his further advantage in having gained a first-hand comparative knowledge of the workings of at least two disparate services, and it is obvious that the case officer’s margin of superiority diminishes, vanishes, or even is reversed.
The Value of His Services
The nature and value of the double agent’s functions depend greatly on his personal ability as well as on his mission for the other service. He can always report on the objectives and conduct of this mission and possibly more broadly on the positive and counterintelligence targets of the other service or on its plans. If he is skillful and well trained, he can do valuable work by exploiting the weaknesses of others: all intelligence officers of any service, despite their training, have some weaknesses. Some are loose-mouthed, some like to drink, others tend to brag.
The case officer may find his agent to be a wonderful fellow and confide in him, putting him in a good position to elicit specific information and making him the recipient of all manner of unsolicited information. The agent may be able to learn the operational techniques, the security practices, the training methods, and the identity of other members of the service. Possibly, if at a high enough level, he may even be able to report the policies and intentions of the government. Although such a double agent is extraordinary, there are on record some whose reports have been of major national importance. Normally, however, the double agent does not have access to such information.
Often a double agent, after a period of time, is able to report on the capabilities of the other service, if not directly at least by giving information on his own handling from which specific capabilities can be inferred. For example, he can report on the type of support given him in servicing dead drops, providing accommodation addresses, arranging transportation, and supplying technical equipment. If he has been issued some modern technical device, say an automatic transmitter, it can logically be concluded that the service has a good support capability.
The double agent often has access through his travels for the other service to positive intelligence on that country, or on third countries of interest to the controlling service. But even when his mission does not afford such opportunities, he is always able to report his observations of the other service. These bits of information can be accumulated until they give a picture of the other service’s administrative practices, its personnel, and possibly its liaison with other intelligence and security services. Debriefing for this purpose in minute detail is time-consuming, however, and it is a real problem to strike the right balance in the agent’s time between extensive debriefing and running him back into the other service.
The double agent serves also as a controlled channel through which information can be passed to the other service, either to build up the agent in its estimation or for purposes of deception. Often operational build-up material is passed first to establish a better reception for the deception material: obviously the greater the stature of the agent in the eyes of the other service, the better the reception of the reports he provides. In the complex matter of deception we may distinguish here between operational deception, that concerning the service’s own capabilities, intentions, and control of the agent, and national deception, that concerning the intentions of the controlling government or other components of it. National deception operations are usually very delicate, frequently involving the highest levels of the government, and therefore require prior coordination and approval at the national headquarters level.
The double agent channel can be used by the controlling service to insert data into the mechanisms of the other service with a number of possible objectives-for example, to detect its activities in some field. The inserted material is designed to induce certain actions on the part of the other service, which are then observed through another operation or group of operations. The material has to be designed very skillfully if it is to deceive the other service and produce the desired reactions. A sophisticated operation of this type is most likely to be used when the stakes are high or the case complicated. Such a situation might arise if a case officer handling several operations wanted to set up still another and needed to find out in advance what the pertinent operational pattern was. The passing of data through the double agent channel for the consumption of the other service for what ever purpose requires a great deal of knowledge about the other service.
A double agent may serve as a means through which a provocation can be mounted against a person, an organization, an intelligence or security service, or any affiliated group to induce action to its own disadvantage. The provocation might be aimed at identifying members of the other service, at diverting it to less important objectives, at tying up or wasting its assets and facilities, at sowing dissension within its ranks, at inserting false data into its files to mislead it, at building up in it a tainted file for a specific purpose, at forcing it to surface an activity it wanted to keep hidden, or at bringing public discredit on it, making it look like an organization of idiots. The Soviets and some of the Satellite services, the Poles in particular, are extremely adept in the art of conspiratorial provocation. All kinds of mechanisms have been used to mount provocation operations; the double agent is only one of them.
There is still another important function the double agent can perform. He can provide a channel for a recruitment or defection operation against the other service. If he is shrewd and personable enough to have succeeded in establishing a psychological ascendancy over his case officer in the other service, he may be able to recruit him or persuade him to defect. If the attempt fails, of course, the whole operation has to be terminated. In a double agent operation that is valuable only for a certain span of time or one that for any reason is about to collapse, there may be an opportunity at the point of termination to use the agent to make a recruitment or defection approach. The agent can be instructed to make his last job a pitch to the other service’s case officer, revealing that he has been under the control of the opposing service for x number of years, pointing out that the case officer’s name will be mud when he returns to his headquarters, and suggesting that he may as well save his skin and make a switch. In this attempt the agent might be limited to planting the seed, or he might carry through the complete recruitment or defection.
