Ethical Espionage

On October 7, as Hamas fighters roared into southern Israel from Gaza, bringing terror and death to anyone they encountered—Israeli soldiers, Bedouins, young people dancing and getting high together, kibbutzniks scooping up small children into desperate arms—I was sleeping in a comfortable hotel room in Georgia. All around me in the sultry darkness of a beautiful resort, many of the US intelligence community’s finest minds were also slumbering. We awoke with the expectation that we would be addressed by CIA director William Burns at the opening of the Cipher Brief’s annual Threat Conference, a yearly gathering of national security professionals from the private and public sectors, plus a few academics and journalists.

Burns didn’t board his flight from D.C. that day. The mood in the carpeted ballroom was somber. Aside from the shock and horror everyone felt at the attacks and the terrible anticipation of what would befall Gaza in response, there were dumbfounded silences as people asked one another versions of the same question. Israel’s legendary security services, Mossad and Shin Bet, were reputed to be the best of the best, the “gold standard,” as former NSA and CIA director Michael Hayden put it. Why hadn’t they known?

These are the moments, in the astonished silence after catastrophes, when most people suddenly become hyperaware of intelligence agencies in their all-too-human reality. For the most part we seem not to like to think about them much unless it’s through the grit-gray, jump-cut, shaky-cam world of spy movies and television. In America in particular, the attitude toward them has always been uncomfortable.

The intelligence services of the US and Britain were profoundly shaped by the chilling view of humanity that emerged during World War II. The bombed-out cities, the millions of deaths, and the horrors of the extermination camps resulted in a visceral sense of mankind’s potential for limitless brutality. And after the war, Stalin—armed with nuclear weapons, in increasing conflict with his former Western allies—seemed to present a new and even more terrifying threat.

In 1954 President Eisenhower commissioned a report on the need for covert operations in a world facing the prospect of nuclear war. Its author, Lieutenant General James Doolittle, wrote of the Soviet menace:

It is now clear that we are facing an implacable enemy whose avowed objective is world domination by whatever means and at whatever cost. There are no rules in such a game. Hitherto acceptable norms of human conduct do not apply.

These words have resonated through decades of intelligence activities. The British and American people, who have only glimpsed this secret world, have developed a strangely ambivalent relationship with it.

On the one hand, an enduring culture in literature and film has been built around the morally compromised intelligence officer as a darkly glamorous character—Ian Fleming’s near-psychopathic James Bond or John le Carré’s jaded Alec Leamas. This attitude has carried over into a fascination with real-life officers and spies who seem to resemble the romanticized fictional ones; the Cambridge Five, for example, have inspired reams of literature. On the other hand, there have been frequent calls in the US for the intelligence services to be disbanded because of their perceived lawlessness and immorality. Since the 1970s many of these calls have come from the left, but now they’re suddenly louder on the far right, most prominently from Donald Trump and his allies, who promote the idea that a lawless “deep state” is invisibly controlling the Washington puppets we imagine to be our leaders.

Calder Walton’s Spies: The Epic Intelligence War Between East and West covers the period from just after the Russian Revolution to what seems to be a fresh beginning of the cold war in the present. The pace is rapid, the style is often clipped, and minor characters are dispatched with unsentimental alacrity. In 1948 Harry Dexter White, the head of the International Monetary Fund, was accused of being a Communist. In the course of one brief sentence he resigns, testifies before the House Un-American Activities Committee, and promptly dies of a heart attack. And in the midst of deadly serious descriptions of covert operations, Walton will drop an odd detail, such as the three-hour shopping trip to Harrods that the legendary Soviet spy Oleg Penkovsky took in 1961 with a courier from MI6, the British foreign intelligence service, while his handler set about finding a tailored suit for Penkovsky on Jermyn Street.

These moments of unexpected bathos and mild incongruity lend the book a certain jauntiness, and this is where Walton begins to reveal his dark enjoyment of the spy genre. Cold war spying lends itself so easily to the register of the absurd that stories about it can be irresistible—the stakes are so unfathomably high, the lives and concerns of individual humans so comparatively minute.

