The Rise and Limits of Just Intelligence Theory

We have seen in the example of James Olson that conscientious IOs are often acutely aware of the ethical dilemmas inherent in HUMINT. Having served undercover for many years, Olson worried that some might see him “as a monster who spent his life doing all these nasty things.” Then, Olson shares his own view: “I did not see them as nasty. . . . I saw them as consistent with the just war theory.”Footnote ⁴ Thus, Olson recognizes the moral complexity of HUMINT and seeks solace and justification in making his actions consistent with the just war tradition (JWT). Its derivation, just intelligence theory (JIT), has indeed dominated intelligence ethics and has influenced both intelligence practice and lawmaking.Footnote ⁵ But can JIT meet the enormous expectations and hopes placed in it by ethically aware intelligence professionals?

JIT seeks to place limits on the whole intelligence cycle by devising concrete criteria for engaging in espionage (jus ad intelligentiam) and for the process of espionage (jus in intelligentia).Footnote ⁶ The jus ad intelligentiam criteria typically include just cause, legitimate authority, right intentions, necessity, and last resort, which are particularly relevant at the institutional and political levels. The jus in intelligentia principles of discrimination and proportionality aspire to provide guidance to individual IOs and agents.Footnote ⁷

Overall, as Kevin Macnish summarizes, JIT is a “helpful tool for thinking about the problems raised in intelligence collection.”Footnote ⁸ It offers useful contributions to intelligence ethics particularly at the levels of political and institutional decision-making, organizational management, accountability, and oversight. Beyond institutional ethics, JIT provides some helpful guidance to individual IOs and agents who find themselves in a very narrow set of circumstances, such as wartime and spying on a known imminent threat.Footnote ⁹ Outside of such circumstances, when applied to HUMINT at the level of its actual practice (as opposed to decision-making about HUMINT), JIT runs into conceptual and practical limits.Footnote ¹⁰ Furthermore, JIT does not offer robust answers to ethical challenges associated with the routine, daily conduct in the professional and personal life of a spy.Footnote ¹¹ Let us now explore that opaque life, which will in turn elucidate the subsequent discussion of JIT’s limits.

Being a Spy: Doing and Living HUMINT

A close study of the actual experience of HUMINT reveals that a truly robust and practical ethics must address not only the professional dimension—that is, doing HUMINT—but also the personal dimension of living HUMINT. The core of the professional dimension is obtaining information through people and from people. To obtain information through people, IOs engage in the practice of recruiting, developing, and handling agents. Beyond information collection, HUMINT is associated with channeling influence, infiltrating a target, maintaining cover, and mapping out one’s social relations. These aims are achieved through routinized practices of deception and manipulation, altering the circumstances so that the target rests their decisions on the wishes of the deceiver rather than on their own will. Thus, HUMINT elicits and abuses trust, and the taught and valued techniques include offering money, flattery, lying, emotional pretense to evoke sympathy and trust, threats, ideological persuasion, and sometimes bribery or blackmail. Furthermore, espionage often leads to, and relies on, lengthy personal relationships that, in turn, make practices of deception and manipulation even more ethically troubling.Footnote ¹² As Cold War CIA officer Drexel Godfrey put it, in HUMINT, “the modus operandi required . . . is the very antithesis of ethical interpersonal relationships.”Footnote ¹³

To illustrate the ethical challenges raised by the practice of HUMINT, take the actions of the ex-CIA IO Rolf Mowatt-Larssen, who had for years incentivized his agents to become traitors and rewarded them for betraying their family, friends, and nation. He describes operations in which agents were manipulated through false promises—never meant to be kept—into surrendering secrets considered vital to U.S. interests. These pledges were sacrificed to strategic imperatives or the practical demands of a given operation. Once the agents’ utility had expired, they were left “out in the cold,” destined in some cases for imprisonment, in others for death. “But what could I do?” Mowatt-Larssen reflects, recalling one such operation in which he paid an agent for his secrets and assured him of safe passage abroad, knowing full well that the agent would in fact be left behind. Days later, the agent was killed, precisely as anticipated. “Was I any better than the greedy spy I had sacrificed for our cause?” he asks, without fully answering the haunting question.Footnote ¹⁴

To complicate matters further, those participating in HUMINT do not clock out at the end of the workday. There is no division between a nine-to-five routine and a life after office hours for IOs and agents. Doing HUMINT also requires living HUMINT. And to live HUMINT goes well beyond handling agents, stealing secrets, influencing others, and infiltrating institutions. IOs and agents also need to actively develop and maintain their cover in virtually all interpersonal interactions outside of their professional life. Thus, they are required to deceive their family members, friends, and people in their wider social circles.Footnote ¹⁵ In living HUMINT, Mowatt-Larssen had deceived—and had to deceive—even his children. They could not know where their father worked.Footnote ¹⁶

