Code Warriors

by Stephen Budiansky

  But Perry’s strikingly upbeat assessment eleven years later showed that a fundamental change had occurred in the interim. The key finding of the Perry Committee was that the theoretical-mathematical advances in cryptanalysis that the Baker Panel and others had kept urging NSA to fund were at last on the verge of paying off:

  During World War II, the U.S. and the UK achieved spectacular success in cryptanalysis which had a profound impact on the execution of the war. We stand today on the threshold of a cryptanalytic success of comparable magnitude….No one can guarantee that we will “break” any specific machine of the new generation, but we do not see the problem as being more difficult—relatively speaking—than the one posed…thirty-seven years ago by ENIGMA.37

  NSA had somewhat halfheartedly agreed back in 1959 to set up an outside think tank, as the Baker Panel had recommended, that would bring in top mathematicians from the academic world for whom a regular career within the agency had always been an unattractive proposition. It was a far cry from the Los Alamos–scale project that had been proposed, and for a number of years—just as NSA had stonewalled other outside groups of experts such as SCAMP—the agency refused to send any real problems to them.38 But the institute, housed at Princeton University under the aegis of an existing military-academic cooperative research organization called the Institute for Defense Analyses (IDA), eventually carried out some particularly groundbreaking work integrating supercomputers into cryptanalysis, having obtained one of the very first all-solid-state supercomputers, the CDC 1604.

  In 1976, NSA retired the IBM Harvest system; the kindest thing anyone said about it was that it had done a “journeyman’s” job on the standard, tried-and-true massive data runs that were the mainstays of business-as-usual cryptanalysis. Others in the agency more frankly called Harvest a “white elephant” that never lived up to its promise and huge expense. One of the major recommendations of the Perry study was that NSA purchase one of Seymour Cray’s revolutionary supercomputers to assist its theoretical, long-term research in cryptanalysis.39 At a cost of $10 million, the Cray-1 was the fastest computer in the world, and would remain so for the next six years. Forty years later a hundred million iPhone owners would be walking around with the equivalent of a Cray-1 in their pockets, but in 1976 it was ten times faster than any other computer in the world, with an innovative “vector” architecture that represented a huge leap forward in cryptanalytic power in the seesawing race of codemakers versus codebreakers. Though standard histories of Cray Research would persist for decades in stating that the company’s first customer was Los Alamos National Laboratory, in fact it was NSA, which, continuing its support of the forefront of computer innovation that began with the founding of ERA in 1946, saw the new Cray-1 as the best hope for achieving its long-sought breakthrough.

  In 1979, the confluence of new techniques developed by the mathematicians at the IDA “Communications Research Division” at Princeton and the Cray-1’s greatly accelerated computing power produced “the height of American cryptologic Cold War success,” in the words of NSA historian Thomas Johnson.40 The cryptanalytic breakthrough against Soviet ciphers, which involved a fundamental enough mathematical result that NSA still refused to declassify any details more than three and a half decades later—likely because it also had implications for NSA’s ability to break digital encryption methods that would assume enduring importance in the Internet era—led to a flow of signals intelligence during the Soviet invasion of Afghanistan, which began on Christmas Day 1979.

  It was a triumph; it was also a last hurrah for the golden age of codebreaking that had begun before World War II with William Friedman’s founding of the Army Signal Intelligence Service in 1921. The shift to reliance on implanted electronic bugs, direction finding and ELINT, and the interception of microwave and satellite communication links was changing forever the nature of the game. By the 1970s the dwindling importance of conventional communication channels led NSA to shut down its last two major radio intercept posts in the United States, both dating from World War II—the Navy station at Two Rock Ranch in Petaluma, California, and the Army’s Vint Hill Station, outside Warrenton, Virginia. Inman pushed to accelerate the development of remotely operated, computer-based intercept receivers to take over the chore of scanning for and collecting the dwindling traffic that still passed on HF manual Morse and radio teleprinter circuits, while upgrading systems for collecting the growing streams of “bauded” signals transmitted by computer modems.41

  U.S. Navy ELINT satellites launched starting in the fall of 1976 were soon to supplant the need for the huge direction-finding bases; by continuously following the radar signals emitted by Soviet, Chinese, and other warships, they made it possible to track from space, in real time, the movements of individual naval units through their entire deployments, anywhere in the world.42 The subsequent mass migration of communications to the flow of computer data over uninterceptable fiber-optic cables in the 1980s was only the culmination of a trend long in the making, in which mathematics and the interception and decipherment of encrypted communications en route to their recipients was giving way to electronic and computer engineering and the infiltration of targeted computers and terminals at their source as the mainstay of the entire signals intelligence enterprise.