Before the gradual opening of the Soviet archives at the end of the cold war, we had an extremely lopsided view of espionage and covert measures during those years; it was like hearing one side of a very heated argument. Other books have begun to rectify this imbalance: Tim Weiner’s The Folly and the Glory (2020) and Thomas Rid’s Active Measures: The Secret History of Disinformation and Political Warfare (2020), for example. But more and more material is becoming available, and Walton seems to have rushed to every archive, East and West, just as the archivist turned the key in the lock so that he could blow the dust off long-held secrets. His pages crackle with the electric thrill of discovery, especially of the physical traces of historical figures, from Churchill’s red initials on World War II decryptions to Stalin’s anomalous writing in green pencil (rather than his usual red or blue) in a letter to a senior intelligence agent bearing bad news that his source can “go fuck himself.” This is all satisfyingly tangible evidence to find in the ghostly realms of secret intelligence.

Moral ambivalence is unavoidable, though. It has always been there. The United States did not have a foreign intelligence agency until 1942, when the Office of Strategic Services—the CIA’s predecessor—was established by President Roosevelt. The OSS conducted both espionage and covert operations—or sabotage—during World War II. At the war’s end President Truman disbanded it, fearing that a peacetime nonmilitary intelligence service (one that, unlike the FBI, wasn’t under the jurisdiction of the Department of Justice) might turn into a secret police like those of authoritarian states.

Such fears were widespread. In 1946 The New York Times reported that the American public was deeply suspicious of the “black arts” of espionage and secret intelligence, as a result of repeated reports of the freewheeling activities of the OSS. On the other hand, largely thanks to the spy literature that had become extremely popular during World War I, intelligence work was considered immensely glamorous. In The Thirty-Nine Steps, published to great acclaim in 1915 and first made into a movie (directed by Alfred Hitchcock) in 1935, John Buchan created the spy Richard Hannay, who then returned in subsequent novels by Buchan. In this cultural atmosphere, Walton tells us, the OSS “attracted adventurers, the children of American dynasties, and movie stars.”

Walton writes that by 1947, with Stalin apparently harboring expansionist ambitions in Europe and beyond (he had already subdued the governments of Poland, Hungary, and Romania), Truman realized that “world affairs were ugly, and required spies.” What Truman didn’t know was that the KGB and the GRU (Soviet military intelligence) had gained enormous advantages over the West in their political subversion capabilities, not least because the Soviet bloc was impenetrable while America was an open book. One of the many fascinating discoveries made by Walton in the Russian archives is the fact that in 1948 Stalin attempted to interfere in Truman’s reelection campaign in favor of the pro-Soviet Henry Wallace.

Oblivious to much of the reality of spying, Truman had bizarrely idiosyncratic attitudes toward the intelligence agencies. In January 1946 he celebrated the establishment of the National Intelligence Authority (which was replaced by the National Security Council in 1947) and the position of director of central intelligence with what the historian Christopher Andrew has described in The Secret World: A History of Intelligence (2018) as “probably the most eccentric lunch in White House history.” Truman personally presented each guest “with a black cloak, black hat and wooden dagger.” He then made Admiral William Leahy, his representative on the NIA, present himself and “stuck a large black moustache on his upper lip.” Truman went on to announce that Rear Admiral Sidney Souers, the first DCI, “had been appointed ‘Director of Centralized Snooping.’”

We can’t be sure what inspired this levity, but it’s not unusual for people to use laughter to disguise deep discomfort. In any case, the National Security Act of 1947 finally established a peacetime foreign intelligence service, the Central Intelligence Agency, to work alongside the Federal Bureau of Investigation, which had been created some forty years earlier and was responsible for domestic intelligence gathering. The National Security Agency was established in 1952, with responsibility for communications intelligence; it was notoriously so secretive that people in government joked that the acronym stood for No Such Agency. Many of Truman’s reservations turned out to be well founded. For almost thirty years the intelligence services conducted themselves without any scrutiny, often in a lawless and amoral fashion that was bound to make the humor turn sour.

The turning point was 1975. After revelations of the CIA’s domestic spying on the antiwar movement, assassination attempts on foreign leaders, and other illegal activities, Idaho senator Frank Church chaired a committee that finally investigated what the intelligence agencies had been doing, and outraged figures on the left, from Gore Vidal to the young Bernie Sanders, called for them to be shut down. There were many legitimate reasons for their outrage. Walton concludes that American and British covert operations during the cold war, with few exceptions, ranged from the unsuccessful to the catastrophic and inhumane.