The nuanced understanding of what it means to do and to live HUMINT derived from experience of actual IOs highlights the limits of institution-centered and legalistic ethical approaches to HUMINT. Instead of addressing the actual reality of being a spy, the dominant approaches emphasize institutions and legalistic procedures and neglect the component of individual moral responsibility. It is true that intelligence collection is an “institutional action” to the extent that individuals operate within certain institutional settings that empower and limit their actions. But it is individuals who operate, not the institution. Concrete, day-to-day moral decisions thus rest on that particular individual, not on an impersonal institution.Footnote ¹⁷ Therefore, it would seem a grave oversight for any robust ethics of HUMINT to give insufficient weight to the importance of individual moral responsibility. And yet, JIT is prone to making this omission due to its excessively prescriptive, checklist-style character and the major conceptual inadequacies at its core, which we shall presently discuss.

The Problem of Circularity in Applying JIT to HUMINT

The real praxis of doing and living HUMINT, as we have seen, is routine, and typically takes place in utterly ordinary circumstances. The routine nature of HUMINT makes ethical considerations associated with being an IO, or an agent, qualitatively different from the classical JWT designed for the extraordinary conditions of war. Even the increasingly “ordinary” continuous quasi-wars—such as counterterrorism and counterinsurgency operations—present conceptually different ethical challenges from those involved in HUMINT.Footnote ¹⁸ Accordingly, the most successful attempts to adapt JWT to intelligence consider revisions of the classical account and recognize that significant changes to the criteria are needed to reflect the specificity of intelligence.Footnote ¹⁹ Even such adaptations run into limits that are clearly revealed in the logical circularity of the just intelligence reasoning, and in the application of the criteria of discrimination and proportionality.

HUMINT weakens the central JWT distinction between knowledge (prerequisite) and the infliction of harm (action). Because action first requires knowledge, the otherwise helpful JWT criteria become circular when applied to HUMINT.Footnote ²⁰ This is unsurprising when we realize that at the core of a just war framework lies an ethical requirement that is often fulfilled via intelligence activity: The legitimacy of a target is established on the basis of collected intelligence. Yet, in routinized HUMINT, the very act of gathering information to determine whether a target is legitimate already raises crucial ethical questions. The circularity distorts the essential interplay of the interrelated JIT criteria; for instance, bending the last resort principle of JWT, since continuous intelligence collection is in effect the first resort.

Intelligence scholars have attempted to address the issue of circularity in diverse ways. Macnish, for instance, has proposed the use of “pre-existing evidence . . . available prior to the decision,”Footnote ²¹ while Ross Bellaby has suggested an intelligence “escalation ladder.” Neither approach, however, resolves the central paradox. Dependence on “preexisting evidence” merely reinforces the circularity by positing that intelligence—the very goal of HUMINT—must already exist before HUMINT can be undertaken; intelligence collection is required to determine whether to escalate intelligence activities.Footnote ²²

Other JIT proponents link intelligence to self-defense, the classical JWT justification for war against an existing or imminent attack and a generally accepted lawful and legitimate casus belli. Most intelligence collection, however, takes place continuously and in a preventive mode, often without any imminent threat. As noted, “Intelligence actions only rarely have the degree of emergency and urgency entailed in the concept of self-defence.”Footnote ²³ Outside of wartime, collection overwhelmingly focuses on potential threats with extended temporal horizons that may never materialize.Footnote ²⁴ JIT, however, understates the distinction between potential emerging threats—the focus of most intelligence activities—and imminent threats that might require self-defense.Footnote ²⁵ If taken literally, the self-defense justification would in fact prove highly prohibitive and ill-suited to doing HUMINT, to say nothing of living HUMINT.

Another attempt to address the circularity proposes new analogies to routine intelligence collection.Footnote ²⁶ One such analogy is sought in permanent nuclear deterrence. David Omand writes that the maintenance of a serious intelligence capability during peacetime helps maintain peace much as nuclear deterrence does.Footnote ²⁷ In suggesting as much, however, Omand omits the fundamental difference in his own analogy: nuclear deterrence might prevent war because nuclear weapons are not used. On the contrary, intelligence is being used—and used continuously, against both enemies and allies, in peace and in war. JIT does not seek to ethically regulate the possession of HUMINT capabilities, but rather their massive usage. Thus, the cumbersome analogy fails to overcome the fundamental conceptual limits in JIT’s applicability to doing and living HUMINT.

