The Spy Wars

New York Times Magazine
September 28, 1980

by Edward Jay Epstein

In July 1977, President Jimmy Carter's secret Special Coordinating Committee, the White House unit that oversees the clandestine activities of the CIA, received a piece of dismaying news: A Central Intelligence Agency spy in the Kremlin, "Trianon," had been apprehended by the K.G.B., the Soviet intelligence service. In 1978, the Soviet press reported that this American spy had been tried for treason and sentenced to death.

"Trianon" was the code name for Anatoly N. Filatov, a 37-year-old aide in the Soviet Foreign Ministry. The CIA had caught him in a sex trap in Algiers in 1976, when he was attached to the Soviet Embassy in Algeria. After being confronted with compromising photographs, Filatov was persuaded or blackmailed, as he is reported to have claimed at his trial, to work as a spy for the CIA when he was reassigned to the Foreign Ministry in Moscow. He was supplied with all the necessary paraphernalia for espionage: a miniature camera for photographing secret documents, a "burst" transmitter r signaling his contact in the American Embassy in Moscow, and a "dead drop" on a Moscow bridge, where he could inconspicuously leave his microfilm for American intelligence agents to pick up.

How he was so quickly caught by the K.G.B. has been a mystery of immense concern to US intelligence. Was he detected through routine Soviet surveillance? Was he exposed by an accidental leak from American intelligence? Or was he betrayed by a mole planted inside American intelligence?

In response to a request from Senator Daniel Patrick Moynihan, Democrat of New York, and Senator Malcolm Wallop, Republican of Wyoming, the Senate Select Committee on Intelligence conducted an investigation into the circumstances that led to Filatov's exposure, which itself opened up a Pandora's box of secrets about the spy war.

When the Senate Select Committee on Intelligence was briefed on the Filatov case shortly after his arrest in 1977, according to one staff member of the committee, it found that the case had thrown the American intelligence community into confusion. Consternation arose because Filatov was apparently the only United States agent in a position of access to secrets in the Soviet Union - he was, in the language of the intelligence world, a "mole." Moreover, incredible as it may seem, he may have been the only mole that the CIA had established inside the Kremlin in more than a decade. According to one high Government official, who was familiar with the major CIA operations between 1969 and 1977, the CIA failed to establish a single productive mole in the Soviet Union between the arrest of Col. Oleg Penkovsky in Moscow in 1962 and the recruitment of Filatov in 1976. This intelligence gap was also cited by former CIA executives and a staff member of the Senate Select Committee on Intelligence.

The only exceptions mentioned by these sources were two Soviet Unite Nations diplomats - codenamed "Top Hat" and "Fedora," recruited by the FBI in New York and a Soviet diplomat codenamed "Igor," recruited by the CIA in Washington, during the 1960s. In all three cases, however, CIA counterintelligence assessed that they were double agent working for the K.G.B., and, in any case, all three returned to Moscow and cut off their contacts to the CIA.

The primary task of any clandestine intelligence service, whether the CIA or the K.G.B., is to establish moles within the enemy's inner sanctum who are in a position to warn changes in its plans and intentions. "No intelligence can function unless it has secret sources," Richard Helms, a former Director of Central Intelligence, pointed out to me.

There are, be sure, other profitable ways of gathering intelligence, such as overhead surveillance of a potential enemy's activities, tapping into its underwater cables by submarines and the interception of communications by powerful antennae. Such reconnaissance requires highly sophisticated equipment, such as spy planes, satellites, underwater robots and computers, but it does not require the operation of clandestine service. And, even in the age of satellites and electronic wizardry, clandestine services are deemed necessary to report on the strategic thinking of an adversary and other clues that cannot be intercepted by remote platform. The clandestine service specializes in the spotting, compromising, recruiting and handling of moles on a regular basis. This is called Human intelligence, or in CIA-talk, HUMINT.

While public debate over the CIA, fueled by Presidential inquiries and Congressional investigations, has narrowly focused on the charge that the agency has abused its power by spying on domestic groups outside its legal purview, the secret concern in intelligence circles, which has not surfaced in any of the many public hearings, is that the CIA is not spying effectively on its principal adversary: the Soviet bloc. As William Harris, a counterintelligence expert from the RAND Corporation put the question: "Why has the C. LA. repeatedly failed to penetrate the Soviet system by recruiting agents?"

