The CIA and Signals Intelligence

Washington, DC, March 20, 2015 – For decades the Central Intelligence Agency has conducted a major signals intelligence (SIGINT) effort that often placed it in competition with other members of the Intelligence Community, according to a significant collection of declassified documentation posted today by the National Security Archive (www.nsarchive.org). As described in a previously Top-Secret multi-volume history of the CIA's role from 1947-1970 — obtained by the Archive through the Freedom of Information Act — the CIA regularly struggled with not only Soviet counterintelligence and international upheavals like the Iranian revolution but overlapping missions and domestic budgetary battles with the National Security Agency (NSA) and other entities during the height of the Cold War.

Among the CIA's successes described in the documents that make up today's posting was the creation of the RHYOLITE geosynchronous satellite program which allowed continuous coverage of missile telemetry and targets in Eurasia. Agency operatives were also able to tap into radio-telephone communications of Communist leaders as they rode in limousines around Moscow, to track Soviet missile launches from two secret stations inside the Shah's Iran, and to intercept Warsaw Pact communications from a tunnel dug under East Berlin.

These achievements were not without bureaucratic costs. The RHYOLITE program raised hackles at both the National Reconnaissance Office (NRO), which oversaw much of U.S. satellite intelligence activity, and the NSA, whose personnel initially found themselves cut out of the program. Overseas, the Soviet limo bugging ended after a news report disclosed it and may also have led to the execution of the Soviet agent who installed the listening devices. After the Shah fled Iran during the 1979 revolution, the founders of the Islamic Republic quickly seized the two sensitive US monitoring sites, handing a major loss to American intelligence.

These and other aspects of the CIA's long involvement with SIGINT are described in over forty documents obtained by Archive Senior Fellow Jeffrey Richelson through Freedom of Information Act requests, archival research, and other websites.

The documents include:

• The History of SIGINT in the Central Intelligence Agency, 1947-1970 (Document 16)
• Guidelines governing the CIA acquisition and retention, via COMINT, of information on U.S. persons, including the one prepared in response to President Obama's presidential directive on signals intelligence (Document 43)
• A note from the President's Foreign Intelligence Advisory Board (PFIAB) concerning the need to improve the CIA's ability to employ audio surveillance devices in pursuit of foreign intelligence (Document 6)
• Several items from different decades discussing the CIA-NSA relationship, and covering topics as diverse as the SIGINT provided by NSA to CIA, and CIA human intelligence support to NSA (Documents 27, 28, 30, 31, 34, 36, 38)
• Memos and an extract from an official concerning the CIA's RHYOLITE signals intelligence satellite program, including the ongoing conflict with the NSA and National Reconnaissance Office over the program (Documents 8, 10, 11, 13)
• A memo written by DCI Stansfield Turner concerning a discussion with Secretary of Defense Harold Brown about modifying a satellite to replace some of the capability lost with the closure of the CIA's TACKSMAN sites in Iran (Document 32)
• Memos prepared in response to the CIA's Family Jewels inquiries, including one indicating testing of a U-2 carried COMINT system (LONG SHAFT ) in the United States (Document 21)

* * * * *

The CIA and Signals Intelligence

By Jeffrey T. Richelson

While the National Security Agency (NSA) and its military components are at the center of U.S. signals intelligence (SIGINT) activities they are not the only components of the U.S. Government that conduct such operations.[1] Three Justice Department entities intercept communications for law enforcement and/or intelligence purposes — the Federal Bureau of Investigation (thru its Data Intercept Technology Unit), the Drug Enforcement Administration (via its Office of Investigative Technology), and the U.S. Marshals Service (specifically, its Technical Operations Group).[2]

But it is the Central Intelligence Agency that, for decades, has conducted a parallel — sometimes complementary, sometimes competitive — signals intelligence effort with NSA. Components of the history of CIA SIGINT activities include, but are not limited to, organizational issues, audio surveillance and cable tapping, ground stations, aerial collection, space-based SIGINT, the CIA-NSA relationship, the creation and operation of the Special Collection Service, and legal controversies and privacy issues.

Organization

CIA headquarters, Langley, Virginia.