The Limits of Jus in Intelligentia

The limits become particularly apparent in the application of the in intelligentia principles of proportionality and discrimination. Omand describes “proportionality” in espionage as “keeping the ethical risks of intelligence operations . . . in line with the harm that the operations are intended to prevent, as part of the balancing act.”Footnote ²⁸ This balance sheet includes different variables: ends and means, the cost of action and inaction, gains and benefits, and the anticipated value of competing ends. Such a balance sheet is supposed to provide a reliable “moral accounting [which] allows us to balance the overall good effect of intelligence knowledge against some of the less desirable methods.”Footnote ²⁹

The criterion of proportionality (challenging to apply even in war) has a very limited applicability to the routine and continuous nature of HUMINT. The first fundamental problem with proportionality when applied to HUMINT lies in the inescapably disproportionate nature of most HUMINT. HUMINT is mostly conducted, as we have shown, against potential threats of a non-imminent sort and against allied and neutral states that rarely become actual adversaries. The inherent potentiality of most HUMINT defies the traditional logic of proportionality, however flawed that logic may be even in other settings. The second problem consists in the objectification of individual persons; their dignity becomes secondary to the benefits to the state. Thus, as Eric Stoddart cautions, proportionality carries an inherent risk. It attempts to objectively measure factors—such as harm to a person—that are fundamentally subjective and possess value beyond quantification.Footnote ³⁰ It is arguably no coincidence that one of the modern states that most seriously mistreated individual persons at home and abroad, the USSR, also had one of the most potent intelligence services. Harm to any individual, and to any group of individuals, was always considered perfectly proportional to the interests of the Soviet state.

The second criterion, discrimination, requires that innocent persons ought not to be harmed and obliges intelligence collectors to assess “the risk of collateral harm . . . [to] the lives of those who are not the intended targets of intelligence gathering.”Footnote ³¹ However, we once again find ourselves trapped in circular logic: Even in the most permissive of circumstances, such as spying on an adversarial IO, how else but through intelligence collection could one establish that a particular target is an IO? The fact that an IO needs to engage in HUMINT in order to collect information about whether a target is legitimate quickly becomes clear.

In echoes of JWT, some JIT theorists use the language of “guilt” for legitimate targets and “innocence” for those who cannot be targeted.Footnote ³² The language of relative guilt and innocence, however, is unsuitable for the actual reality of intelligence collection in general, and HUMINT in particular. Does merely possessing information make one “guilty” of something? Does a professional or family association with, say, a senior government official provide justification to be spied on? More problematically still, does one incur guilt by potentially possessing information or, alternatively, by knowing someone who might possess it? HUMINT routinely targets all of the above and, in most cases, the language of guilt and innocence is wholly inappropriate.

Consider, for instance, CIA’s HUMINT operation to locate the former al-Qaeda leader Osama bin Laden. CIA IOs engaged medical professionals, among others, who had conducted vaccination in the areas where bin Laden might have been hiding. The medical staff had not transgressed a norm that would make them deserving of being targeted and dragged into espionage.Footnote ³³ We see further that individuals routinely get involved in HUMINT, often without their knowledge. An IO’s family and friends, as well as potential targets, are all being deceived and manipulated on a daily basis, as Olson describes in his memoirs. His children did not know that their father was an IO.Footnote ³⁴ However, if a foreign intelligence service sought to establish whether Olson indeed was an IO and were to that effect to collect information about him, they would probably target his children—survey their routines, tap their phones, monitor their social media—as HUMINT-related activities often seek details about a target from third-party individuals who are associated with the potentially legitimate target. A strict application of the discrimination criterion would permit neither the use of medical staff nor engaging a target’s associates. This strict application would render HUMINT unrealizable.Footnote ³⁵

The Limits of Democracy-Centered Legalism

The final limit to a wider applicability of JIT lies in its prevalent use to justify acquiring intelligence on behalf of a (morally superior) liberal democracy against an “authoritarian” or “nondemocratic” state. Both scholars and practitioners often advocate for harsher HUMINT methods against citizens of states that fall outside of the liberal-democratic community. JIT presupposes that a similar HUMINT practice is acceptable specifically for preserving the national security of liberal democracies. However, its proponents simultaneously argue that the just intelligence way of thinking “should be universal.”Footnote ³⁶ Seumas Miller’s justification illustrates the broader argument. He writes that, with respect to authoritarian states, there are few if any constraints on intelligence gathering and analysis.Footnote ³⁷

On the general level of espionage, it is hard to see how JIT, with its clear preference for liberal democracies associated with the collective West, could be widely embraced and relied on by professionals in other political systems and in different parts of the world. In a revealing contradiction, JIT has a profound regard for liberal democracies on the basis of their professed respect for universal human rights and, simultaneously, diminishes the importance of these rights by justifying HUMINT against nondemocratic states.Footnote ³⁸ Olson offers a typical argument along these lines. He describes spying “against totalitarian, evil, oppressive, atheistic communism . . . we thought we were on the right side of that. And so we had no qualms about doing what we had to do for our country.”Footnote ³⁹

However, on the level of individual HUMINT, the democracy-centric argument implies that the personhood of individuals living outside of liberal democracies holds a lower value and, as such, might be subject to harsher HUMINT methods. This is highly concerning in part because HUMINT rarely targets the highest political leadership—the real or imagined “authoritarians”—but rather relatively “common” citizens who might have, or be able to mediate, access to desired information. It is therefore not only logically flawed but also ethically problematic to base the superiority of the liberal democracy on its respect for universal human rights, while denying the universality of these rights to individuals who happen to live under a different regime.