Within the CIA itself, this question has been the center of a bitter and destructive debate that has persisted unresolved for some 20 years. On one side of the issue, it is argued that the K.G.B. has successfully established its own moles with the CIA and other US intelligence services, and that these moles report to Moscow the secret plans and sources of the CIA, thereby making it impossible for the CIA to recruit, or keep secret,- its own moles.

Tennant Bagley Jr., who was the deputy chief of the CIA's Soviet Bloc Division in the mid-1960's and was responsible for countering the activities of Soviet intelligence, explained to me that "it takes a mole to catch a mole." According to his view, the two most successful moles that the CIA ever recruited, Col. Peter Popov (195358) and Colonel Penkovsky (1961-62), were both caught by Soviet intelligence because they had been betrayed by a K.G.B. mole, or moles, working in American intelligence. Bagley claimed, moreover, to have seen during his tenure in the CIA direct evidence of a mole "feeding back," as he put it, operational plans of the CIA to the K.G.B. "In one case, Soviet intelligence clearly knew about an elaborate CIA plan to recruit a Soviet-bloc diplomat in Switzerland," he pointed out. He knew of no productive mole that the CIA had recruited in the Soviet Union since the capture of Penkovsky in 1962. (Bagley retired from the C-I-A. in 1972.) He accounted for this failure in blunt terms: "It is impossible for the CIA to maintain any secret sources if it is penetrated."

This argument was carried much farther by James Jesus Angleton, who served as the CIA's counterintelligence chief until 1975. Angleton, theorizing on the basis of information supplied by Soviet defectors, believed that he had pinpointed the K.G.B. "Penetrations," as he called them, in the Soviet Bloc Division Of the CIA In 1963, he began purging or transferring four possible suspects. When these "administrative measures," as he called them, did not result in ferreting out the mole or plugging the apparent leak, Angleton took more drastic action. In 1960s, he explained to me, he completely "cut off" the entire Soviet Bloc Division from information about highly sensitive cases. This step led to the near paralysis of the Soviet Bloc Division, which was then responsible for all CIA intelligence activities in the Soviet Union and Eastern Europe.

on the other side of the debate, a large number of CIA officers, such as William Colby, who became Director of Central Intelligence in 1973, who believed that the mole issue was divisive, demoralizing and ultimately a dangerous distraction. They argued that Popov, Penkovsky and other CIA moles were caught by the K.G.B. either through routine surveillance procedures or because of a blunder or mishap in American intelligence - and not through any information supplied by a mole. These intelligence officers viewed the deductive search for moles as "sick think," as Jack Maury, a former head of the CIA's Soviet Bloc Division, described it to me. Indeed, William Colby blamed the failure of the CIA to recruit agents in the Soviet Union on the mistaken fear that there was a mole in the CIA who would quickly betray them. When he became Director, he fired Angleton and transferred other counterintelligence officers who had worked under him. He also did away with the tight compartmentalization' of information that Angleton had insist Colby explains in his autobiography that he took these actions because he believed that Angleton's "ultra conspiratorial turn of mind had, at least in recent years, become more of a liability than an asset to the agency-"

The dismissal of Angleton did not end the debate. When the K.G.B. uncovered Filatov, the CIA again had to come to grips with the possibility that Soviet intelligence had a source in the agency. Even though CIA officials told the Senate Select Committee On Intelligence that Filatov's detection had come about because of an inadvertent statement to the press by one of national security adviser Zbigniew Brzezinski's deputies on the National Security Council, a number of counterintelligence officers believed that Filatov had been betrayed by a mole in the CIA

In fact, the CIA had cogent evidence in its files testifying in no uncertain terms to the capacity of Soviet intelligence to recruit and sustain moles in highly sensitive positions in American and other Western intelligence services. In the early 1960's, the CIA uncovered, through the services of its own anonymous spy, a well organized complex of Soviet moles that included not only American but also French, German, Israeli, British, Swedish and NATO officers.

Most of these agents, according to their public admissions, were induced to work for the K.G.B. by financial rewards or sexual blackmail rather than an ideological sympathy with Communism. Some were enlisted under "false flag" arrangements in which, for example, former Nazis were recruited by a K.G.B. front that Pretended to be a secret Nazi conspiracy. They all continued spying for long periods of time, and, in some instances, such as in west Germany, provided the, K.G.B. not only with secrets but also with control of the intelligence apparatus itself. In the West German case, according to Tennant Bagley's analysis for the CIA, the moles were able to manipulate the careers of their fellow officers so as to promote and strategically place other K.G.B. moles. In this sense, the mole complex was self-perpetuating; and between 1960 and 1978 more than two dozen K.G.B. agents would be uncovered in the NATO alliance.