  —

  The Soviets had their last hurrah, too, appropriately involving the spies they had always excelled at recruiting and exploiting right in the very heart of the U.S. and British intelligence agencies. On January 14, 1980, an FBI wiretap on the Soviet embassy in Washington recorded this telephone conversation:

  First person: May I know who is calling?

  Caller: I would like not to use my name if it’s all right for the moment.

  First person: Hold on please. Sir?

  Caller: Yes, um.

  First person: Hold the line, please.

  Caller: All right.

  Second person: Ah, Vladimir Sorokin speaking. My name is Vladimir.

  Caller: Vladimir. Yes. Ah, I have, ah, I don’t like to talk on the telephone.

  Sorokin: I see.

  Caller: Ah, I have something I would like to discuss with you I think that would be very interesting to you.

  Sorokin: Uh-huh, uh-huh.

  Caller: Is there any way to do so, in, ah, confidence or in privacy?

  Sorokin: I see.

  Caller: I come from—I, I, I am in, with the United States government.

  Sorokin: Ah, huh, United States government. Maybe you can visit.

  Making a last-minute change in his arranged meeting time, the unidentified caller just managed to duck an FBI surveillance team outside the embassy the following afternoon. All the FBI got was a photograph of his back.43

  Five years later a KGB defector named Vitali Yurchenko, who had been the KGB’s duty officer at the embassy the day the caller had come in for that first meeting, identified him: he was Ronald Pelton, an NSA cryptanalyst and Russian linguist who had worked on the agency’s most sensitive collection projects. In April 1979 he had filed for bankruptcy and three months later resigned from NSA.

  Three days after learning of Yurchenko’s identification, the FBI tracked Pelton down: he was working as a boat and RV salesman in Annapolis, Maryland, and living with a new girlfriend in an apartment in downtown Washington, where they spent most of their time drinking and taking drugs. Fearing Pelton would flee the country, the FBI at one point had two hundred agents detailed to keeping him under twenty-four-hour surveillance. With no evidence to prove he had engaged in espionage, the FBI agent in charge of the case decided to confront him directly in the hopes of getting a confession. The gambit paid off. After listening to the recording of the telephone call to the Soviet embassy, Pelton offered a full account of his work for the Soviets. In exchange for $35,000 (he had asked for $400,000), he had arranged to meet with KGB officials at the Soviet embassy in Vienna on two occasions, submitting to lengthy interrogations. He told them about A Group’s success in breaking Soviet cipher machines
, U.S. SIGINT satellites that targeted microwave telephone links throughout the Soviet Union, the U.S. embassy listening post, and an extremely secret Navy-NSA project that had deployed a submarine to install a tap on an undersea cable used by the Soviet Pacific Fleet’s headquarters in Vladivostok for its operational communications. The Soviets responded in 1981 by making an across-the-board change in their military encryption systems, bombarding the U.S. embassy with microwave jamming signals, and dispatching a salvage vessel to retrieve the cable tap from the floor of the Sea of Okhotsk.44

  During Pelton’s trial in 1986, President Ronald Reagan personally called the publisher of the Washington Post, Katharine Graham, to urge the newspaper not to run a planned story about the cable tapping operation, named Ivy Bells, even threatening prosecution under Section 798 of the Espionage Act. Post editor Benjamin Bradlee finally agreed to delete some details about Ivy Bells from the story, even though he concluded that “the Russians already know what we kept out.”45