By arguing from the standpoint of a liberal democracy, JIT implicitly renders regime type a decisive moral variable, such that the ethics of spying are understood as a function of the regime it serves. This view might be defensible at an institutional level but is deeply problematic when considered through the prism of individual IOs. As history shows, even democracies can require their intelligence professionals to engage in morally wrongful actions. Furthermore, regimes of all types justify their actions by referring to their supposed righteousness. During the Cold War, governments would routinely justify actions with statements akin to “Action X protects communism, therefore it is good; action Z advances socialism, therefore it is just.” And it is certainly arguable that there was little inherently either good or just in socialism in Eastern Europe, let alone the Soviet Union. More generally, the structure of the argument is simply wrong, whether made by a socialist government or a democratic one. Moreover, there is no generic liberal democracy (just as there was no generic socialism), but only countless variations. There may well be a relatively good and just liberal democracy. But the assessment is too complex, too contextual, and too subjective to take the regime type as an a priori justification.

JIT’s focus on democratic legal procedures does not resolve the issue—certainly not from an individual IO’s perspective. Complying with the law does not automatically entail acting ethically. Attempts at serious ethical reflection on HUMINT based primarily on legal norms, combined with the nature of a regime, are incomplete and, often, hypocritical and self-serving. Moreover, a foreign intelligence service is obliged to operate within the boundaries of its national law in every country, regardless of the regime. There is nothing inherently ethical about a particular legal framework and, correspondingly, about lawful action, as law is an instrument of state power—the same power that controls and uses the intelligence services. Angela Gendron, a thoughtful proponent of JIT, remains skeptical about building an ethics of spying around legalism. As she acknowledges, the fact that “an intelligence service is legally mandated and subject to oversight does not mean that every objective it pursues, nor every means to do so, will be either lawful or morally justifiable.”Footnote ⁴⁰

Finally, there is the legal dimension of operating abroad, which is equally, if not more, problematic. HUMINT in foreign intelligence collection operations involves a “constant breaking of the rules . . . such operations are necessarily extra-legal and sometimes illegal.”Footnote ⁴¹ An exoneration by one country’s own law while breaking those of another renders an overemphasis on legal arguments either void or lacking in good faith.Footnote ⁴² It strips the law of any pretense for an objective and universal moral value. A strict application would logically create a situation where each state conducts lawful HUMINT according to that state’s legal framework, without any a priori relation of that framework to ethics.

The Limits of Rights-Based Revisions of JIT

An interesting alternative reinterpretation of JIT has been proposed by Cécile Fabre. She starts by posing two central and related questions. First, is an IO morally permitted to collect certain information? And second, what is the IO permitted, or even obliged, to do to obtain that information?Footnote ⁴³ She then proceeds to a general defense of espionage when it serves a “just foreign policy,”Footnote ⁴⁴ which she links, crucially, to the protection of fundamental rights (the definitions of which she leaves somewhat vague). The rights-based approach leads Fabre to argue that intelligence collection is ethically defensible as a means “to thwart violations of fundamental rights or risks thereof, in the context of foreign policy writ large, subject to meeting the requirements of necessity, effectiveness, and proportionality.”Footnote ⁴⁵ She thus partially embraces the JIT criteria and maintains that, first, espionage is legitimate to counter rights violations; and, second, that continuous collection is unjustified unless there is prior evidence of a threat to rights violations.

What seems at first to be a very modest and reasonable argument would in fact prohibit most intelligence collection activities, which in practice do not have any direct, and often even indirect, impact on the defense of fundamental rights—except in the broadest and vaguest of senses.Footnote ⁴⁶ If Fabre’s principles are applied strictly, most espionage as it is practiced would have to be terminated; if the principles are interpreted broadly and permissively, then they become void of meaning. Of course, one could defend spying using a more permissive interpretation of Fabre’s framework by arguing that intelligence about a potential war might possibly protect someone’s rights in an indeterminate future. Thus, depending on interpretation, Fabre’s proposal could either exclude most espionage or, on the contrary, permit almost any instance of it.

Fabre admits as much herself, acknowledging that “it is entirely possible—perhaps even likely—that much of what intelligence agencies have done and are currently doing is morally wrong, at the bar of the account I defend here.”Footnote ⁴⁷ Acknowledging the problem, Fabre eventually slips into what is in effect only a revised version of JIT. She argues that, generally, espionage-related practices such as deception, manipulation, and lying are justified when they are necessary, effective, and proportionate means of neutralizing threats to fundamental rights.