The CIA's signals intelligence effort has, in one way, come full circle since its inception — inasmuch as the CIA's efforts were originally conducted by part of an office with a larger mission and today are carried out by another office with a broader mission. Thus, until the formation of the Deputy Directorate for Research in 1962 and the creation of the Office of ELINT (Document 17, Document 24) there was no central office charged with managing the agency's electronic intelligence effort. Instead, ELINT activities could be found in the Office of Communications, (Document 1) and several other agency components.

One of those components that continued in operation after the creation of the ELINT office was Division D (Document 2) — an element of the Plans directorate's Foreign Intelligence Staff — which had been established to serve as conduit for the transmission of communications intelligence from NSA to the CIA. Other components that would be involved in SIGINT operations included the Office of Special Activities (1962-1974) — which was responsible for agency U-2 missions, some of which engaged in ELINT collection (Document 4) — as well as two offices that developed satellite reconnaissance systems (the Office of Special Projects, established in 1965, and the Office of Development and Engineering, which replaced the special projects office in 1965 and was disestablished in 2012.)

OEL remained a component of the Directorate of Science & Technology (which replaced the Research directorate in August 1963) until February 1977, when it and Division D were merged to form the Office of SIGINT Operations (OSO). Then in August 1993, OSO (Document 33) was merged with the second Office of Special Projects that had been established in 1988 to handle measurement and signature intelligence projects - to form the Office of Technical Collection (OTC), which remains in existence as of today.

Audio Surveillance & The Berlin Tunnel

A Soviet officer inspects intercept equipment in a tunnel under East Berlin used by the CIA in the 1950s. (Photo - German Federal Archive)

Although much signals intelligence is gathered through remote collection, the CIA has also relied on an assortment of operations that gain direct access to a communications link or to a location where conversations of interest are taking place. A study on clandestine collection in Latin America during the 1960s reported that the CIA had managed to install audio devices in the homes of many key personnel and those devices had produced much of the CIA's intelligence on the region. During the 1970s, one target of CIA audio devices was Nguyen Van Thieu, President of South Vietnam. Presents given to Thieu by the CIA — including television sets and furniture — came equipped with audio devices. The agency also attempted to install devices in the office and living quarters of the South Vietnamese observer to the Paris Peace Talks.[3]

Several years before those operations, the President's Foreign Intelligence Advisory Board had examined CIA efforts with regard to development and use of audio surveillance devices, reporting (Document 6) to President Kennedy that "the limited capabilities of present equipment have precluded the [CIA] from successful audio surveillance penetration of intelligence countries within the Soviet bloc" and that "the full potentialities of audio surveillance for intelligence purposes have not been thoroughly explored as a major scientific undertaking." The board made several recommendations - including that the CIA and Defense Intelligence Agency jointly conduct an operational analysis of "existing and potential use of audio surveillance as an intelligence collection technique" and "estimate the results which might be expected, within defined dollar limitations." The President approved the board's recommendations.

Years earlier, a much more elaborate effort was conducted by the CIA when it tapped three Soviet and East German communications cables from a tunnel under the Soviet sector in Berlin — which became the subject of a Clandestine Services history (Document 15). The project, codenamed PBJOINTLY, commenced in 1952 and was planned by individuals from a number of CIA components — including the chief of Division D. Construction of the tunnel was completed in late February 1955. The operation ran for eleven months and eleven days until Soviet forces "discovered" the tunnel on April 21, 1956.

Information from the intercepted material was disseminated via a control system designated REGAL. The Clandestine Services history identified an assortment of political, military, scientific, and operational intelligence derived from the intercepts — including details of the Soviet program to implement the decisions of the 20^th Party Congress (including measures to suppress unrest among Soviet nuclear scientists), the doubling of Soviet bomber strength in Poland, the organization of the Soviet Baltic Fleet headquarters, the location of approximately one hundred Soviet air force installations in the USSR, and identification of several hundred individuals associated with the Soviet atomic energy program.