  Pelton, who did more damage to NSA’s work than any spy since William Weisband, was sentenced to three consecutive life sentences. The arrest of another Soviet agent during that “year of the spy,” as 1985 came to be known, offered another valedictory chapter to the conventional cryptologic battle in which the United States and the Soviets had been locked for four decades. John Walker’s long run as the linchpin of a spy ring that had provided the Soviets the monthly key lists of the KL-47 cipher machine abruptly came to an end when his very drunk but very vengeful ex-wife phoned the FBI and turned him in. For sixteen years, Walker and a fellow Navy cryptotechnician he recruited after his retirement from the Navy in 1976, Jerry Whitworth, made regular copies of the lists just before destroying them in accordance with routine security procedures. Walker would leave them at dead drops around the Washington area for collection by his KGB handlers. The Soviets even supplied him with an ingenious custom-made “rotor reader,” as the FBI termed it, a pocket-sized battery-powered device to recover the internal wiring patterns of new rotors for the KL-47: placing the rotor into its circular slot and turning a contact through each sequential position would cause a lamp on a miniature readout board to light showing which contact on the output side corresponded with each on the input side.46

  The value of Walker’s material was abundantly clear both in the considerable sums the Soviets paid—unlike the usual KGB chicken feed, Walker received more than $1 million, Whitworth $400,000—and the extraordinary lengths their handlers took in arranging countersurveillance measures for the drops. Convicted in 1986 of espionage and tax fraud, Whitworth was given a 365-year prison sentence; Walker died in prison in 2014, a year before he would have become eligible for parole, by then an aging relic of a conflict that for the rest of the world had ended twenty-five years earlier in one of the most astonishing, and peaceful, collapses of a totalitarian system in the history of nations.47

  *

  *In a 1967 case, Katz v. United States, the Supreme Court reversed its 1928 precedent in Olmstead and ruled that private electronic communications were covered by the protections of the Fourth Amendment against unreasonable search and seizure and that a warrant was generally required for a domestic wiretap.

  EPILOGUE

  The Collapse of the Wall, and a Verdict

  A number of cracks had exposed the fragility of the Soviet hold over its Eastern European empire, not least the courageous challenge to Communist rule by Poland’s Solidarity labor movement that began in August 1980. But most of all was the fact that a regime whose authority had always rested upon ideological certainty, the ruthless use of military force, and an all-pervading atmosphere of fear and secrecy was trapped by its own past. When Mikhail Gorbachev was chosen general secretary of the Soviet Communist Party in 1985, becoming the first reform-minded Soviet leader since Khrushchev and the first in decades who was not geriatric, ailing, or (like his immediate predecessor, Konstantin Chernenko) literally senile, his efforts to promote openness and a restructuring of the social and economic system only laid bare how little there was holding up the entire house of cards. In Hungary in 1956 and Czechoslovakia in 1968, Soviet leaders had sent tanks to end popular protests, but once it became clear that Gorbachev would not intervene to suppress the stirring “count​errevo​lutiona​ry unrest” in Eastern Europe, all the Soviets had left was bluff, and that could not last forever; any small shove could bring the whole edifice down.

  It came on November 9, 1989, when, facing weeks of growing antigovernment street protests, the hard-line East German Communist regime desperately tried to save itself with a concession to the protestors, announcing an easing of travel restrictions to the West. Within minutes, crowds of East Berliners surged to the Berlin Wall checkpoints; the guards, caught by surprise and lacking instructions, finally just opened the gates and let them through. In the end it was the spontaneous action of ordinary people, not the calculations of think-tank strategists, intelligence analysts, and Pentagon contingency planners, that brought the ceremonial—and real—end of the Cold War. East and West Berliners surged atop the wall that had symbolized the front line of the conflict, bearing hammers and chisels, and began knocking it to pieces while East German troops stood by and watched.1 By the end of the year the Marxist-Leninist regimes in Hungary, Bulgaria, Czechoslovakia, and Romania were ousted, joining Poland’s earlier escape from the Communist orbit. It was inevitable that the wave Gorbachev had unleashed would not spare the Soviet Union itself, which disintegrated two years later in a dissolution that brought an end to the one-party rule Lenin had so ruthlessly secured for his Bolsheviks seven decades earlier.

