The Countermeasures of Others: When Can States Collaborate in the Taking of Countermeasures?

Abstract

In the last few years, states have advanced various proposals for cooperation in the use of countermeasures. In this Article, we ask whether, and if so under what conditions, states may lawfully collaborate in the taking of countermeasures against other states. We distinguish five different types of collaboration: (1) independent but coordinated action; (2) secondment; (3) joint action; (4) aid and assistance; and (5) what we term “proxy countermeasures”—the idea of taking a measure at the request and on behalf of another state. We consider the permissibility of each, both where the acting state is itself entitled to resort to countermeasures and where it is not. We also draw attention to certain legal and policy considerations relating to, and to plausible avenues for, the development of international law.

1. Introduction

States collaborate with each other all the time. They train each other's officials, fund projects, sell weapons, and undertake joint operations. States forge and act within alliances, sometimes informal, sometimes institutionalized. This collaboration is regulated by international law. Thus, in certain circumstances, states will bear joint responsibility where their collaboration entails the commission of an internationally wrongful act. Various rules of complicity give rise to responsibility where states facilitate the wrongful acts of another. The doctrinal aspects of this regulation have received extensive attention in recent international legal scholarship.¹

States regularly collaborate, too, in the adoption of sanctions or restrictive measures against perceived violators of erga omnes and erga omnes partes obligations. And recently, states have begun to raise the possibility of collaboration in the taking of countermeasures.² For instance, there has been extensive argument about the legality of a proposal to transfer seized Russian assets to Ukraine as a countermeasure in response to Russia's invasion of Ukraine.³ This may involve a number of allies acting in collaboration. Relatedly, in recent statements on the application of international law to cyberspace, states have stressed the importance of alliances in that context,⁴ putting forward views that states may act together in responding to malicious cyberoperations, including by means of countermeasures. In particular, these statements have proposed that collaboration in the taking of countermeasures may take place through support,⁵ collective action,⁶ or where one state acts on behalf of another.⁷ More widely, states may seek to act jointly as a regional bloc in response to some wrongful act which directly injures one of them, a situation that is likely to increase in the light of growing multipolarity and the inadequacies of the UN's collective security system. For example, in December of 2023, the EU's Anti-coercion instrument came into force, which includes the possibility of resorting to countermeasures.⁸

For the most part, collaboration in the taking of countermeasures has not been treated distinctly or drawn out precisely, either in the practice or views of states, or in the related scholarly commentary. This may be due to a variety of factors which, we think, have tended to obscure the question. First, it is possible that cooperation in the taking of countermeasures does occur in practice: injured states may have sought help from allies, or allies may have offered it, in responding to violations of their rights. But this practice may not be public. Second, while in many instances states impose sanctions and other restrictive measures in response to another state's alleged violations of international law in concert, at least some of these measures do not involve breaches of the rights of the target state. Even where they do seem to involve such a breach, this may not be recognized by the state taking the measure. As such, the question of collaboration in the taking of countermeasures has not been raised or considered on its own terms by these states. Third, collaboration in the taking of countermeasures has also tended to be collapsed into a different, though related, question.⁹ Most collaboration in response to wrongs has so far taken place in the context of sanctions regimes imposed on perceived violators of fundamental rules of the international legal order. For this reason, focus has been on the permissibility of countermeasures in the general interest—on “States other than the injured State” (in the sense of Article 48 of the Articles on State Responsibility) taking countermeasures in response to the violation of an erga omnes or erga omnes partes obligation. As we discuss below, the question of the lawfulness of countermeasures in the general interest is a critical aspect for thinking through the permissibility of collaboration, but it is only part of the picture.

The recent developments just mentioned—the proposal to confiscate Russian state assets, countermeasures in cyberspace, and the EU Anti-coercion Instrument—all distinctly raise the possibility of collaboration in the taking of countermeasures. At the same time, though, these positions and proposals do not specify, for the most part, the forms that collaboration can take. Collaboration can range from as little as a statement of support, to participation in the countermeasures of an injured state, to even taking such a measure at another state's request. These different forms of collaboration may, and often do, fall under different rules of international law. By implication, their regulation and, indeed, their permissibility, may not always be the same. For this reason, analyzing the form that collaboration takes and the rules that are applicable to each will be essential in understanding the extent to which, if at all, such collaboration is addressed by, and permitted by, the legal order.

In this Article, we consider five forms of collaboration between states, and assess whether a state acts lawfully when collaborating with an injured state in taking countermeasures. We distinguish: first, independent but coordinated action; second, the secondment of an agent or organ to another state; third, joint conduct; fourth, the provision of assistance to another state's countermeasure; and fifth, the idea that a state may take a countermeasure at the request and on behalf of an injured state. In practice it may often be difficult to determine the exact boundaries between certain forms of collaboration, especially where the terms of the collaboration are not clarified by the various participants. But at a doctrinal level, it is important to distinguish them since the legal rules applicable to each vary and the permissibility of the collaborating state's conduct may—and does, as we will argue—differ from case to case.

