The True Story of the Soviet Engineer Who Became a Spy and Saved the Federal Government $1 Billion

HUMINT remains the hardest, most dangerous, least scientific but also most powerful intelligence discipline of them all.

A KH-11 ‘Keyhole’ spy satellite costs, give or take, around $2 billion. In 1979, the US Air Force estimated that Soviet engineer Adolf Tolkachev had saved Uncle Sam about as much–in just his first year as a CIA agent. David Hoffman’s intriguing The Billion Dollar Spy (Doubleday, 2015) tells Tolkachev’s story through to his execution by the Soviets but in the process also raises some important issues about the intelligence struggles of the Cold War–and ones with growing resonance today. In particular, Tolkachev, codenamed CKSPHERE, epitomizes the risks and rewards of human intelligence, HUMINT, and in the process underscores why it is needed now more than ever.

Tolkachev, like so many agents, was motivated by a mix of ideological opposition to the system around him and what he called its “impassible, hypocritical demagoguery,” as well as the opportunity to feel like he was making a difference–and making his and his family’s life a little more comfortable in the process. Having decided that he wanted to be a spy, he demonstrated considerable nerve and raw, innate tradecraft by staking out the gas station where US diplomats fueled their cars and approaching them. After four approaches that were ignored or rebuffed (in fairness, the risk of a KGB set-up was considerable), the fifth time he tried to pass a letter for the CIA station chief, it reached its destination.

“Walk-ins,” potential agents who volunteer their services, are amongst the greatest banes and blessings in intelligence. The chance is good that they are a “dangle,” a fake would-be agent who would be used to gather intelligence and pass on disinformation, or else simply some maladjusted wannabe. And yet, and yet, often the best agents are walk-ins, and they come without the need for lengthy, risky operations to identify and recruit new sources. Canadian Sub-Lieutenant Jeffrey Delisle, for example, perhaps Moscow’s most damaging agent in recent years thanks to his position at the HMRCS Trinity intelligence fusion center, simply turned up at the embassy in Ottawa and asked to speak to a representative of the GRU, Russian military intelligence.

It took the CIA a year to decide he was both who he claimed to be and, considering his job at the top-secret Scientific Research Institute for Radio Engineering (NIIR), of potential value. That this took so long reflected not just a cultural hangover from the days of James Angleton (the CIA’s director of counter-intelligence from 194-75), who had a blinkered belief that every walk-in was a Russian plant, but also the practical difficulties of meeting Tolkachev. After all, while the Soviets could not adequately feed themselves or build a decent car, counter-intelligence was something about which they themselves were very serious and which they did very well. Moscow was no longer a “denied area” (one in which HUMINT was essentially impossible), but with the level of scrutiny of embassy staff and a constant environment of surveillance, it was still a very hard target.

Once Tolkachev had been recruited, then he proceeded to exceed his handlers’ dreams of productivity and even the Americans’ ability properly to process the material he was providing. NIIR was at the heart of a range of Soviet military R&D programs, from new radars to missiles, and he was in a position to plunder their secrets. If anything, he was more eager than his handlers, scorning their efforts to get him to be more cautious as he photographed thousands of pages of documents.

Even luck, meticulous planning and nerve only go so far, though. In 1985, he was arrested by the KGB, seized and stripped so that he did not even have a chance to take the “L-pill” suicide capsule he had badgered his handlers to provide. The next year, he was tried and shot, the victim not of his own enthusiastic activities but two Soviet spies within the USA: serving CIA counterintelligence officer Aldrich Ames and sacked CIA officer Edward Lee Howard.

Nonetheless, in his seven years of work, even though he never received the kind of payments he demanded and expected–and the penny-pinching of the Agency is in itself an interesting and depressing tale–Tolkachev gave the United States an appreciable extra edge in their Cold War competition with the Soviets. Indeed, this was an advantage which would outlast the USSR; as Hoffman describes in the epilogue to this book, it meant that the US fighters deployed during the 1991 Gulf War were optimized to fight and defeat the leading-edge Soviet warplanes in Saddam Hussein’s arsenal. This is an advantage that persists even now, as actual and potential adversaries from Serbia (Kosovo War, 1998-99) to Syria field the same aircraft.

