Pragmatism and Principle: Intelligence Agencies and International Law

“The hell with international law,” Secretary of State Dean Acheson reportedly said during the Cuban Missile Crisis. “It’s just a series of precedents and decisions that have been made in the past.”[1] Although arising in a different context, Secretary Acheson’s observation also aptly summarizes the traditional attitude of states toward international law and intelligence gathering. Intelligence activities are a prime candidate for legal pragmatism—and especially its skepticism toward doctrine.[2] More than this, as Professor Ashley Deeks asserts in her recent article in this publication,[3] commentators have often contested international law’s precise remit in relation to at least some intelligence activities. Reviewing much of this same literature, I have argued that “the international community seems content with an artful ambiguity on the question.”[4] A more recent assessment points to international law’s “policy of silence” as the starting point in understanding the discipline’s relationship to intelligence gathering.[5]

This uncertainty should not, however, be overstated. It has arisen most often, and most credibly, in relation to true intelligence activity in the narrow sense of information collection. “Intelligence” is the “product resulting from the collection, processing, integration, evaluation, analysis, and interpretation of available information concerning foreign nations, hostile or potentially hostile forces or elements, or areas of actual or potential operations.”[6] International law is usually silent in relation to even the most notorious aspect of intelligence—peacetime spying or espionage, through the covert collection of information outside of an armed conflict. And indeed, in the absence of definitive, subject-matter specific law in the area, analysts have arrived at dramatically different conclusions about international law’s relationship with spying.

For instance, in his review, Professor A. John Radsan partitions the academic commentaries on the topic into three categories: those that regard espionage as illegal; those that see it as “not illegal”; and those that describe espionage as neither legal nor illegal.[7] A fourth approach abandons the debate of whether “intelligence gathering” or “espionage” is per se legal or illegal and instead subdivides the world of intelligence collection into constituent state acts. That is, it disregards a preoccupation with form (“intelligence collection”) and instead examines law governing specific conduct (for example, invasive surveillance, conduct of diplomats, interrogation, and so forth).[8]

But international law is much less agnostic in relation to the exercise by states of physical powers on the territories of other states, or in relation to human beings—conduct that when done secretively may fall within the scope of what is generally called “covert action.” Covert action designed to directly affect or influence people or the course of events often engages primordial rules of international law, particularly those of state sovereignty and the concomitant prohibition on intervention in the affairs of other states. International law has not singled out covert action and imposed redundant injunctions against activities that would already be unlawful if done overtly. But the absence of an extra-special admonishment against covert action does not amount to the same silence identified by commentators in discussing the lawfulness of spying. The surreptitious nature of a state’s conduct does not change its legal status, although it may change the politics surrounding it. There is, therefore, no principled basis to conclude that covert action per se falls into an area in which, to quote the famous S.S. Lotus case, states are permitted a “wide measure of discretion.”[9]

That is not to say that covert action lacks for justifications. One defense of covert action against legal formalism may rely on exceptionalism, urging the virtue of the cause prompting states to exercise covert powers.[10] In this respect, covert action may sometimes be assigned the same label as NATO’s 1999 Kosovo air campaign: illegal, but legitimate.[11] A second, possibly related justification may be simple realpolitik. States will not tarry over legal formalism when supreme security interests are at stake. However, both exceptionalism and realpolitik are unappealing justifications in a multipolar world in which many states may now be in a position to partake in potentially destabilizing covert actions. Moreover, their simple invocation risks abandoning legal formalism without superimposing workable policy guidelines to regulate covert action.