  Ronald Reagan had given it a decisive shove, too, refusing to accept the Cold War as a permanent condition the world had to live with and deploying his theatrical skills effectively in what was after all a war of global theater and public perception as much as one of military force. But it was in the broadest sense a victory of George Kennan’s long policy of containment, the patient belief that if the Soviets could be held in check and if the world did not blow itself up in the meanwhile, the weight of all the contradictions of Soviet power would cause it to fall in on itself. William Odom, NSA’s director during the last three years of the Reagan administration, wryly suggested to a reporter that the U.S. intelligence agencies hold a ceremony at CIA headquarters, declare victory, run down the flag from the flagpole out front, and dismantle the entire intelligence system and use the opportunity to start over and do it right this time. It was a sarcastic comment on the crazy jury-rigged American intelligence structure, with its perpetual internal bureaucratic warfare, tangled lines of authority, and wasteful inefficiency and duplication, but it probably also was an admission that even in helping to attain the victory of containment over Soviet Communism the intelligence agencies had often failed spectacularly at crucial moments, and had left in their wake an often sordid trail of transgressions against law, morality, decency, and basic American values.

  NSA’s most important Cold War achievement was its least visible contribution, but one that undergirded the entire American ability to hold off Soviet military power, and that was its ability to offer the minute-by-minute assurance that no Soviet tank regiment or warship could move and no nuclear-armed bomber or missile could take off without the president knowing about it. Before the deployment in the 1970s and 1980s of the first operational system of space-based infrared sensors to detect missile launches in real time, NSA’s global signals intelligence network was the mainstay of tactical warning. The reassurance that the U.S. nuclear force could not be caught in a surprise attack and destroyed on the ground greatly eased the nightmarish need to contemplate pushing the button first in a crisis. The equal reassurance NSA was able to provide the White House during the Suez and Cuban crises that the Soviets were backing away from their saber-rattling threats was of incalculable importance in averting a runaway escalation that was the greatest fear throughout the Cold War confrontation, the danger that even a minor flashpoint could accelerate through miscal
culation into an exchange of thermonuclear weapons in which millions would die before it could be brought under control, if it even could once it began.

  The other enduring triumph of U.S. signals intelligence in the Cold War was the technical systems NSA worked out, after much bureaucratic delay and confusion, to supply enemy radar tracking data lifted from intercepted signals directly to U.S. fighter aircraft; in the skies over Korea and Vietnam these systems repeatedly proved their worth by extending the warning time American pilots had of approaching MiGs and hugely increasing U.S.-to-enemy kill ratios. This was the genesis of a more comprehensive system of direct support to U.S. troops by NSA’s SIGINT operations centers that would reach its maturity in the wars in Iraq and Afghanistan.

  NSA’s worst failures in the Cold War were the result of its becoming the victim of its own success, and the resulting overreliance on SIGINT by political leaders, which led to two profound miscalculations that framed the American tragedy in Vietnam, the Tonkin Gulf incident and the Tet Offensive. But NSA was not blameless in these fiascoes, nor even in cases such as MacArthur’s disastrous decision to ignore the accurate and timely warnings that American cryptologists were able to provide of the impending Chinese intervention in Korea—because all were ultimately rooted in the great dysfunctional flaw that went back to the very start of the U.S. signals intelligence structure in the pre–World War II days and that NSA, both out of bureaucratic inertia and self-interest, never attempted to remedy. This was the fiction that signals intelligence was not intelligence but information, which NSA was not to analyze but merely pass on in its raw form for others to interpret. That, plus the classification rules that always treated SIGINT as more secret than other secrets, led to a tunnel vision in the way intercepts were handled; the absence of definitive analysis by an experienced intelligence officer who could see the whole picture and was trusted by those at the top to do so made it possible for generals and cabinet officials and presidents who were overconfident of their own judgment to do with SIGINT whatever they liked. It led equally to McNamara’s seizing on intercepts that confirmed his preconceived beliefs and MacArthur dismissing those that contradicted his. Never having the authority to perform its own analyses, NSA never explicitly developed an ability to do so, yet always did by default with varying degrees of professionalism and success.