In the early 1950s, for example, CIA director Allen Dulles was responsible for the programs code-named Bluebird and Artichoke, which the journalist James Risen describes in The Last Honest Man: The CIA, the FBI, the Mafia, and the Kennedys—and One Senator’s Fight to Save Democracy (2023) as “brutal interrogation experiments on prisoners of war and other ‘expendables’ at black sites in Germany, Japan, and elsewhere.”¹ These programs were succeeded in 1953 by MK-ULTRA, which researched “mind control” and was notorious for its cruelty as well as for shockingly bad scientific practices. Experiments with drugs on often unwitting subjects in prisons and mental hospitals—in at least one experiment, Risen writes, they were kept on LSD for as long as seventy-seven consecutive days—along with the use of electric shocks and sensory deprivation to brainwash people or depattern their minds, were horrifically inhumane.

Walton tends to speak of the agencies’ worst episodes sotto voce. He sums up MK-ULTRA in a few words, as a program that “used electroshock and drugs like LSD on subjects, experiments that led to at least one death.”

His account of Operation Phoenix is even more laconic. He describes it as “a tragedy for the CIA” and “a counterinsurgency program to root out the Vietcong” that “became a torture and assassination program.” Under the program, which was initiated by the CIA and had access to CIA funds, at least 20,000 people were killed between 1968 and 1971. William Colby, who led Phoenix, described the people killed as members of the “Vietcong infrastructure,” but according to a congressional report, many of them were innocent civilians. In House hearings, K. Barton Osborn, who worked for Phoenix, testified that US personnel had participated in the murders and tortures. These crimes largely took place under President Nixon, with Henry Kissinger as his national security adviser—men whom Walton portrays as great minds of foreign policy. Some readers will raise eyebrows.

But most importantly, in Walton’s view, there was scarcely a US covert action that was a long-term strategic success, with the possible exception of intervention in the Soviet–Afghan War (a disastrous military fiasco for the Soviets) and perhaps support for the anti-Soviet Solidarity movement in Poland. The point of writing this history is a moral one: hindsight can help us do better. People may debate Walton’s judgment on particular episodes in the cold war, but thanks to his astoundingly deep research they will do so armed with invaluable new information.

The activities of the US intelligence services are now examined by Congress, revealed by whistleblowers, and reported by rigorous investigative journalists. Though much undoubtedly remains concealed from the public, the prospect of exposure and even prosecution presumably acts as some measure of constraint on immoral or lawless operations. By contrast, a Russian defector told Walton that “the only limitation for the Russian security service (FSB) is operational effectiveness—not legal, ethical, or moral considerations.”

For Walton, our responsibility is clear: we have to identify the kind of intelligence activities that will strengthen rather than weaken liberal democracies in a world threatened by resurgent authoritarianism. We may never live up to our highest ideals, but we at least should have clarity about what they are and accountability when we fall short of them. We need an open debate about the proper function of intelligence services at a time of new threats and rapid technological changes.

Though it hasn’t yet reached the public arena, a literature on the ethics of intelligence gathering has been burgeoning in the academic field of intelligence studies and now in philosophy departments as well. Cécile Fabre’s Spying Through a Glass Darkly: The Ethics of Espionage and Counter-Intelligence will undoubtedly open up new avenues of investigation so that others can branch out from the longer-standing philosophical concern with just war theory.

Fabre maintains that in some circumstances espionage is not just morally permissible but morally mandatory. No doubt this conclusion will be more surprising to philosophers than to members of the intelligence services. But she wants to think carefully about which circumstances justify what kinds of actions and to provide a clear account based on a coherent and consistent set of principles.

Fabre’s argument is rooted in claims that she finds intuitively persuasive about individual rights. This is characteristic of much political philosophy in the analytic tradition: writers such as John Rawls and Ronald Dworkin have presupposed a common liberal democratic culture as a basis for our intuitions about what is reasonable. For Fabre, these rights entail duties of assistance even to people who don’t live within our borders. Thus her analysis transcends nationalistic biases and notions of duty to the nation as a primary obligation.

There will be many counterarguments to her purely moral considerations. But she convincingly insists on the legitimacy of individuals’ right to secrecy “in respect of sensitive political, military, strategic, and economic information about their community,” so long as these secrets are necessary to “protect their security, democratic agency, and economic rights.” Correlatively, she also claims that espionage and counterintelligence activities are

morally justified as a means, but only as a means, to thwart violations of fundamental rights or risks thereof…subject to meeting the requirements of necessity, effectiveness, and proportionality.

After weighing with careful arguments what is permissible within these bounds, she concludes that morally justified intelligence and counterintelligence activities may include “deception, treason, manipulation, exploitation, blackmail, eavesdropping, and computer hacking,” but not mass surveillance.