Confronted with this problem, Professor W. Michael Reisman and Judge James Baker have described the “myth system” of international law—that is, its doctrine—and juxtaposed it with the quite different state practice in the area of covert actions.[12] Reconciling the doctrinal myths of international law with this “operational code” means “that determinations of lawfulness in particular cases must . . . use a more comprehensive, consequentialist, and policy-sensitive approach.”[13]

In her article, Professor Deeks takes up this challenge by dividing international law applicable to intelligence activities into two “baskets”: a basket of rules that “is relatively detailed and focuses on protecting individuals”—such as international humanitarian and human rights law—and a basket of state-focused “rules such as respect for state sovereignty and territorial integrity, [that] regulates state-to-state activity.”[14]

She then advocates a hybridized, sliding-scale approach to international law and intelligence activity, urging more robust legal compliance in relation to individual-protecting rules and a policy closer to realpolitik in relation to state-centric rules.[15] Her motivation for doing so is largely pragmatic: If legal formalism wishes to “gain traction among states with robust intelligence capacities, it must allow states to adapt their international law interpretations to the special circumstances engendered by secret state activities, accepting that states require greater flexibility in interpreting some bodies of international law.”[16]

In this Essay, I offer a response to this model and review the rules of international law as they relate to intelligence activities. In Part I, I dissect the concept of “intelligence activities” and distinguish international law as applicable to spying from that relevant to covert actions. I urge that while international law is silent on spying per se, it is engaged by specific activities that rise to the level of intervention in a state’s sovereign affairs and that transgress the bar on the extraterritorial exercise of enforcement jurisdiction. There are, therefore, international norms that may readily be violated by at least some sorts of covert actions, above and beyond human rights principles that protect individuals. Ambiguity exists, but should not be overclaimed.

In Part II, I contemplate the virtues of tempering legal formalism in favor of a sliding scale in the area of international law and intelligence activities. While sympathetic to the necessity for pragmatism, I ask whether the sliding scale may result in the weakening of norms better served by being honored in the breach rather than abandoned in the name of realism.

I. International Law and “Intelligence Activity”

Noting the difficulty of definition, Professor Deeks describes intelligence activity as “both intelligence collection and covert activities undertaken by intelligence services, except for uses of force that would implicate Article 2(4) of the U.N. Charter, such as targeted killings overseas.”[17] This definition sweeps wide, and on its face includes everything that falls short of the threat or use of force against another state’s territorial integrity or political independence prohibited by Article 2(4), from open-source information collection to even covert assassinations. This is especially true if “covert action” is defined in accordance with U.S. law: “[A]n activity or activities of the United States Government to influence political, economic, or military conditions abroad, where it is intended that the role of the United States Government will not be apparent or acknowledged publicly.”[18]

A. A Typology of Intelligence

Most commentators focusing on international law and intelligence activities have defined their subject matter more narrowly than Professor Deeks, confining their topic to spying, “peacetime espionage,” or surreptitious intelligence collection.[19] Here, there is little doubt that international law contains no emphatic prohibitions. Where their discussions touch on covert actions, observers have often been much less equivocal about international law’s uncertainty.[20] Put another way, there is a regularly accepted legal dichotomy between spying and covert activity. On the spying side of that divide, international law is largely silent, promoting efforts to piece together rules from various collateral international law disciplines.[21] There may be instances, for example, where methods of intelligence collection trigger international human rights law,[22] although claims in this area should not be exaggerated.[23] On the covert action side, international law is much more certain, a matter I discuss in the next Section.

B. International Law and Covert Action

The frequent starting point for many discussions of international law and both espionage and covert action is the Lotus principle, a reference to the Permanent Court of International Justice’s 1927 judgment in Turkey v. France.[24] Commentators have pointed to this case in urging, “what is not prohibited is permitted in international law.”[25] And building on this doctrine, Commander Michael Adams has urged the existence of a security-preoccupied jus extra bellum—the “state’s right outside of war.”[26]

But permissiveness as the default position on the exercise of state power does not displace rules that do prohibit, or at least constrain, its exercise. This indeed was the position reached by the International Court of Justice (“ICJ”), confronted with the Lotus principle in the Nuclear Weapons Advisory Case.[27]