Fabre’s defense suffers from an underappreciation of what doing and living HUMINT actually means, and, correspondingly, what some of the most serious ethical issues are on the individual and interpersonal levels. Her framework is unrealistic except for very limited circumstances, such as spying on an evidently rogue regime (and, even then, with significant restrictions) and during wartime. She is aware of the prohibitive implications of her framework and concludes with a rather somber admission that the “question is what to do with those institutions.”Footnote ⁴⁸ Fabre’s question is extremely relevant—and one that I do not consider in this article. Instead, Fabre’s conclusion (without accepting her rights-centered premise) leads me to inquire what individuals in these institutions can and should do. To address that question, I have first demonstrated the limits of JIT in providing ethical guidance to individuals involved in HUMINT. Having revealed the limits of the current discussion, I propose a paradigm shift at two levels: I indicate how this paradigmatic shift can open up new possibilities for ethical reflection and, in doing so, enrich the now rather stuck discourse. Finally, I invite ethicists and practitioners coming from diverse ethical traditions to engage in this discussion, for my effort is but an initial turn toward fresh approaches outside of the exhausted JIT-inspired reasoning.

Institutionalized Individuals and Moral Outsourcing

A renewed focus on individual moral responsibility for (in)action, combined with institutional practices that constantly encourage and facilitate competing ethical inputs into decision-making, can contribute to filling the gaps in JIT, support conscientious IOs, and strengthen intelligence agencies. I hold that each IO is a moral agent with ultimate responsibility for their conduct and ought not become an institutionalized individual who outsources that responsibility to the collective ethos of an organization, government, or state.

The collective shortfall in moral reflection and action has been clearly—and darkly—articulated by Reinhold Niebuhr. Against the background of two world wars, Niebuhr convincingly argued that individual moral deficiencies are dramatically augmented at the level of collectives, such as nations and powerful state institutions. The action on behalf of, or within, a collective is precisely where Niebuhr identifies the greatest moral risks and the causes for much of the violence in societies and international relations. The individuals’ capacity for moral deliberation weakens as that individual becomes a part of the institution’s collective mentality. In every human group, writes Niebuhr, “there is less reason to guide and to check impulse . . . and therefore more unrestrained egoism than the individuals, who compose the group, reveal in their personal relationships.”Footnote ⁴⁹

A group—such as a highly secretive and powerful intelligence agency—can amplify egoistic impulses, erode conscientious resistance, and displace personal responsibility in favor of faceless collective responsibility and abstract notions of justice. Such moral corrosion is a natural risk when IOs ground their ethics primarily in institutional authority and legalistic justification. Such justifications reflect an insufficient grasp of the “brutal character of the behavior of all human collectives, and the power of self-interest and collective egoism in all intergroup relations.”Footnote ⁵⁰ They push “institutionalized individuals,” in Niebuhr’s phrase, to the invention of “romantic and moral interpretations of the real facts, preferring to obscure rather than reveal the true character of their collective behavior.”Footnote ⁵¹

In general, institutionalized individuals wield immense power through their respective institutions. Through the complex and seemingly all-powerful machinery of a state institution, even an otherwise powerless individual can cause immeasurable harm. In the intelligence world, risks associated with the exercise of that power are further accentuated by the necessarily limited oversight of individuals’ actions. Given that so many examples from HUMINT are classified, let us use cases from outside the HUMINT domain to illustrate with disturbing clarity what happens when institutionalized individuals act without a firm sense of ethical responsibility in organizational cultures that neither promote nor reward them for doing so. Though these cases are all from Nazi Germany, the implications extend directly to HUMINT, where ethical dilemmas are typically less distinct—and the risk of overlooking them all the more imminent. Furthermore, the readiness of individuals to cooperate in manifestly evil acts serves as a stark warning for IOs, who typically operate in far less morally binary environments—settings in which the internal moral warning signals may be considerably more muted; if individuals could cooperate in something so clearly wrong as the Nazi regime, think of how much easier it is to become complacent in less extreme circumstances. First let us consider the rank-and-file Reserve Police Battalion 101 that operated during World War II, five hundred ordinary men deployed by the Nazis to occupied Poland to contribute to the Final Solution. The institutionalization through the police turned the common civil servants from the Hamburg area into murderers who killed some eighty-three thousand civilians—men, women, and children—between 1942 and 1943. Christopher Browning reminds us that the “policemen faced choices, and most of them committed terrible deeds.” The institutional pressure in bureaucratized societies, continues Browning, sets behavioral norms and exerts enormous pressure on individuals to accept these norms. However, Browning observes that, despite the pressure, even in Battalion 101 “some refused to kill and others stopped killing.”Footnote ⁵² He concludes—as I do—that “human responsibility is ultimately an individual matter”Footnote ⁵³ that can make a profound difference, sometimes between life and death. For our purposes, we learn that even individuals at the lower levels of an institutional hierarchy can make a profound difference if they assume—and, more poignantly, as the examples show, fail to assume—personal moral responsibility.