Soviet discovery of the tunnel appeared to result from unfortunate circumstances related to weather. But subsequently it was learned that George Blake, an officer with the British Secret Intelligence Service with knowledge of the operation, had been under Soviet control since 1952, when he was a prisoner in North Korea. However, the KGB, in order not to risk revealing Blake's treason, refrained from interfering with the operation until a plausible alternative excuse presented itself. Nor did they try to insert disinformation into the tapped communications cables due to concerns over the number of people who would learn of such an operation.[4]

Ground Stations

CIA intercept operations have also involved funding foreign ground stations as well as operating their own. Norway began operating SIGINT collection sites on behalf of the CIA in 1955. For the 1966-1967 fiscal year the CIA budgeted part of $104,000 to replace one of the principal ELINT receivers at a Norwegian-operated site at Kirkenes (which had a subsidiary outpost, designated METRO, at Korpfjell). The sites intercepted communications, telemetry, and other electronic signals. Included were data on launches out of the White Sea, on air-to-air and air-to-ground missile launches, and on Soviet practice firings from the Barents Sea. The operation also provided COMINT coverage.[5]

But the most important sites operated by the CIA during the Cold War were located in Iran, and known as TACKSMAN I and TACKSMAN II. The first site, a telemetry intercept station, was established in the late 1950s in an ancient hunting castle at Beshahr (not to be confused with Bushehr, location of a well-known nuclear reactor on the eastern shore of the Persian Gulf), on the southeastern corner of the Caspian Sea — with the objective of collecting signals from the Tyuratam test range, which the CIA believed would become a major Soviet test facility.[6]

In 1965 and 1966, the Office of ELINT sought to improve its ability to monitor missile tests emanating from Tyuratam and antimissile activity at Sary Shagan. It established a second telemetry intercept station in northeastern Iran at Kabkan, forty miles east of Mashad. TACKSMAN II was only 650 miles southwest of Tyuratam. As did the Beshahr site, it had a communications intercept capability to permit monitoring of test range communications. The sites could do what no other U.S. intercept sites could do — monitor the last moments of the firing of a missile's first stage, resulting in a greater degree of confidence in determining missile dimensions and throw weight. At their peak, the Iranian stations provided about 85 percent of the hard intelligence the U.S. acquired on the Soviet ICBM program.[7]

President Jimmy Carter considered the sites sufficiently important that he told his ambassador to Iran, William Sullivan, that intelligence cooperation between the CIA and Iran should continue despite the Shah's poor human rights record. But in January 1979, the Shah, in the wake of escalating protests and riots, fled the country — and the sites were eventually seized by the new government. In the wake of that loss, the issue of finding alternatives to the sites became a priority. Thus, in late October 1979, Director of Central Intelligence Stansfield Turner and Secretary of Defense Harold Brown discussed (Document 32) financial and technical issues related to adding a telemetry intercept capability to an existing SIGINT satellite system — while preserving the existing missions.[8]

Aerial SIGINT Collection

CIA ELINT operations — often directed at gathering signals of foreign radar systems — also involved multiple aerial platforms. The U-2 spy plane, while predominantly used for photography, could also be equipped with intercept equipment. One U-2 mission (Document 4), flown on January 16, 1961, took the aircraft over the South China Sea, Laos, South Vietnam, and North Vietnam. Its ELINT receivers gathered signals from four different bands and "verified previous estimates of the number of long range P-band radars" in North Vietnam, and "the South and Southwest Air Defense Districts" of China.

In August 1965, the Office of ELINT was involved (Document 12) in evaluating modifications to the intercept equipment on a drone, designated LONG ARM, to gather data on the SA-2 surface-to-air missile and radar system employed by North Vietnam air defenses near Hanoi.

A particularly extensive CIA electronic intelligence activity was the 'Quality ELINT' effort (Document 14) — which employed a variety of specially-equipped aircraft to gather data on radars operated by the Soviet Union, China, East Germany, and Cuba. Thus, between August 1963 and October 1966, a combination of Air Force RB-47H, C-97, and C-135 aircraft conducted seventeen missions carrying OEL's Power and Pattern Measurement System (PPMS). The PPMS gathered data on the Tall King, Spoon Rest, Knife Rest, Bar Lock, Fan Song, Hen House, and Knife Rest radars.