The unraveling of this complex did not occur through any ordinary security procedure but through an accident of history that could not reasonably be expected to reoccur in the intelligence war. This story began with a letter sent on April 1, 1958, to the American Ambassador in Switzerland, Henry J. Taylor. Taylor promptly turned the letter over to the CIA station chief in his embassy.

Tennant Bagley, one of the CIA officers who took control of the case, recalled to me that the letter was written in fluent German, and that the author, who claimed to be a high-ranking officer of a Communist intelligence service, refused to divulge his name or even nationality. The mysterious author suggested, according to Bagley's recollection of the case, that there were moles in Western intelligence who would betray him if he identified himself. He therefore proposed helping Western intelligence put "its own house in order," presumably by ferreting out the moles, before he would consider defecting to the West. He signed the letter "Heckenschiitze."

In his initial reports, sent to mailing addresses supplied by the CIA, "Heckenschiitze" rapidly identified seven Soviet spies. These included a British admiralty aide at the Portland Naval Base, named Harry Houghton, who had been supplying the KGB with secret information about United States nuclear submarines; Col. Israel Beer, an Israeli military historian who, in fact, was an Austrian who had emigrated to Israel 20 years earlier, pretended to be an Orthodox Jew and gradually won the confidence of Prime Minister David Ben-Gurion and other Israeli leaders; and Col. Stig Wennerstrom, the Swedish air attache in Washington, who was actually a general in the KGB

"Heckenschiitze" also provided a document that caused serious embarrassment at the British Secret Service - a purported list of 26 Polish officials compiled by British agents in Warsaw as potential targets for recruitment. This list, "Heckenschiitze" explained, had come from the KGB When Bagley and other CIA officers evaluated the list, the question arose: How could the KGB have obtained such a sensitive document unless it had a mole inside the British Secret Service?

The British intelligence asserted that the names could have been taken out of the Warsaw telephone directory. The denials were so heated that even the usually suspicious Angleton was prepared to believe that the anonymous mole was a dispatched agent attempting to sow discord between the American and British services.

Then, to everyone's astonishment, a researcher in the CIA's Eastern European Division discovered that British intelligence had sent essentially the same list to the CIA a year or so earlier. It now became clear to the CIA officers handling the case that the list had not been lifted from the Warsaw phone book, but from the secret files of British intelligence.

Allen Dulles, then the Director of Central Intelligence, presented this evidence to his British counterpart, and, after several months of investigating those who had access to the list, British intelligence traced the probable leak to the safe of George Blake. Blake, a Dutch-born intelligence officer, had rapidly risen in the ranks of the British Secret Service through a remarkable string of successful recruitments of Communist officers in Germany. Could such successes have been purposely provided by the KGB to enhance Blake's standing?

During his interrogation, Blake admitted that he had spied for the Soviet Union since 1952 and that he had passed virtually every important document the British Secret Service had in its files to the KGB

The depth of this KGB penetration into British intelligence stunned the CIA When the British diplomats Guy Burgess and Donald Maclean had defected to the Soviet Union in 1951, Harold (Kim) Philby, an officer in the British Secret Service, also had come under suspicion and, in the early 1950's, he had been effectively retired. The Philby case was now reopened. Then, after Blake's confession, Anthony Blunt, a former officer in the British security service (MI5), who had retired at the end of the war, was confronted by British interrogators and, in return for a grant of immunity, admitted that he had served as a Soviet mole. (In 1963, Philby defected to Moscow, thereby clearing up any doubts about his loyalties, and about his loyalties, and, then, Blake escaped from prison, and also went to Moscow.

Heckenshulttze next turned his attention to the West German intelligence service, the BND, headed by General Reinhard Gehlen, Hitler's former intelligence chief against the Russia. The BND worked closely with the CIA, which had created it.

"Heckenschiitze" reported in 1959 that he had been told by a high-ranking KGB officer that the BND had been thoroughly infiltrated by Soviet intelligence, and that many of its top officers had been blackmailed by the KGB into cooperating with it. He stated that of the six BND Officers who had visited CIA head quarters in Washington in 1956. and Allen Dulles with met, two were KGB Moles. This lead was specific enough to identify immediately one member of the group, Heinz Felfe.