1. Sovereignty and Nonintervention

Some covert actions that have consequences on individuals obviously engage human rights principles (for example, detention, disappearances, and torture). Here, however, I focus on a less precise but equally obvious constraint on covert action: sovereignty. Sovereignty contains several ingredients, one of which is the principle of nonintervention—part of customary international law.[28] Professor Deeks correctly observes that the precise content of the broad principles such as sovereignty and nonintervention can be nebulous.[29] There are, however, at least some markers. For instance, in Nicaragua v. United States, the ICJ concluded that, at minimum, the principle of nonintervention

forbids all States or groups of States to intervene directly or indirectly in internal or external affairs of other States. A prohibited intervention must accordingly be one bearing on matters in which each State is permitted, by the principle of State sovereignty, to decide freely. One of these is the choice of a political, economic, social and cultural system, and the formulation of foreign policy.[30]

In the particular context of the Nicaragua matter, the ICJ concluded that prohibited interventions included “methods of coercion,” even when these fell short of use of force.[31] On a similar basis, some commentators have concluded that to constitute unlawful intervention, “the interference must be forcible or dictatorial, or otherwise coercive, in effect depriving the state intervened against of control over the matter in question. Interference pure and simple is not intervention.”[32]

Notably, coercion in this context likely means more than direct, physical compulsion. As one authority describes it, “[c]oercion in inter-State relations involves the government of one State compelling the government of another State to think or act in a certain way by applying various kinds of pressure, threats, intimidation or the use of force.”[33] These strictures would clearly implicate some forms of covert action. Thus, commentators have suggested coercive interference includes manipulation of “elections or of public opinion on the eve of elections, as when online news services are altered in favour of a particular party, false news is spread, or the online services of one party are shut off.”[34]

2. Enforcement Jurisdiction

Noninterference is not the only international rule engaged by covert actions. The Lotus decision itself acknowledges more general boundaries. Right after asserting its famous adage that “[r]estrictions upon the independence of States cannot . . . be presumed,” the court voiced an equally famous observation:

[T]he first and foremost restriction imposed by international law upon a State is that—failing the existence of a permissive rule to the contrary—it may not exercise its power in any form in the territory of another State. In this sense jurisdiction is certainly territorial; it cannot be exercised by a State outside its territory except by virtue of a permissive rule derived from international custom or from a convention.[35]

The exercise of state power is known as “enforcement jurisdiction,” and the prohibition on the imposition of nonconsensual enforcement jurisdiction extraterritorially, that is, on the territory of another state, remains a bedrock principle of international law: “[T]he legal regime applicable to extraterritorial enforcement is quite straightforward. Without the consent of the host State such conduct is absolutely unlawful because it violates that State’s right to respect for its territorial integrity.”[36]

Because enforcement jurisdiction rules do impose definite limitations on the powers states may exercise on the territory of other states, the legality of the covert action depends entirely on its nature. International law certainly precludes nonconsensual, extraterritorial conduct jure imperii—that is, involving the exercise of government functions.[37] And so, it reaches a state’s use of physical force on the territory of another state (such as an arrest and abduction).[38] A state agent entering a foreign territory in his or her official capacity without permission also transgresses this standard.[39] As Professor Simon Chesterman argues, the limitation on extraterritorial enforcement jurisdiction

would clearly prohibit unauthorized entry into territory; it would also extend to unauthorized use of territory, such as Italian claims that CIA agents abducted an Egyptian cleric in Milan in February 2003 to send him to Egypt for questioning regarding alleged terrorist activities. It would also cover the use of territorial airspace to transfer such persons as part of a programme of ‘extraordinary renditions.’[40]