RHYOLITE

Pine Gap SIGINT site, Alice Springs, Australia. (Photo - Open Source)

Not long after becoming head of the Directorate of Science and Technology in August 1963, Albert Wheelon was reading a story in the New York Herald Tribune about the Syncom II satellite program. The article discussed what was then a revolutionary means of communications, first suggested by science and science fiction writer Arthur C. Clarke, that allowed communications far beyond the horizon. Signals were transmitted from a ground station to a satellite and then back down to another ground station.[9]

The Syncom satellites flew 22,300 miles above various points on the equator — in geostationary orbit. At that altitude and location, the satellites revolved around the earth at the same speed as the earth turned on its axis. In effect, they hovered over a single point on the equator. In addition, at their high altitude, about one-third of the earth was in view of each satellite. It occurred to Wheelon that it might be possible to employ such an approach to intercept telemetry signals emitted during Soviet missile tests — "the number one U.S. intelligence priority" at the time, according to a NSA historian (Document 33) — and relay them to a ground station.[10]

On September 23, an official of the Office of Special Activities noted (Document 8) that "the existence and success of Syncom II represents a significant intelligence collection opportunity." Wheelon assembled some key CIA officials to explore such ideas. Possibly as a result of that exploration, the same OSA official produced another memo (Document 10) that asserted "it is technically feasible to develop a synchronous satellite capable of the missions described" in a document reviewed by the author and that "it will press the state of the art." He also asserted that "it is important that we proceed with the basic study as a first step in a cost effectiveness review."[11]

The program to develop such a satellite, designated RHYOLITE, became another battle in the prolonged secret war between Wheelon and Brockway McMillan, the director of the National Reconnaissance Office (NRO) - who had something to say about the program since it would be conducted as part of Program B, the CIA component of the NRO, and funded through the reconnaissance office's budget. In a memo to Wheelon (Document 11, Document 39) in early November 1964, McMillan raised a number of objections to the effort, including that the Soviets could take defensive action to deny information to the U.S., and that the intercepts could be achieved by building "a covert, and relatively inexpensive, element of a satellite openly developed for other purposes" — which could include weather monitoring or communications relay.

McMillan's opposition continued through his departure from the NRO. In his exit memorandum of September 30, 1965 (Document 13), addressed to the secretary of defense, he wrote that "it is hard for me to believe" that the project "would justify so large an expense. In addition, he complained that "no alternative collection schemes were compared." Despite his opposition the RHYOLITE program (subsequently renamed AQUACADE) would produce four successful missions, beginning in June 1970.[12]

The CIA-NSA Relationship

Not only was RHYOLITE an issue between the CIA and NRO, it was also a source of acrimony between the CIA and NSA. According to an official NSA history (Document 33), "CIA cleared no one at NSA" with regard to RHYOLITE and that only changed in the summer of 1965 when General Marshall Carter, who had been serving as deputy director of central intelligence, became director of NSA and arranged "to clear a handful of NSA people and sent them to CIA to learn about the [RHYOLITE] program." The history goes on to state that "the road proved rocky in the extreme" since the "CIA wanted no NSA participation at all, and in the early months did a great deal to shut NSA out." Eventually, a truce was reached and CIA accepted NSA's suggestion that communications intelligence become a secondary mission of RHYOLITE satellites.

The RHYOLITE program was not the first or last time that the CIA-NSA relationship would be a matter of concern at either the high- or working-levels of DoD, CIA, or Congress. Indeed, the conflict between the U.S. central intelligence organization and its SIGINT collector(s) went back to World War II — before there was even a CIA or NSA. And once there was a CIA, it had to fight for access to COMINT (Document 16).

In February 1963, after both the CIA and NSA had long been established, the DoD deputy director for defense research and engineering wrote (Document 7) to Roswell Gilpatric, Robert McNamara's deputy, noting his desire to determine DCI John McCone's "feelings concerning the functioning and adequacy of management within NSA" and wondering whether Gilpatric had "any words of approbation" for NSA's director.

In June 1976, E. H. Knoche, the deputy director of central intelligence, wrote (Document 26) NSA director Lew Allen that it was his "strong belief that it is in our mutual interest to move rapidly toward a harmonious and lasting understanding between our two agencies in the SIGINT area." He added that toward that end he was commissioning a task force to determine the best ways to bring about such an understanding.