A former Nazi Officer, Felfe, was the deputy chief of West German counterintelligence. Like Blake, Felfe had risen to his high Position through a series of "successes." West German security Police immediately placed Felfe under close surveillance, andcaught him transmitting secrets. The surveillance led to the arrest of a number of other moles in West German intelligence, including Hans Clemens, the man in charge, ironically enough, of the surveillance team inBonn. (Felfe, after being convicted of espionage, was traded to East Germany for a group of West German spies.)

A classified 1973 review of the memoirs of General Gehlen by Angleton's deputy, Raymond Rocca, termed the Felfe case a "crushing defeat" for the BND and concluded that the West German government had been "thoroughly penetrated".

"Heckenschiitze" finally decided to defect to the United States in 1960, after more than 30 months

service as an anonymous mole. His reason: The KGB had found out about certain documents that he had sent to the C-I-A. and asked his help in tracking down the leak. "Heckenschutze" now knew that there was a leak in American intelligence. On Christmas Day, he arrived with his wife at the American military mission in Berlin, and was met by a contingent Of CIA officers. He identified himself as Michael Goleniewski, the vice chairman Of Polish military intelligence. He further informed the Americans that he had hidden away a cache of documents in Warsaw.

When the CIA retrieved these documents, it found thousands of pages of polish and Soviet military bulletins containing United States military secrets that could only have come from high level sources in NATO and the United States Defense Department.

Goleniewski was given an Office in Washington, where he worked with his debriefing Officers attempting to "elaborate," as he put it, the various clues. He believed, for example, that he could pinpoint the leak in the CIA that had betrayed him. He revealed that Polish intelligence had known about a 1959 CIA plan to recruit a Polish diplomat in Switzerland .

The C. I.A. did not pursue the lead, according to Goleniewski. They spent, he claimed, "only a few hours" on this subject, and never brought it up again.

Before the debriefing could be completed, Goleniewski presented the CIA with still another surprise, He informed his case officers that "Goleniewski had merely been a cover name he had used in Polish intelligence. His real name was Grand Duke Aleksei Nicholaevich Romanoff. He further explained to the bewildered men from the CIA that his father, Czar Nicholas, had secretly escaped from Russia to Poland after the Bolsheviks had seized power, Goleniewski told his astonished audience that he was now heir to the Czar's fortune.

When news Of these disclosures reached Richard Helms, then Deputy Director for Plans, he realized that the CIA, had a potentially embarrassing Problem on its hands. Goleniewski had been the most productive agent in the entire history of the C-I-A-, revealing more than a dozen Soviet moles; the CIA, however, could not be put in the position of supporting his wild claim to the Czar's fortune. In 1964, the CIA severed its relations with its former spy.

Almost exactly one Year after Goleniewski had defected in Berlin, a KGB security Officer named Anatoli Golitsin defected from the Soviet Embassy in Helsinki, Finland, and was taken by the CIA- to Washington, where he was turned over to Angleton for questioning.

Even though he held a relatively low rank in the KGB, he said he had attended Moscow staff meetings in which the penetration of Western intelligence services was discussed. Like Goleniewski, he suggested that the KGB had infiltrated its moles in the C. I A., the British Secret Service, NATO, and French Intelligence, Indeed, much of the data that he furnished on this mole complex seemed to parallel that provided earlier by Goleniewski. Golitsin asserted additionally, however, that the KGB had managed to place Its agents in France in cabinet level positions close to de Gaulle.

This Golitsin leads focused suspicion on the French Deputy Prime Minister, but they were insufficient for French intelligence to take any action. Golitsin demanded an immediate payment of $1 million for his information, and received a substantial portion of it from the CIA

According to Philippe de Vosjoli, who had been the liaison between the CIA and French intelligence in Washington, and was brought in on the case, Golitsin insisted that at least six French intelligence officers were Soviet moles. After Golitsin provided clues that fit two colonels in French intelligence, both were allowed to from the service.

Golitsin further described a Plan that French intelligence had devised to spy on American nuclear-missile sites. The information that French spies collected in the United States in this operation would. according to Golitsin, be channeled to the KGB through its moles in French intelligence. It turned out that top French officers in Paris had ordered the espionage but De Vosjoli had never been informed about it. When De Vosjoli inquired about it after learning about it from Golitsin, he received orders from his superior's in Paris to now organize the spy networks in the United States that Golitsin had outlined. As far as de Vosjoli was concerned, this order demonstrated that French intelligence was being controlled by KGB moles and used to collect information for the Soviet Union, not France. He Protested the scheme, pointing out that France had no interest in spying on American missile sites' When his orders were not changed, he resigned from French intelligence and, after being informed that he would be assassinated if he returned to France, he went into hiding in the United States.