The most famous exercise of covert extraterritorial enforcement jurisdiction was the Israeli abduction of Nazi war criminal Adolph Eichmann in 1960, when Mossad agents covertly snatched Eichmann from Argentina.[41] Argentina’s foreign minister protested, declaring the conduct “contrary to international norms,”[42] while Argentina’s ambassador to the United Nations called the kidnapping an infringement of Argentina’s sovereignty.[43] Argentina submitted a complaint to the UN Security Council, precipitating an unusual resolution from the Council.[44] That resolution declared that acts such as the kidnapping “affect the sovereignty of a Member State,” “cause international friction,” and may “endanger international peace and security.”[45] The Security Council further called on Israel to offer reparations.[46] Following negotiations, Argentina and Israel settled the matter, but also issued a joint communiqué acknowledging the role of Israeli nationals in the breach of Argentine sovereignty.[47]

Eichmann-style abduction constitutes an unequivocal exercise of state powers. Some commentators further assert that because of the limitations on enforcement jurisdiction, states are “also disentitled to carry out investigations in a foreign country, if it is their purpose to pursue and enforce its prerogative rights such as its criminal, administrative or fiscal jurisdiction.”[48] Examples include “gathering information in one State for enforcing revenue laws of another State.”[49] More generally, without consent, a state may not “send its police officers, even if they are in civilian clothes, into foreign States to investigate crimes or make enquiries affecting investigations in their own country. Nor can it allow spies or informers to operate abroad.”[50]

As already noted, there is considerable doubt as to the validity of the last statement concerning spying. There is also a view that noncoercive, peaceful investigations undertaken by one state on the territory of another involving the collection of (at least) information concerning the antitrust and tax activities of its expatriates comply with international law.[51] It stands to reason, however, that the international legality of that investigation becomes more doubtful where the territorial state’s laws are breached in the course of the investigation. As one commentator urges in discussing extraterritorial state action, “the local law should be used to determine whether the pertinent exercise of sovereignty can be viewed as a valid exercise of State authority.”[52]

C. International Law and the Cyber Headache

Until recently, there was an obvious territorial element to covert actions—and indeed, almost all intelligence activity —that eased the assessment of legality. For instance, an agent was either acting physically on the territory of a foreign state, or not.

The communications revolution, however, has changed the physical locus of at least some state action and has therefore created awkward questions for geocentric international law. Does the simple act of a state reaching out from a computer on its own territory to penetrate a server in the territory of another violate the server state’s sovereignty? Does this amount to the exercise of extraterritorial enforcement jurisdiction?

The answers to these questions are not obvious. Certainly, the consequence of the hack may color the assessment of its legality. Considerable analysis has focused on whether a cyber act visiting physical damage on the receiving state constitutes a use of force within the meaning of jus ad bellum rules.[53] More difficult are circumstances when the penetration falls short of physical destructiveness, but involves the more passive co-option of, for instance, foreign government communications networks to monitor communications or spread corrupted data.

Intrusiveness of this sort plausibly amounts to the exercise of state power on the territory of another state, raising sovereignty concerns. In 2007, a Canadian Federal Court judge concluded that intrusive surveillance (presumably involving electronic wiretaps) conducted by the Canadian Security Intelligence Service (“CSIS”) on the territory of another state without its consent would violate that state’s sovereignty.[54] A second judge then distinguished that holding where the intercept, while directed abroad, took place entirely from Canadian soil.[55] It was never entirely clear to this author from the limited public record in this case how an intrusive intercept of a foreign communication could be done within Canada without reaching out (electronically) and hacking communications overseas, in presumptive violation of some foreign law. More significantly, subsequent controversy stemmed from CSIS’s nonobservance of this Canadian territorial expectation. CSIS, in coordination with Canada’s signals-intelligence service, outsourced the intercept function to (unnamed) “Five Eyes” partner intelligence agencies, which include the U.S. National Security Agency.[56] Intrusive surveillance was not, therefore, confined to the territory of Canada and was instead conducted by foreign partners.[57] Moreover, this intrusive surveillance by Five Eyes members was not limited to those agencies’ own territories and involved intrusive surveillance in third-party states.[58] In effect, CSIS had outsourced conduct that the first judge had viewed as inconsistent with international law.