Of course, these measures can only be morally permissible if they’re also prudent and effective. As John Sipher, who spent twenty-eight years in the CIA’s National Clandestine Service, told me:

We at CIA do not engage in blackmail—probably less for moral reasons than the fact that it doesn’t work. It is much more effective to get people who want to work with you, rather than those who hate you and are looking for any way to escape the situation.

A former senior British intelligence officer told me that MI6 considered blackmail both unethical and ineffective.

The kinds of judgments that have to be made on the basis of secret intelligence are always probabilistic and made under conditions of uncertainty. Many consequences are unpredictable. Fabre points out that this means the intelligence services have not only moral obligations but a duty to be as diligent as possible in ensuring that their reasoning is “undefeated by countervailing reasons.”

But our epistemic duties and moral duties can come into conflict. Procuring evidence can involve inflicting harms. Acting on the basis of intelligence can cause further (sometimes extremely grave) harms. What constitutes “reasonable belief” in the objective correctness of a judgment can be a highly vexed issue.

And the nature of sources in intelligence creates additional problems, since intelligence officers will have differing levels of confidence in HUMINT (human intelligence), which can be unreliable, or SIGINT (signals intelligence), which can be manipulated. The best we can hope for might be what philosophers call a “reflective equilibrium” (though Fabre doesn’t use the term), which involves particular judgments and general principles being brought into line with one another through discussion and deliberation, but which in this case would also require mutual comparisons of certainty levels based on past experience.

Walton provides examples that show how demanding this process can be in an environment where we’re being actively deceived. The US and British intelligence services have twice imagined that Russia was an ally while its intelligence operatives were in fact conducting espionage and implementing active measures against them. The first time was during World War II. Walton tells us that as soon as the Soviet Union became an ally of the West after the German invasion in 1941, the British Foreign Office’s strategy “was based on a gentlemanly but naive belief that allies do not spy on allies.”

At one point during the war Finnish intelligence officers got hold of a Soviet code book, which they sold to the OSS. When the US secretary of state, Edward Stettinius Jr., and the OSS director, William Donovan, discovered this, they ordered that it be returned. Meanwhile, the Soviets had infiltrated Allied governments, militaries, and intelligence agencies at the highest levels, including the Manhattan Project. The extent of what Stalin’s spies got away with is in retrospect quite astonishing.

But Walton emphasizes that something very similar happened during the war on terror. Putin’s Russia was treated by many as a trusted ally. The vast majority of intelligence resources, in both the UK and the US, were redirected from hostile foreign states to international terror networks; Russia offered information and cooperation in tracking them down. It gradually became evident that Russia’s “war on terror” was largely a pretext for the subjugation and devastation of Chechnya, and that the Russians had once again exploited our naiveté. Their joint counterterrorism efforts with the US served a counterintelligence function for them, as primarily a means of gathering information on America, and they yielded little actionable intelligence for the US.

The CIA, after decades of experience with the KGB, was especially skeptical about this cooperation. Several senior officers wrote an op-ed in The Washington Post in 2020 explaining that they never had faith in these efforts. The people in the agency responsible for relations with Russia would have told policymakers that “the Russian intelligence services view the CIA as the enemy, and they routinely see cooperation as just another means to undermine U.S. interests.”² But the warning went unheeded by many. And open debate, even among the national security community, would have been impossible when the highest levels of secrecy were deemed necessary.

This illustrates one of Fabre’s important points: accountability can never be fully reconciled with the need for secrecy. The public can’t know what is and isn’t being legitimately concealed from them. We have to rely on trust. Fabre can offer no abstract, theoretical solution to this problem of selective disclosure:

The solution lies in the proper vetting and ethical training of intelligence officers; in fostering constant awareness amongst citizens and officials of the dangers of a culture of excessive secrecy; and in normatively directed institutional design of intelligence oversight.

This sounds appealing, but as a realistic possibility in either Britain or the US today, it seems excessively optimistic.

Fabre is right to assume that this ethos would have to extend across every aspect of policymaking, intelligence, and oversight. The division of labor involved in intelligence is much less simple than intelligence training manuals imply. It was once thought that the functioning of the intelligence services could be described in terms of a cycle, comprised of planning, collection, processing, analysis, and dissemination. In the academic discipline of intelligence studies, this is considered unrealistic.

Peter Gill and Mark Phythian have expressed convincing skepticism about each stage of this process. They point out that “in practice, states do not invest in vast and expensive intelligence collection capabilities and then wait for policymakers to determine the targets of collection.”³ Collection and analysis, they point out, are overlapping processes that inform one another. Many agencies are “called upon to implement the policy responses arising from their own collection and analysis.” Another layer of complexity is added when we factor in oversight and accountability.