A large number of documents that Goleniewski had left for the CIA in the tree trunk in Warsaw contained information stolen from the NATO command. There was, for example, a top-secret June 1960, report on .intelligence objectives elaborated by the commanding staff of NATO. Goleniewski claimed that some of these documents had come from a French source. married to a Communist, Who had once been associated with the French war college.

In August 1963, French intelligence photographed a NATO official passing an attache case full of NATO documents to a Soviet Embassy official. He was Georges Paques, a former director of studies at the war college who had been an aide to nine French ministers. During his interrogation, he confessed that he had been spying for the Soviet union for some 20 years.

Then, in 1968, Hermann Ludke, a rear admiral in the west German Navy and the deputy chief of logistics for the NATO command, was identified by West German security police as a KGB_ SPY. Two weeks after his interrogation began, Admiral Ludke was found dead; he had been shot with a rifle. German officials declared his death an apparent suicide. The same day that Ludke was killed, Gen- Holt Wendland, the deputy director of west German intelligence, was found shot to death in his headquarters, another alleged suicide. Goleniewski claimed that he had pointed to Wendland as a key Soviet mole in West German intelligence under the code name "Organizer" as early as 1961- General Wendland had been the prime target of a West German security investigation. and had undergone interrogation prior to his death, He now was presumed to have been a Soviet mole for some 22 Years, according to a CIA officer who had been privy to the investigation. Within two weeks, four other German officials, who were reported to be suspects in the Ludke-Wendland cases, died violently, all alleged suicides.

Behind a ring of three barbed-wire electrified fences at Fort Meade, Md., is the headquarters of America's most secretive intelligence service the National Security Agency (NSA.). Even though it has more employees and a larger budget than any other American intelligence including the CIA. Even though its very existence had been classified a secret in the mid 1950s, such secrecy is considered necessary because it is responsible for protecting the security of the channels through which the leaders Of the United States Government, military forces and intelligence services communicate with one another. In most cases, the NSA designs the ciphers, encoding machines and protected lines through which the nation's most closely guarded secrets are transmitted . Any breach of this system can have disastrous consequences.

Aside from protecting the nation's secret communications, the NSA intercepts and deciphers the secrets of foreign governments. Such-signal intelligence includes intercepts of telephone and radio signals, telemetry from missiles and electrical impulses from radar and sonar. Vast quantities of information about the testing, capabilities and deployment of Soviet weaponry are derived from the NSA's electronic intelligence, or ELINT. Information about Soviet intentions comes from its code and cipher operations, which is known as Communications intelligence, or COMINT.

Despite its aura of secrecy, NSA has had multiple penetrations by Soviet intelligence. On July 22, 1963, Victor Norris Hamilton, a Syrian-born research analyst at NSA headquarters, turned up in Moscow and announced that he was defecting. He had been presumably an agent of the KGB In Moscow, he joined two other former NSA employees, Bernon F- Mitchell and William H Martin, who had defected 10 the Soviet Union three years earlier. While working as KGB moles at NSA head quarters, they had provided the Soviet Union with information about the technical capabilities and locations Of the super secret sensors that the NSA had employed against it, and also with data about the NSA's codes and breaking techniques.

One day after Hamilton defected from the NSA, Jack E. Dunlap, an employee of the NSA since 1958, was found dead of carbon monoxide poisoning - an apparent suicide. One month later, when Dunlap s wife found sealed packets of Government documents in the attic of their house, it was reported that he was a Soviet agent.

Col. Thomas Fox, the chief Of counterintelligence of the Defense Intelligence Agency at the time of the investigation, told me that Dunlap, a native of Bogalusa, La. had been recruited by the KGB. while employed at the NSA communications-interception base at Sinop, Turkey. He had met there Major General Garrison Coverdale the chief of staff of the NSA. General Coverdale then selected Dunlap to be his personal driver at NSA. headquarters at Fort Meade. General Coverdale further arranged for Dunlap to receive top-secret clearance and a position in the NSA.'s traffic-analysis division. Since the general's car had "no inspection" status, Dunlap could drive off the base with documents hidden in the car and then return without anyone knowing that the material had been removed from the base.