Occasionally a service runs a double agent whom it knows to be under the control of the other service and therefore has little ability to manipulate or even one who it knows has been successfully redoubled. The question why a service sometimes does this is a valid one. One reason for us is humanitarian: when the other service has gained physical control of the agent by apprehending him in a denied area, we often continue the operation even though we know that he has been doubled back because we want to keep him alive if we can. Another reason might be a desire to determine how the other service conducts its double agent operations or what it uses for operational build-up or deception material and from what level it is disseminated. There might be other advantages, such as deceiving the opposition as to the service’s own capabilities, skills, intentions, etc. Perhaps the service might want to continue running the known redoubled agent in order to conceal other operations. It might want to tie up the facilities of the opposition. It might use the redoubled agent as an adjunct in a provocation being run against the opposition elsewhere.
Running a known redoubled agent is like playing poker against a professional who has marked the cards but who presumably is unaware that you can read the backs as well as he can.
Sometimes, although infrequently, double agent operations are started for propaganda purposes. A Soviet-controlled provocateur works for a Western service for a year or two and is then pulled back home, where he is surfaced on the radio and in press interviews to denounce his former Western spy masters. More frequently the Soviets use this trick to get added mileage from an operation that is dying anyway.
Finally, liaison services running a double agent jointly against an adversary quite naturally use this opportunity to assay each other’s capabilities. There is nothing perfidious in this practice as long as it is kept within bounds. Unless the U.S. service operating from a friendly country, for example, can realistically gauge its host’s capabilities in such vital matters as physical surveillance,, phone taps, and hostile interrogation, the operation is likely to go awry.
Since a good deal of nonsense about control sometimes crops up in our thinking about double agents, a definition is first in order. Control is the capacity of a case officer (and his service) to generate, alter, or halt agent behavior by using or indicating his capacity to use physical or psychological means of leverage. A case officer does not control an agent the way he controls an automobile. And a case officer working overseas does not control a double agent the way a policeman controls an informer. The intelligence officer who thinks of control in absolutes of black and white does his operation a disservice; the areas of gray predominate.
First, the U.S. case officer running an operation abroad usually lacks executive powers. Second, the very fact that the double has contact with the opposition affects control. For example, pressure exerted bluntly or blindly, without insight into the agent’s motivation and personality, may cause him to tell the truth to the adversary as a means of escaping from a painful situation. Before the case officer pushes a button on the agent’s control panel he should know what is likely to happen next. Finally, the target service inevitably exercises some control over the double agent, if only in his performance of the tasks that it assigns to him. In fact, it is a primary principle of the counterintelligence service not to disrupt hostile control of the positive half of the operation and thus tip its CI hand. Even if the positive side is being run so poorly that the misguided agent is in danger of coming to the attention of local authorities whose intervention would spoil the CI aspect too, the case officer must restrain his natural impulse to button up the adversary’s operation for him. At the very most, he can suggest that the agent complain to the hostile case officer about insecure practices, and then only if the agent’s sophistication and relationship with that case officer make such a complaint seem normal.
Complete physical control of the double agent is rare in peacetime situations. Normally it is achieved only over the agent captured in war. Limited physical control, however, may be exercised in varying degrees: an agent may have his home in an area where he is subject to complete surveillance or he may live in an uncontrolled area but work in a controlled installation.
The degree to which an agent’s communications can be controlled runs closely parallel with the degree to which he is physically controlled. Communications control, at least partial, is essential: the agent himself is controlled to a considerable extent if his communications are controlled. But even when his communications are completely controlled, a well trained agent doubled against his will can appear to be cooperating but manage at an opportune moment to send a signal to his own service indicating that he is under duress. A number of captured wartime Soviet, British, and German agents did manage to get off such signals.
With only partial control, if the agent is in communication with the opposition service through a courier, dead drop, or live drop, some control or surveillance has to be established over these meetings or servicings. The double agent who makes trips in and out of the area where he can be physically controlled presents a multiplicity of problems.
Assessing His Potential
Acquisition of a double agent may be the result of a deliberate follow-up of leads, or it may be opportunistic. The counterintelligence screening process that forms part of security programs produces many leads. Others may arise in the course of positive operations. Opportunistic acquisition, as of a walk-in, has the disadvantage of being unexpected and therefore unplanned for: the decision to run a double agent should be made only after a great deal of thought, assessment, and evaluation, and if the candidate comes as a volunteer, the service may have to act without sufficient time for reflection. In this situation the necessity of assessing the candidate conflicts also with the preservation of security, particularly if the officer approached is in covert status. Volunteers and walk-ins are tricky customers, and the possibility of provocation is always present. On the other hand, some of our best operations have been made possible by volunteers. The test of the professional skill of an intelligence organization is its ability to handle situations of this type.