The question of whether invasive (but nondestructive) cyber penetration of this sort truly breaches international law was addressed, in part, in the Tallinn Manual on International Law Applicable to Cyber Warfare.[59] A private project involving international experts, the Manual constitutes the most comprehensive treatment of the topic to date; however, it does not resolve the doubt. In keeping with the discussion above, the Manual urges, “international law does not address espionage per se. Thus, a State’s responsibility for an act of cyber espionage conducted by an organ of the State in cyberspace is not [to] be engaged as a matter of international law unless particular aspects of the espionage violate specific international legal prohibitions.”[60] On the topic of cyber operations going beyond spying, the Manual notes:

            A cyber operation by a State directed against cyber infrastructure located in another State may violate the latter’s sovereignty. It certainly does so if it causes damage. The International Group of Experts could achieve no consensus as to whether the placement of malware that causes no physical damage (as with malware used to monitor activities) constitutes a violation of sovereignty.[61]

Nevertheless, it concluded that “intrusion into another State’s systems does not violate the non-intervention principle . . . even where such intrusion requires the breaching of protective virtual barriers” such as firewalls or the cracking of passwords.[62] The litmus test is the concept of coercion, discussed above.

However, the Manual does not address the supplemental question of whether remote intrusion onto the territory of another state through cyber means constitutes an unlawful exercise of enforcement jurisdiction. At the very least, applying the doctrine discussed above, it seems likely that a cyber intrusion that requires the manipulation of cyber assets in a foreign state (through hacking or otherwise) does constitute an exercise of extraterritorial state power. This is not like remote sensing involving passive sensors located outside the territory of the state. Instead, this involves the transmission of electrical impulses in a manner that changes (and does not simply observe) the status quo in a foreign state. While it is true that the physical intrusion is minimal, I am not aware of any authority demonstrating that the legality of enforcement jurisdiction depends on the scale of the physical presence.[63] Indeed, to the extent that hacking violates local law, the intrusion is probably better described as an “encroachment of high intensity,”[64] an assertion consistent with the observation above on the role of territorial law in determining the legitimacy of the foreign state’s extraterritorial conduct. This is especially the case where, as here, international treaties oblige states to prohibit cyber hacking.[65] This is not, in other words, an idiosyncratic local law.

II. Sliding Scale, Legitimacy, and Legality

From the discussion above, the state of international law in relation to peacetime intelligence activities might best be described as follows:

• Intelligence collection is not per se regulated by international law, although sufficiently intrusive collection can be tantamount to covert action.
• Covert action is regulated by international law to the extent it amounts to coercive interference into the affairs of another state or the nonconsensual exercise of state powers on the territory of another state.
• In both instances, the precise nature of the intelligence activity may trigger application of more specific international rules concerning, for instance, human rights or other specialized regimes implicated by the state conduct at issue.

Ambiguity occasionally exists in the precise application of these rules, but only on the margins. For instance, it is not entirely clear what state conduct constitutes the improper jure imperii. Does it include, for instance, simple investigations? Does it reach cyber intrusions? It seems safe to say, however, that the more kinetic or physical the state conduct and the more inconsistent with territorial state laws, the more likely it is to amount to a wrongful exercise of enforcement jurisdiction. And the more consequential the impact on the foreign state, the more likely it constitutes intervention violating that state’s sovereignty. While greater definitional precision would always be useful, these are workable standards on which any legal advisor adequately apprised of the facts should be able to give advice.