The following two examples concern individuals with executive authority. The first, the story of Franz Stangl, the commander of the extermination camps in Sobibor and Treblinka, testifies to the failure of individual responsibility due to refusal to bear the costs of one’s decision, combined with unrestrained personal ambition. In Gitta Sereny’s classic account, Stangl repeatedly defended his decisions to remain in his institutional position based on a supposed lack of alternatives; if he refused to stay in the role, he thought he would have been imprisoned or killed, perhaps with his whole family. Thus, he continued in the position until around nine hundred thousand people had been murdered under his command. Only hours before his death from a heart attack did Stangl recognize the importance of assuming individual responsibility and accept its price. “I should have died. That was my guilt,” concludes Stangl.Footnote ⁵⁴ A “moral monster,” to use Sereny’s strong words, is not born but made through myriad daily decisions. It follows for Sereny that collective morality depends on an individual’s ability to make responsible ethical choices as they exercise power within their institution.Footnote ⁵⁵ Individuals’ failure to exercise responsible moral agency, particularly when their power is augmented through an institution, can have tragic consequences for peoples and nations. They can do much damage regardless of their respective institutional seniority.

Another such institutionalized individual, Adolf Eichmann, provides our next example. As recorded by Hannah Arendt, Eichmann demonstrates the detrimental effect of Stangl-like ambition combined with a refusal to engage in critical questioning of orders, and in moral reflection on one’s actions to implement these orders. Eichmann reposed his individual conscience in the faceless collective; he outsourced his ethical deliberation. When the Nazi state collapsed, Eichmann realized that henceforth he would have to “live a leaderless and difficult individual life, I would receive no directives from anybody, no orders and commands would any longer be issued to me.”Footnote ⁵⁶ Had he been an ethically responsible individual in his institutional role, he would have made different moral choices. Crucially for us, Eichmann’s bureaucratic complacency was far from unique; the sui generis historical context turned him into an Eichmann. As Arendt described, the “trouble with Eichmann was precisely that so many were like him . . . [in] that they were, and still are, terribly and terrifyingly normal,”Footnote ⁵⁷ a stark warning for IOs with their immense institutionalized power.

Similar to Stangl, Eichmann had been incapable of accepting the price of making an ethical decision. He asserted that “none of us had a conviction so deeply rooted that we could have taken upon ourselves a practically useless sacrifice for the sake of a higher moral meaning.”Footnote ⁵⁸ The decisions of several of Eichmann’s compatriots proved him wrong.Footnote ⁵⁹ In fact, a group of spies, among others in the German Resistance, sacrificed their very lives for a higher moral meaning.

Ethically Responsible Spies

Spies involved in different roles in the conspiracy against the Nazi regime manifested individual ethical responsibility in and through HUMINT. In a leadership role, the head of the Abwehr (the intelligence department of the German military), Admiral Wilhelm Canaris, strived to systematically weaken the Nazis—which he officially served—due to profound ethical disagreements. Thus, Canaris actively covered for conspirators, recruiting them to the Abwehr to allow them to travel freely, and also had close ties to the British intelligence and, likely, secretly conversed with Winston Churchill through clandestine channels and intermediaries throughout the war.Footnote ⁶⁰ Rank-and-file Abwehr intelligence officers and government functionaries Hans Oster and Hans von Dohnanyi shared Canaris’s concerns and also took the ethical responsibility to fight the Nazi regime through espionage. We “must do more than pray,” repeated Oster even as he was developing his traitorous network of agents.Footnote ⁶¹

Two of the key principled agents were Josef Müller and the Lutheran pastor Dietrich Bonhoeffer. With the consent of Canaris, Müller and Bonhoeffer were formally recruited as Abwehr agents to have a robust cover for their foreign travels that could withstand the scrutiny of the Gestapo, a competing Nazi intelligence service. Serving the conspiracy, Müller collected information on Poland and Russia, and liaised with the Vatican, to seek the pope’s formal blessing for tyrannicide.Footnote ⁶² The group concluded that “treason had become true patriotism and what was normally patriotism had become treason.”Footnote ⁶³ The secret police did not share their view and eventually arrested Bonhoeffer on the charge of espionage, together with Müller and their spymasters, including Oster, von Dohnanyi, and, later, Canaris himself. Eventually, they paid with their lives, except Müller who had escaped execution through the intervention of his high-level Nazi contact.

The conspiring spies’ motivations and actions are unintelligible outside of their strongly held ethical convictions that were incompatible with the actions of the Nazi government. Furthermore, their letters and writing make clear that peace served as the guiding telos driving their actions. Their experience further demonstrates that “it is possible for an individual to transcend the interests of himself and his group”Footnote ⁶⁴ without becoming a Promethean caricature. Firm embeddedness in an institution, combined with freedom from enslavement by that institution, is a testament to the power of individual responsibility within a collective, and to the impact of ethical actions through that collective. This dimension of individual moral responsibility within institutions remains a fundamental omission in current HUMINT scholarship and practice.