Moreover, Dunlap had other high-level connections in the NSA The Carroll Report, a secret Defense Department document (part of which I received through a Freedom of Information Act request) named after Gen. Joseph F. Carroll, who was asked to investigate the case, noted that Dunlap had helped a colonel at the NSA. base pilfer some "expendable items of Government property" from his office. From this incident, the report deduced, "Dunlap had already had experience in circumventing NSA. procedures under relatively high level tutelage." The implication was that he had expanded his access to secret files by offering to help officers appropriate furniture and other articles from their offices.

When General Coverdale left Fort Meade in August 1959, Dunlap was reassigned as a driver to the new NSA. chief of staff, General Watlington. By continuing his chauffeuring, Dunlap retained access to the "no inspection" vehicle necessary for smuggling documents on and off the base.

The Carroll Report makes it clear that Dunlap was interrogated by NSA investigators just before he died. According to Colonel Fox, the Defense Department investigating team did not establish any connection between Dunlap and the three NSA employees who fled to Moscow. Since four KGB. moles had been uncovered in the NSA., the agency found it necessary to change its secret codes, encoding machinery, security procedures and entire modus operandi.

While Dunlap was chauffeuring around the NSA chief of staff at Fort Meade, the KGB developed another mole at the pinnacle of American military intelligence Lieut. Col. William Henry Whalen. Colonel Whalen who was the intelligence advisor to the Army Chief of Staff. Since Colonel Whalen, as intelligence adviser, could demonstrate a "need to know," he had access to virtually all military planning and national intelligence estimates. In return for money, he regularly supplied secrets to his Soviet case officer over a three-year period , even after he had retired from the Army because of a physical disability. According to his subsequent indictment, the highly classified data sold to the KGB included "information pertaining to atomic weaponry, missiles, military plans for the defense of Europe, estimates of comparative military capabilities, military intelligence reports and analyses, information concerning the retaliation plans by the United States Strategic Air Command and information pertaining to troop movements. " He gave away, in short, a wide range of national secrets available to the Joint Chiefs of Staff. (Pleading guilty in 1966 to charges of conspiring with a Soviet agent to divulge national defense documents, Colonel Whalen was sentenced to 15 years in prison, and paroled after six years.)

Through the services of Dunlap and Whalen, the KGB succeeded, as Angleton put it, in "opening the window" on virtually all American intelligence-gathering activities in the Soviet bloc. Just as the CIA was able to ferret out KGB moles by tracing the documents that Goleniewski provided from Moscow to their source,, the KGB could presumably trace the military intelligence reports and analyses that Whalen provided to whatever traitors existed in the Soviet intelligence apparatus. During this period, 1958 to 1963, the KGB did in fact succeed in catching the CIA's two prize moles in Moscow, Peter Popov and Oleg Penkovsky. Both were executed.

Even in the light of these past Soviet successes in penetrating the NSA and Defense Department, there is considerable resistance in the intelligence community to confronting the possibility that the KGB has used the same techniques and resources to establish new and undetected moles in American intelligence. For one thing, there is little bureaucratic incentive for searching for moles: If the search is a failure, it will be viewed as a demoralizing witch hunt; if it is successful, it will completely undercut trust in the past work of the intelligence service. Just as the British Secret Service resisted the idea that it had been infiltrated by KGB moles even after it bad received the incriminating documents from Goleniewski, the FBI elected not to pursue evidence of a mole. For example, William C. Sullivan, Assistant Director of the FBI for Domestic Intelligence until 1971, claims that J. Edgar Hoover, the FBI Director, refused to allow him to move against what he was convinced was a Soviet mole in the FBI's New York office. In his autobiography, Sullivan describes how he discovered the leak and, unable to identify the mole, proposed transferring, one by one, all personnel out of the suspected section. Hoover replied, "Some smart newspaperman is bound to find out that we are transferring people out of the New York office," and flatly rejected the request. The source of the leak had not been removed from the office, or further identified, when Sullivan retired. Similarly, the CIA has relied on polygraph examinations to uncover moles, even though there is no empirical evidence that they work. In 1978, for example, a 23-year-old watch officer in the CIA named William Kampiles sold to the KGB atop-secret manual explaining the technical operations of the KH-11 satellite system that is used over the Soviet Union. When the CIA investigated, it discovered that there were at least 13 other missing KH-11 manuals. Kampiles had passed all his polygraphs.

The strategy denial is of course self-fulfilling. So long as a secret service denies it is possible to penetrate it, it is unlikely to find evidence of such penetration.