Two months later, the chairman of the task force reported (Document 27) to Sayre Stevens, head of the CIA's intelligence directorate, that he had visited several CIA analytical offices and heard several complaints. Objections included, but were not limited to, "NSA gists are not complete enough and sometimes miss the important points, to be found in the raw transcript;" "NSA analysis is incomplete, unreferenced and often indiscriminately mixes SIGINT and collateral;" and "NSA quite often refuses to provide transcripts or 'technical data.'" The chairman also reported that "almost everyone agreed" that it would be useful to station an NSA liaison team at CIA.

That same month, a memo (Document 28) to the CIA's assistant comptroller for requirements and evaluation examined the evolution of the CIA-NSA relationship. It noted that "in its early years NSA looked respectfully and appreciatively to CIA for guidance as to what it should collect and produce" but that "as time passed and its budget double, tripled, and quadrupled, NSA began to swell its corporate chest and develop a personality and style of its own." The memo noted that if the SIGINT task force concluded the Agency had an important role to play in the SIGINT field "it will require vigorous, imaginative and unified action ... especially at the top, to convince Congressional and other critics of the rightness of our cause."

Over the next thirteen months various aspects of the CIA's SIGINT activities were addressed by senior officials. In September 1976, DDCI Knoche wrote (Document 29) to Deputy Secretary of Defense Robert Ellsworth concerning CIA compliance with an executive order pertaining to SIGINT activities and noted the existence of "certain legal limitations and authorities unique to CIA." Around June 1977, DCI Stansfield Turner inquired (Document 30) whether any limitations could be placed on the foreign intelligence information provided to him by NSA. That fall, Turner wrote (Document 31) to the chairman of the House Committee on Appropriations, George H. Mahon, noting that an interagency SIGINT agreement had been signed between the director of NSA and the deputy director of central intelligence in January and that he had received a copy of the implementation plan produced by the NSA director and the DDCI. He went on to provide some specifics about the plan, including that it gave the NSA director "review and approval authority of the CIA SIGINT program."

In more recent years, a key aspect of the CIA-NSA relationship has centered around the changes in communications technology — the movement away from remote transmission (including via satellite or microwave) toward cable-carried communications. Thus, in 1999, the NSA's deputy director for operations noted (Document 36) that one reason for strengthening the NSA's strategic alliances was the increased importance of CIA HUMINT support to NSA operations — support which was undoubtedly a factor in many of the NSA successes described in the documents leaked by Edward Snowden.[13]  Similarly, a 2004 article (Document 38) in the CIA's in-house journal Studies in Intelligence noted that subsequent to the 9/11 attacks, NSA was "moving their strategic partnership ... into a new era of collaboration."

The Special Collection Service & Gamma Gupy

Headquarters of the joint CIA-NSA Special Collection Service, just north of Washington DC. (Photo from Cryptome.org)

The interagency SIGINT agreement noted by Turner in his letter (Document 31) to appropriations committee chairman George Mahon followed an investigation of NSA and CIA SIGINT activities orchestrated by a senior committee staffer. The resulting agreement was referred to in an official NSA history (Document 35) as the "Peace Treaty."

Redacted from the declassified version of that history was the discussion of one important consequence of that treaty — the merger of the CIA embassy-based eavesdropping efforts managed by the Office of SIGINT Operations (since it absorbed Division D in the 1977 merger with OEL) with those of the NSA. The result was a new joint operation designated the Special Collection Service (SCS), with headquarters separate from either the CIA or NSA. While the existence of the SCS is not classified, all details of its mission are. The SCS was a subject of a leaked briefing (Document 41) to the Pacific SIGINT Development conference in March 2011. By 2010 there were 96 SCS sites around the world, including 74 manned location, and 14 unmanned remote sites.[14]

One pre-SCS embassy-eavesdropping operation was designated GAMMA GUPY, and involved, in the late 1960s and early 1970s, intercepting the radio-telephone conversations of Soviet Politburo members - including General Secretary Leonid Brezhnev, President Nikolai Podgorny, and Premier Alexei Kosygin - as they drove around Moscow. Originally no attempt had been made to scramble or encipher the conversations. After columnist Jack Anderson disclosed the operation in 1971, the Soviets began enciphering their limousine telephone calls to plug leaks. Despite that effort, the United States was able to intercept and decipher a conversation between Brezhnev and Minister of Defense A.A. Grechko that took place shortly before the signing of the SALT I Treaty. Grechko assured Brezhnev that the heavy Soviet SS-19 missiles under construction could fit inside the launch tubes of lighter SS-11 missiles, making the missiles permissible under the SALT treaty. In general the intelligence was not earthshaking but did, according to a former intelligence official, provide "extremely valuable information on the personalities of top Soviet leaders."[15]

According to a document (Document 42) produced by the National Counterintelligence Executive (NCIX), a component of the Office of the Director of National Intelligence, the operation involved a CIA asset — a mechanic who worked on the limousines. The CIA, according to the NCIX, lost contact with the asset after the Jack Anderson article appeared in the Washington Post on the operation. The asset "was never heard from again and presumed killed."