I argue that HUMINT ethics with a stronger emphasis on individual moral responsibility would not lead to more frequent treasonous behavior or to institutional paralysis caused by disobedience. The wartime context cited above was extreme, as were the actions of the conscientious spies. Absent extreme circumstances, the IOs’ individual ethical responsibility lies not in standing outside of their collective structures, but in becoming free and responsible within them. In this dialectical relationship, the individual as a free moral agent scrutinizes collective practices, and the institutionalized mechanisms in turn provide checks against individualistic excesses. The freedom (not) to act liberates the IO from uncritical submission to the agency, government, or state; at the same time, the institutional ethos shapes the individual and guards against destructive hyperindividualism. Depending on the context—and hence the preference for contextual rather than prescriptive ethics—the same ideals of individual morality may, at times, be best expressed through group loyalty and, at other times, need to be restrained for the collective good.

In routine intelligence settings, the effects of the proposed ethical shift would be gradual, long term, constructive, and—crucially—function across the hierarchy of an intelligence agency. On the institutional dimension, agencies would embed continuous processes of ethical deliberation into the organization’s fabric; neither operational nor strategic decisions would be taken without thorough, collaborative, and inclusive deliberation, informed by individuals’ ethical reflection. The leadership level should include ethical considerations as an important variable in their calculations. Furthermore, they ought to encourage subordinates to challenge decisions from their varied ethical perspectives, thus limiting biases associated with groupthink, and create an atmosphere that allows open and honest engagement on ethical grounds across the hierarchy. Furthermore, the rank and file ought to seek regular opportunities for constructive critical engagement with their peers and superiors. Finally, IOs at all levels must have the moral courage to refuse to take any action that they find morally unacceptable—and they must be ready to bear the consequences. As former CIA director John Brennan remembers, “I have told officers that I don’t want [them to do] something that they feel is inconsistent with their personal ethics and values.”Footnote ⁶⁵ Brennan’s words might or might not be truthful. But they articulate how it should be. The ability to assume responsibility, and the courage to suffer the consequences if necessary, makes the difference between the Stangls and Eichmanns of history, on the one hand, and the von Dohnanyis, Osters, Müllers, and Bonhoeffers, on the other.

A greater space for individual ethical responsibility will not weaken spy agencies but make them considerably stronger. Internally, the proposed approach will contribute to more sound decisions, thanks to a robust process that involves individuals coming from diverse ethical traditions. Operationally, an agency with ethically sound objectives and a solid reputation for due consideration of morality of individuals’ actions may attract more high-quality agents that are reliable and sustainable for a long-term collaboration, as shown by the respective examples of Müller and Bonhoeffer, who had refused to spy for the Gestapo but willingly joined the conspiracy. Publicly, an established, thorough ethical scrutiny may limit moral failures and associated backlashes. The improved reputation may, in turn, stimulate popular support and political willingness to approve robust funding and to grant wider powers to a well-functioning intelligence agency.

Espionage, Peace, and Responsible Statecraft

For ethical responsibility to be effective at both the individual and the institutional levels, it must be guided by some minimally shared direction. One direction can be found in peacemaking. A peace-focused teleological ethics of HUMINT brings at least three benefits. First, it provides a common telos, however narrow, for constructive discussions between different ethical traditions and diverse philosophical, religious, and political systems. The telos of peace—even if imperfect and contested—can offer a minimal common ground that may transcend particular viewpoints and ideologies. As the quest for some form of peace is not unique to liberal democracies, the nature of the regime becomes secondary in the moral reasoning.

Second, the imperfection of peace points to a world that is painted countless shades of gray, rather than existing in the binary of good and evil. In the grayness, demands for peacemaking are ever evolving. Accordingly, a peace-oriented teleological HUMINT ethics would naturally be less checklist-like and more contextual—characteristics that capture the dynamism of doing and living HUMINT.

Third, such a telos strengthens HUMINT by offering an alternative to the more restrictive tendencies of JIT, human rights–based approaches, and deontological traditions. In contrast, a peace-oriented framework would be more permissive, responding to the pressing need of responsible governments to sustain effective foreign intelligence services. At the same time, given the risk of abuse inherent in peace-centered teleological reasoning, any justification for expanding the remit of intelligence services requires careful and critical consideration.