Gill and Phythian reject the traditional idea of an intelligence cycle in favor of David Omand’s suggestion to think of the intelligence process as an interactive network rather than a hierarchy in which information is collected and passed up from analysts to senior intelligence officers and ultimately policymakers. Omand, the author with Phythian of Principled Spying: The Ethics of Secret Intelligence (2018), is a former head of GCHQ (Britain’s signals intelligence organization), and his description is rooted in decades of experience working in this kind of network.

One consequence of characterizing intelligence in this way is the recognition that soundness of ethical judgment is absolutely indispensable at every level of the intelligence community. But if this recognition were widespread, would we have outsourced so much essential intelligence work to the private sector, where profit is the fundamental motive? The agencies hire huge companies (the most prominent being Leidos, Booz Allen Hamilton, CSRA, SAIC, and CACI International) to do both analysis and operations, as well as many cybersecurity companies and others run by people with niche specialties (for example, analysts of the deep web and experts on cryptocurrencies) that the general public has never heard of.⁴ All of these are working closely with our intelligence agencies and are subject to the concerns that Omand, Gill, and Phythian raise once we view intelligence gathering as an interactive network.

With a revolving door between the public and the private sectors, it’s all too easy to envisage the gradual replacement of ethical motivations with financial incentives. The Israeli Intelligence Corps’s famous Unit 8200, which is at the cutting edge of cyber warfare, decryption, and signals intelligence, trains its personnel in highly sensitive government work, for instance encryption skills. These personnel often have the aim of eventually doing much more lucrative work for private companies, whose encryption may in turn have to be decrypted by government agencies.

AI and other technological developments will continue to generate new practical and ethical challenges. At the end of his book Walton reflects on the ways in which intelligence work has changed, particularly as traditional espionage is increasingly replaced by the use of open-source data—freely available information as opposed to information from clandestine sources. Machine learning is becoming pivotal in this transition, employing familiar techniques such as natural language processing, facial recognition, pattern recognition, and predictive analysis. It’s clear that we will have to pay careful attention as AI technologies evolve. But we’re also left wondering whether these technologies are already making the old tradecraft obsolete, whether the long, glamorous film noir that was twentieth-century spying has finally come to an end.

The intelligence failures of October 7 shed light on some of these questions. Walton, like me, was at the Cipher Brief Threat Conference that day, as the intelligence community grappled with the fact that this wasn’t just an Israeli intelligence failure but an international one. It was thought at the time that no one had foreseen the attack. It seemed as though Hamas had defeated all the complex machinery of surveillance, on which governments around the world spend billions of dollars, by employing extremely strict operational security of the old kind, for example by restricting communications to word of mouth and a small set of hardwired phones.

There’s some truth to this account of Hamas’s low-tech invisibility. But The New York Times reported in November that the attack had in fact been foreseen.⁵ Three months earlier, in July, a veteran female analyst with Unit 8200 warned her superiors about a daylong training exercise conducted by Hamas. It closely followed a plan set out in a document known as Jericho Wall, versions of which had been collected since 2016. Intelligence officials who were aware of the plan had a fixed belief that Hamas didn’t have the necessary capabilities to carry out the attack it outlined. When the analyst raising the alarm was brushed off by a colonel, she insisted that Hamas was hoping to initiate a war and said, “I utterly refute that the scenario is imaginary.” Did those who dismissed her concerns show the greatest possible diligence in ensuring that their reasons were, in Fabre’s words, “undefeated by countervailing reasons”?

The Border Defense Corps, a unit made up entirely of young women, also picked up on suspicious activity leading up to the attacks, using the oldest surveillance equipment of all: their eyes and their collective judgment. They told Haaretz that their warnings fell on deaf ears because of ingrained sexism, a problem that has apparently led to deplorable working conditions and even attempted suicide in their ranks. On October 7 they were simply left to be slaughtered.

AI hasn’t yet replaced human judgment or absolved us of the responsibility to exercise it in close cooperation with others. Neither is the old tradecraft dead. But it’s not glamorous. This is not the cold war. Humanity has gone on creating a hellish world of hot wars. Tens of thousands of innocents are dying in Gaza, in Ukraine, and in the many invasions and wars around the world to which the media pay little attention. Our best hope for the future of our intelligence agencies has to be that liberal democracies will employ them to help restrain the brutality of a relentlessly cruel species.