The Family Jewels and Legal Issues

Cover page of Document 18.

Fourth Amendment protections with regard to electronic surveillance as well as general concerns over government eavesdropping have meant that CIA communications intelligence activities, if they involve U.S. persons, have run the risk of either falling into the category of being "potentially embarrassing," illegal, or both. Thus, CIA COMINT activities were among the topics represented in the CIA "Family Jewels" collection, investigated by the Justice Department, and the subject of a number of guidelines documents.[16]

COMINT-related Family Jewels memos concerned CIA electronic surveillance operations directed against drug trafficking activities in South America (Document 18), Division D intercept operations carried out at a CIA communications facility (Document 19), and Project LONG SHAFT — described in the memo (Document 21) simply as a NSA/CIA COMINT collection activity, and elsewhere as a package to be flown on U-2 aircraft and employed to intercept microwave communications of foreign targets, including those in China. However, with regard to testing of the system against U.S. communications, the memo stated that "in the broad sense this activity could be labeled illegal" although to the CIA's knowledge "nothing sensitive was picked up." Another memo (Document 20), foreshadowing the more recent, post-Snowden debate over bulk collection of telephone metadata, reported that the CIA had asked an AT&T official for copies of telephone call slips relating to US-China calls but that the general counsel's office did not believe that request violated the Communications Act "since eavesdropping was not involved."[17]

The CIA electronic surveillance activities directed against drug trafficking operations in South America became the subject of an extensive 1976 report (Document 25) by the Department of Justice. It reported on those activities, purported sources of authority for those operations, possible violations, applicable laws and statutes, and possible defenses, then summarized the possible violations and defenses. The final section includes the recommendation that "the inquiry be terminated in all respects for lack of prosecutive potential."

As a means of avoiding either embarrassment to the CIA or indictment of personnel, the agency has produced a number of guidelines concerning CIA SIGINT activities. Thus, the 2002 and 2006 "minimization procedures" (Document 37, Document 40) specified, inter alia, when information about the identities of United States persons could be retained or disseminated outside the CIA. The most recent guidelines (Document 43) were promulgated in response to President Obama's January 2014 directive, "U.S. Signals Intelligence." The guidelines cover general policy, collection, the use of SIGINT collected in bulk (distinguishing between acceptable and unacceptable uses), retention and access, dissemination, compliance, and responsibilities. It gives the director of the CIA the authority to "approve any exception to any provision of this regulation that is not required by the Constitution or a statute."

CIA SIGINT and Secrecy

The history of the CIA's signals intelligence activities, even from decades earlier, is still often shrouded in secrecy - with much of that history either completely classified or with only small portions of those activities appearing in declassified documents. Some of that history, such as the origins of the Office of ELINT (Document 5, Document 16, Document 24), the Berlin Tunnel operation (Document 15), or the Quality ELINT program (Document 14), can be found in some detail in declassified documents. But, despite the small discussion of the TACKSMAN sites in one memo (Document 32), a much more detailed account — an article in the fall 1991 issue of Studies in Intelligence, "TACKSMAN: A SIGINT Success Story" — has been denied in its entirety in response to Freedom of Information Act requests. And no information about the CHESTNUT sites established in China in the wake of the loss of TACKSMAN, while reported in the press, appears to have been declassified. Likewise, while memos (Documents 8, 9, 10, 11, and 13) and portions of an official NSA history (Document 34) add to the history of the RHYOLITE program, that is largely because of reporting on the program in books and articles, which permit an understanding of what the redacted histories and memos are actually referring to. Similarly, without media reporting the Special Collection Service would either not be known to exist or its activities would largely be a mystery.[18]