The sad truth of the matter is that the great HUMINT assets of the Cold War tended to be walk-ins, not targeted recruits. There were exceptions, such as diplomat Alexander Ogorodnik and KGB officer Anatoly Filatov, but tellingly they were recruited in Bogota and Algiers respectively, not Moscow. The USSR itself remained a particularly difficult environment in which to operate, ever once, as Hoffman neatly describes, a new generation of much more can-do case officers began to rise to the challenges.

This helps explain why the USA came to depend so heavily on its panoply of technical intelligence means, from spy satellites to the electronic eavesdropping of the NSA. It played to America’s technological strengths, it obviated the need to operate within the Soviet security state, and it was safe. After the unexpected downing of the U-2 spy plane in 1960, even the skies were not open to US intelligence gathering, so they had to resort to orbit and ether, instead.

Such means as signals intelligence (SIGINT) and image intelligence (IMINT) undoubtedly have their virtues. They can pick up on everything from troop movements to construction, and were responsible for all kinds of revelations. Even trivial details can have significance: it was pictures of baseball diamonds in Angola in 1975, for example, that helped prove the presence of Cuban troops (Angolans at the time certainly were not the fanatics of “beisbol” the Cubans were). However, they also have great limitations. They may simply not be looking in the right direction; the other side may be able to deploy what the Russians call maskirovka, strategic concealment, to hide their actions and intent; or the clues may simply be missed because of poor analysis. Indeed, the best technical intelligence work generally works in conjunction with, or as a result of good HUMINT. The Soviets’ best SIGINT collection from the US embassy in Moscow, for example, was carried out thanks to listening devices planted by human agents.

Drawn from extensive work in the CIA’s archives as well as his usual extensive interviews, Hoffman’s fascinating account not only illustrates the inner workings of the Cold War intelligence struggle, it also reinforces one crucial point: human beings are still the ultimate intelligence tool. No spy satellite can choose to look in a different direction because it suspects there may be something interesting there. No electronic intercept can tell you if a statement was made deadpan or tongue in cheek. Telemetry may show you where a plane is now, it can’t tell you where it is going next. No spy ship can recruit other ships to spy for you too.

Of course there are always risks with human agents. They can be turned. They can be so seduced by the opportunity to make money or feel important that they start inventing what they cannot legitimately discover. They can get things wrong (after all, not all of them are primarily copying documents, as Tolkachev was, but they might be drivers or aides recounting conversations overheard, or sights seen). But when we look at HUMINT failures, they tend to be failures of analysis and management. Consider CURVEBALL, the Iraqi defector who so assiduously sold the line that Saddam Hussein had WMDs. The problem was not so much with his lies so much as that–despite ample reasons to mistrust his information–a political leadership so eagerly wanted to believe him.

If anything, the modern world offers not just so many new technical intelligence capabilities–hacking emails, new higher-resolution and multi-spectral spy satellites that can read a car license plate, the RQ-180 stealth spy drone–but also ways to get round it. Terrorists hide their chatter in seemingly-innocuous chat rooms and no spy satellite is likely to spot a bomb in a sports bag. The Iranians use motorcycle couriers precisely because they known any electronic communications can be intercepted; even if they are not decoded, the simple fact of traffic is a data point. The Russians managed to maintain complete communications secrecy before seizing Crimea and even now the holy grail of Western intelligence services is finding out what his real aims and intentions may be.

Yet a human agent in the rights place at the right time may be and often is the best source to foil a terror attack. If the Iranian communications are delivered on paper and by hand, then you need to get at whoever is writing or delivering them. Putin doesn’t tweet or email his innermost thoughts, and his closest allies are both loyal and closely watched – but one day a bodyguard, valet, cook or driver might be able and willing at least to relay snatches of what is said in the inner Kremlin circles where real policy is made. And, of course, as Tolkachev’s fate demonstrates, the most powerful counter to a human spy is another. In short, the reasons why brave men and women engaged in the kind of intelligence war Hoffman describes, ones in which the risks are of suicide or the firing squad, still apply, and HUMINT remains the hardest, most dangerous, least scientific but also most powerful intelligence discipline of them all.

(Image via vagant/