An argument in favor of powerful intelligence agencies inevitably runs up against the Kantian deontological approaches to espionage, which argue that the practice of HUMINT is practically irreconcilable with the categorical imperative. Indeed, Kant himself went beyond abstract concepts and explicitly condemned spying. He suggested that espionage is a “diabolical art” that “exploits the dishonesty of others” and, as such, is “intrinsically despicable.”Footnote ⁶⁶ Thus, to interpret Kant in modern terms, the duty one owes to human beings is considered superior to what one owes to national security or interests of the state.Footnote ⁶⁷

If national interests are conceptualized within the broader notion of peace—not as a static binary opposite to war, but as a dynamic process conducive to human flourishing, authentic personal development, and the common good—then their pursuit and protection need not be a priori immoral, as Kant believed.Footnote ⁶⁸ As perpetual peace is never attained and the specter of war and violence is ever present, it follows that an imperfect peace must be pursued ceaselessly as a sine qua non for individual and societal development. This view recognizes negative peace as a prerequisite for sustainable contributions toward positive peace. Positive peacemaking, grounded in the protection of a negative peace, requires suitable tools. One such tool is HUMINT. Acknowledging HUMINT’s role in sustaining negative peace implies a vision of positive peace, even if never fully attainable. The objective of peacemaking, combined with the aspiration for genuine peace, can thus teleologically guide the choices of IOs and the operations of their agencies.

As a Kantian critique correctly emphasizes, HUMINT fundamentally thrives on, and reinforces, the hermeneutics of suspicion. The belief in states’ capacity for treachery justifies intelligence collection against foes and friends alike. Spying on temporarily allied states (for there are no eternal alliances) follows the Russian maxim favored by Ronald Reagan: “Doverai, no proverai,” or “Trust but verify.”Footnote ⁶⁹ Thus, states dishonestly (and naturally) claim that intelligence collection focuses on “acquiring the secret intelligence of hostile states.”Footnote ⁷⁰ In reality, much collection is focused on “potentially hostile states”—that is, on current partners that pose no immediate threat. But, as Niebuhr asserts, the duplicity of a state is a necessity for that state’s survival.Footnote ⁷¹ Such practice is consistent with a reality of peace that is fundamentally different from Kant’s image of a perpetual peace where, perhaps, espionage might be unnecessary.

Absent perpetual peace, however, the prevalence of espionage both testifies to the reality of negative peace and contributes to its maintenance. HUMINT indeed cannot deliver harmonious, risk-free relations between states. But such relations are an illusion in the reality of negative peace. Within that reality, HUMINT is a fundamental component of responsible statecraft. It is an expression of, and testament to, the human condition where fear and mistrust shape relations between states. While predicated on suspicion, HUMINT can also reduce uncertainty, dispel fear of the unknown, and establish channels of communication with the mistrusted. In this sense, HUMINT can be conceptualized as a tool of responsible statecraft that facilitates, or informs, nonviolent conflict prevention and resolution, guides policymakers, and thereby contributes to the making of peace. Although often practiced through morally troubling methods, HUMINT can nonetheless serve as an effective instrument of peacemaking.

Such a conceptualization of HUMINT is both realistic and idealistic. Realistic because it recognizes the need for espionage in a world full of actual and potential threats; and idealistic as it aspires to mitigate those threats peacefully. Rather than positioning HUMINT in opposition to peace-emphasizing ethical approaches, this teleological account turns HUMINT into an essential and powerful tool of responsible peaceful statecraft in a fundamentally dangerous world.

Reframing HUMINT Ethics: Power, Conscience, and Peace

Both academic and practical intelligence ethics require a sharper focus on individual moral responsibility. Such a focus can enrich an overinstitutionalized, deindividualized discourse that currently fails conscientious IOs, fosters the outsourcing of moral judgment from individuals to collectives, and thereby weakens intelligence agencies. While JIT offers important contributions, normative frameworks alone do not fully correspond to the dynamics of doing and living HUMINT. Therefore, HUMINT ethics would benefit from reframed approaches that are contextual.

The approach advanced here encourages individuals to assume ultimate responsibility for their (in)actions, while proposing institutional practices that promote critical ethical engagement across all levels of intelligence decision-making, from rank-and-file staff to agency leadership, from operational planning to strategic management. This emphasis reduces two principal risks: the sidelining of ethics in decision-making and the dangers of conformist groupthink. Future discussions about HUMINT ethics should accordingly foreground the role of potentially ethical individuals within potentially unethical organizations. A teleological ethics with peace at its core—one that places high value on individual moral responsibility within institutional settings—can produce intelligence services that are both more efficient and more ethical.

Developing such an ethics is a collective debt owed to the individual spies who dedicate their lives to protecting communities and who would benefit from more robust guidance in navigating the moral complexities of HUMINT.Footnote ⁷² For HUMINT is the sphere where power and conscience meet: Every intelligence officer embodies the power of the agency, yet must exercise conscience to use that power responsibly. The effort and desire to bring ethics more tangibly into the routine practice of HUMINT, repeatedly expressed by Olson, Mowatt-Larssen, and other well-meaning officers in liberal democracies—and, arguably, beyond—can find firmer grounding if the current approach is enriched through emphasis on individual responsibility directed toward peace.