The take-home point is this: To the extent that commentators are inclined to treat intelligence activities as a unique area immunized from international law or subject to some special, more relaxed lex specialis, they exaggerate considerably. The residual question is, however, the one that animates Professor Deeks’s article: Namely, how should international lawyers respond to the reality that states do and will engage in intelligence activities, regardless of the niceties of international law? Building on the projects of other authors who have suggested their own criteria,[66] Professor Deeks’s solution is a “sliding scale” that retreats from legal formalism in an effort to graft principled policy constraints on intelligence activities.[67]

The resulting guidelines hinge on four variables: (1) risk of error and quantum of harm, (2) state or non-state target involved, (3) the specificity of the international rule applicable to the situation, and (4) covert action done in support of a goal for which other, overt activities are permissible (for example, election bribery undertaken to influence policies that could be influenced overtly through foreign assistance).[68] Applying these factors, Professor Deeks urges:

When engaged in intelligence activities that target actors not associated with governments, states should interpret strictly (in favor of the target) international rules that clearly address themselves to the type of harm the intelligence service is contemplating inflicting and that function to minimize the risk that a state will erroneously undertake a particular harmful activity against an individual. In contrast, when states undertake more traditional intelligence activities that primarily implicate the equities of other states, states should be permitted greater flexibility in interpreting relevant international law.[69]

The Deeks model has the significant virtue of incorporating international law into the intelligence activity calculus. It opens the door to a wide margin of appreciation for states where that law is ambiguous, but also where the consequences are less dire. It narrows that margin where the law is more precise, and especially where the conduct affects human rights. In this manner, it offers a way of prying open the door for law to an area of state conduct in which international law has figured modestly, if at all, in state decision making.

However, at its core, the sliding scale “anticipates and accepts gradations of interpretation of international law.”[70] A clear objection to the approach—as to any form of pragmatism—is that it does not operate to give international law primacy in any instance. Unquestionably, the subordination of international law is the way of the world: International law colors state discourse without governing the outcomes of state decisions, at least for matters of high politics. The residual question, however, is whether it should also be the way of the law. It is quite one thing to say that international law is sometimes ignored. It is another to say its content should vary according to a pragmatic calculus.

Professor Deeks’s bet is that an approximation of international law in intelligence activities, leavened by other considerations, is better than an indifference to it. This is a reasonable compromise of real utility to legal advisors fighting a rear-guard action against expediency. But to the extent these compromises are conflated as the rules themselves, this system risks moderating the (ideally, constraining) political risk that accompanies violations of international law. Legal formalism may not reflect the way things are done, but it is often the looking glass through which state conduct is evaluated. As Professor Nigel White argues, international legal rules may be weak compared to contingent preoccupations that drive state security behavior, “but given that the latter is just a short-hand term for power and self-interest, the formal laws remain as constraints, no matter how weak, on power.”[71]

Put another way, international law approximates a grammar of international relations. Like the grammar of any language, it does not dictate precisely what is said. But it does bind how that thing can be said. With grammar (at least in the English language), no central authority dictates its proper form. Instead, that form is decided organically through shared use. It is subject to change, sometimes even radical change, but it is almost always possible to say a particular usage is grammatical or not at any particular point. And those who use the language are then judged on their command of grammar, creating peer incentives toward conformity with generally-accepted usage.

The trouble with pragmatism as a tool for deciding the actual content of international law is that it loosens these “grammatical” constraints, unmooring international relations from any fixed (or at least slowly evolving) shared index of propriety. For its part, a formalist defense of international law in intelligence activities is not (just) a form of rule-bound inflexibility or naïve idealism. It also stems from a policy preoccupation: It is better to protect law, and accept that questions of expediency may deprioritize legality in the calculus conducted by states, than to “collapse[] any distinction between law and politics, between breach and compliance.”[72]

Professor Deeks’s guidelines are compelling policy, but in trying to bridge the gap between what Reisman and Baker call the “myth” and the actual “operational code” of international law, they are inspired by, but merely approximate, doctrine.[73] As such, they are simply a contingent choice. That makes them arbitrary: They are fully mutable as between states and governments. States would differ in the emphases they place on elements of the Deeks calculus—and on whether they accept those elements at all. And as for governments: They come and go, and some may be willing to place a heavier thumb on those variables, thus permitting a wider range of arbitrary state action. International lawyers should, however, always be able to speak clearly to the legality of this conduct, whatever the mood of any given administration.