When a double agent candidate appears, judgments are needed on four essential questions in order to decide whether a potential operation . exists, whether to run the candidate, and whether the service has the capability to do so.
Has he told you everything? Enough information can ordinarily be obtained in one or two sessions with the candidate to permit testing by polygraph, investigation of leads, and file checks. These steps must be taken very quickly because it is not possible to un-recruit a man. The two areas of possible concealment which are especially dangerous are prior intelligence ties and side-commo.
Does he have stayability? This term combines two concepts-his ability to maintain access to the counterintelligence target for the foreseeable future, and his psychological stamina under the constant (and sometimes steadily increasing) pressure of the double agent’s role. If he lacks stayability he may still be useful, but the operation must then be planned for short range.
Does the adversary trust him? Indications of adversary trust can be found in the level of the communications system given him, his length of service, the seniority of the adversary case officer, the nature and level of requirements, and the kind and extent of training provided. If the opposition keeps the agent at arm’s length, there is little prospect that doubling him will yield significant returns.
Can you control his commo both ways?Control of communications on your own side can be difficult enough, especially if the agent lives in hostile territory. But control of adversary channels is hard under even the best of circumstances. It requires a great deal of time, technical skill, and-as a rulemanpower.
Negative answers on one or even two of these questions are not ground for immediate rejection of the possible operation. But they are ground for requiring some unusually high entries on the credit side of the ledger.
The initial assessment is made essentially through interrogation, used in a broad sense to include friendly debriefing or interview. The interviewing officer may be relaxed and casual, but underneath the surface his attitude is one of deliberate purpose: he is trying to find out enough to make an initial judgment of the man. A human being in a stress situation is a complicated personality, and the interviewing officer must penetrate below the surface, sensing the man’s emotions and mental processes. For instance, if an agent walks in, says he is a member of another service, and reveals information so sensitive that the other -service would surely not give it away just to establish the informant’s bona fides, there are two possibilities: either the agent is telling the truth or he is attempting a provocation. Sometimes the manner in which the man conducts himself will suggest which of the two it is.
In addition to establishing the individual’s true identity and examining his documents, the officer should get as many details as possible on the service he belongs to and his position in it. His job may be such that it is necessary to make a fast initial judgment: for example, he may be one of the two or three intelligence officers in a small office where a prolonged absence would cause suspicion.
It may be more difficult to determine the reason why the agent presented himself than to establish who he is and what service he represents, because motivation is a complex of mental and emotional drives. The question of the double agent’s motivation is approached by the interviewing officer from two angles-the agent’s professed reasons and the officer’s own inferences from his story and behavior. The agent may profess a love for democracy, but the officer cannot elicit any convincing evidence of such a love. Some of the agent’s reasons may not ring true. To decide between what the officer thinks the motive is and what the agent says it is is not easy, because double agents act out of a wide variety of motivations, sometimes psychopathic ones like a masochistic desire for punishment by both services. Others have financial, religious, political, or vindictive motives. The last are often the best double agents: they get pleasure out of deceiving their comrades by their every act day after day.
Making the judgment about the agent’s psychological and physical suitability is also difficult. Sometimes a physician or psychiatrist can be called in under some pretext. For the most part, however, professional assistance is not available, and the interviewing officer must rely upon his own skill in assessing human beings and understanding what makes them tick. Such skill can be acquired only by experience.
Experience suggests that some people who take to the double agent role-perhaps a majority of willing ones, in fact have a number of traits in common with the con-man. Psychiatrists describe such persons as sociopaths. From the point of view of the double agent operation, here are their key traits:
- They are unusually calm and stable under stress but cannot tolerate routine or boredom.
- They do not form lasting and adult emotional relationships with other people because their attitude toward others is exploitative.
- They have above-average intelligence. They are good verbalizers-sometimes in two or more languages.
- They are skeptical and even cynical about the motives and abilities of others but have exaggerated notions about their own competence.
- Their reliability as agents is largely determined by the extent to which the case officer’s instructions coincide with what they consider their own best interests.
- They are ambitious only in a short range sense: they want much and they want it now. They do not have the patience to plod toward a distant reward.
- They are naturally clandestine and enjoy secrecy and deception for its own sake.
In brief, the candidate must be considered as a person and the operation as a potential. Possibilities which would otherwise be rejected out of hand can be accepted if the counterintelligence service is or will be in a position to obtain and maintain an independent view of both the double agent and the case. Perhaps such independent collateral can be acquired from another operation, in being or in the offing.
The officer’s estimate of the potential value of the operation must take into consideration whether his service has the requisite personnel, facilities, and technical support; whether running the operation will prejudice other activities of his government; whether it will be necessary or desirable, at the outset or later, to share the case with foreign liaison; and whether the case has political implications.
Running the Operation: Do’s and Don’ts
The following principles apply to the handling of all double agent operations in varying degrees. In composite they form a check-list against which going operations might be periodically reviewed-and given special examination with the appearance of danger signals.
- Remember that testing is a continuous process. Use the polygraph early and run later tests as well. Be alert for changes in agent motivation. When you can do so securely, employ such additional means as further records investigation, checking out of operational leads, technical analysis of documents and equipment, surveillance and countersurveillance, mail and telephone taps, and substantive analysis of reporting. Although 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.
- Train the agent, but only as a double. Give him training as needed in security of the doubled part of the operation, CI reporting, cover as a double, the handling of technical equipment used for CI purposes, etc. But do not poach on enemy territory by teaching him the skills he needs for adversary purposes. An “inexplicable” improvement in his work would draw suspicion.
- Be careful about awakening in the hostile service an appetite which cannot later be satisfied without giving away too much. Do not furnish build-up material that transcends the agent’s access or that will rouse adversary interest in sensitive areas. In general, let the agent carry out his adversary assignments on his own instead of spoonfeeding him, although there are exceptions to this rule
- Require the agent to report and, as security permits, turn over to you everything he gets from the other sidemoney, gifts, equipment, documents, etc. If he is permitted to hold out anything he may grow confused about which side he is working for. But do not be too rigid in following this rule. It may be better, for example, instead of confiscating his payments from the adversary, to put them into a third country bank account and promise him the lump sum upon successful termination.
- Avoid interference. Oblige the other service to solve any problems that arise from the agent’s activity on its behalf. For example, if the agent is arrested or threatened with arrest by local authorities, the counterintelligence officer should not rush to his aid. The threatened agent should take his problem to the adversary, who may be forced to surface a new asset in order to help him. It should be explained to the agent that you are not indifferent but on the contrary too concerned about his security to blow him by meddling.
- Be constantly alert for hostile provocation. The opposition may create a security crisis for the agent, or he may at their instigation report such a crisis. If he does, examine the claim thoroughly and test it.
- If the adversary appears to be a Satellite service, do not lose sight of the possibility that the agent is being manipulated behind the scenes by the Soviets, probably without the Satellite’s knowledge.
- Keep analyzing the agent as well as the case. Do not be satisfied to fix a label (such as “anti-Communist”) to him instead of learning to understand him.
- If the agent is to pass classified U.S. information to the adversary, keep precise records of what was passed, which department or agency cleared the release, and the dates.
- Do not plan a deception operation or pass deception material without prior headquarters approval.
- Do not reveal your service’s assets or CI knowledge to a double. It is vital that double agents be run within the framework of their own materials-the information which they themselves supply. Junior CI officers, especially, may be tempted to impress double agents with the omniscience of their service. The more you keep from an experienced double the information he should not have, the more he will be reassured that his own safety is in good hands.
- Prepare all briefings carefully. Have the agent rehearse his instructions. If you think it advisable, brief him on resistance to interrogation; but be cautious, if you do, about revealing to him the specifics and scope of your knowledge of the adversary.
- Mirror-read. Look at the operation from the viewpoint of the hostile service. But be careful not to impute to it the motives, ideas, methods, or other characteristics of your own service. Do not put the adversary in your place; put yourself in his, a task which requires both knowledge and understanding of him.
- Do not run the operation in a vacuum. Be aware of any political implications that it may have, locally or internationally.
- Do not hesitate to ask for help.
- Review the case file periodically. Restudy of the operation sometimes throws into relief facts previously ignored, misinterpreted, or improperly linked to one another. As new information develops, it will throw a new light on the old facts. And review cover now and then-for your service, yourself, the agent, and your meetings with him.’ Consider whether new developments require any changes.
- Decide early in the operation how it will be terminated if the need arises. Do not merely drop it without further steps, leaving an unsupervised hostile agent in place. If he is to be turned over to a local security service, try to make the transfer while there is still some equity in it for them.
- If the operation is joint, weigh, its probable effect upon the liaison relationship.
- Keep a full record, including dates, of all adversary assignments given the agent.
- Report the case frequently, quickly, and in detail. The hostile services are centralized. Pitting against them the limited resources of one U.S. officer or field installation means giving them needlessly favorable odds. Only timely and full reporting to your headquarters will permit it to help you effectively.