The focus of our analysis is on the question whether, and if so under what conditions, a state acts permissibly when it collaborates with an injured state in its countermeasures against the wrongdoing state (or “targeted” state). If the injured state acts in accordance with the substantive and procedural requirements for the taking of countermeasures against its wrongdoer, it will act permissibly. For the collaborating state, two possibilities arise. First, where the collaborating state possesses its own independent entitlement to take countermeasures against the wrong-doing state in question, few difficulties arise. As we will argue, this state will act permissibly when it collaborates with another state that is also entitled to take countermeasures against the same wrongdoer. The difficulty that arises in this scenario is that of identifying the circumstances in which the collaborating state will itself be entitled to take countermeasures against the same wrong-doing state. This will obviously be the case where the two collaborating states are injured by the state targeted by the countermeasure, but it also raises the broader—and long disputed—question of whether a state other than an injured state is entitled to take countermeasures when the prior breach violates an obligation erga omnes or erga omnes partes owed to it. Second, more difficult issues arise where the collaborating state is not itself entitled to take countermeasures in the circumstances. As we show below, the extent to which—if at all—the collaborating state also acts permissibly by cooperating in the injured state's countermeasures in these circumstances is unclear. This is likely to be the more common scenario in practice, insofar as the permissibility of countermeasures by states other than an injured state in response to erga omnes breaches remains controversial and, in any case, is only applicable in relation to a narrow set of obligations. For this reason, this scenario—where the collaborating state possesses no entitlement to take countermeasures—will be our primary focus.

To this end, the Article is structured as follows. Part II sets out some foundational aspects of the law of state responsibility and our framework for analysis. It addresses the idea of countermeasures as a justification to otherwise wrongful conduct, and the ongoing controversy about who is entitled to take countermeasures. Part III forms the bulk of our analysis—it concerns the lawfulness of collaboration in each of our five scenarios where the collaborating state has no entitlement of its own to take a countermeasure. Here, the most complicated question concerns the lawfulness of a collaborating state's assistance to another state's countermeasure. We show that it is plausible to think such assistance is lawful, but that certain considerations of principle relating to essential features of countermeasures challenge this view. In addition, our analysis in Part III addresses an idea that has emerged in the context of the application of international law to cyberspace—that a state may take a countermeasure at the request and on behalf of another state where it would otherwise not have an entitlement to do so. We term this proposal “proxy countermeasures” and show that at present it has no basis in international law, and cannot be justified by analogy to collective self-defense. Thereafter, Part IV briefly considers situations where the collaborating state possesses its own entitlement to act—showing that in all five collaboration scenarios, it acts lawfully. Finally, Part V turns to questions of policy that arise in relation to collaboration in the taking of countermeasures, and sets out potential directions for the development of the law. Part VI concludes.

II. Foundations and Framework

This Part provides a framework for our analysis to come. It will, first, discuss the role of countermeasures in the international legal order, their function as a justification to otherwise wrongful conduct, and the ongoing controversy about who is entitled to take countermeasures, particularly in response to the breach of an erga omnes or erga omnes partes obligation. Second, it will turn to collaboration, noting the limited engagement in the practice and literature with the idea of state collaboration in respect of countermeasures thus far, and the recent developments pointing in that direction.

A. Countermeasures in the International Legal Order

Countermeasures are a mechanism for the enforcement of state responsibility. Under the law of countermeasures, states whose rights have been breached by another state can, in response to the breach and with a view to inducing compliance with the targeted state's obligations of cessation and/or reparation, forego performance of an obligation owed to that state. The regime of countermeasures is set out in the Articles on State Responsibility (ARSIWA), adopted in 2001.¹⁰ In order to be permissible, countermeasures must satisfy a series of substantive and procedural conditions. Substantively, countermeasures must be a response to a prior wrongful act, and must be adopted by the injured state in order to induce compliance by the wrong-doing state.¹¹ They must be temporary,¹² reversible,¹³ and proportionate.¹⁴ There is no requirement that the injured state target an obligation connected or closely related to the prior breach: reciprocity is not needed, though the choice of obligation is limited by the exclusions listed in ARSIWA Article 50.¹⁵ Procedurally, before resorting to countermeasures, the injured state must call upon the wrong-doing state to comply with its obligations and, subject to urgency,¹⁶ notify the wrong-doing state of its intention to take countermeasures and make an offer to negotiate.¹⁷ ARSIWA leaves open the question whether states who are not injured by a breach may nevertheless take countermeasures when the breach affects an erga omnes (partes) obligation owed also to them—a question to which we return later.¹⁸

The regime in ARSIWA reflects a careful balance between competing considerations. On the one hand, there was the desire to ensure that international obligations are respected and injured states’ rights enforced: in the absence of centralized dispute settlement mechanisms, international law must afford states the power to seek this aim unilaterally. As the Argentinian delegate put it in the United Nations General Assembly's Sixth Committee: without countermeasures it was “the injured State [that] could find itself unable to bring about the cessation of wrongful acts in a timely manner, and could turn out to be the main victim of a decentralized system of international sanctions.”¹⁹ On the other hand, there was the desire to curb the potential for abuse which is inherent in unilateral self-help. There was ample historical evidence of the abuse of reprisals by powerful states against weaker ones, often on the basis of spurious or trivial claims.²⁰ Unsurprisingly, certain states that had been subjected to such abuse resisted the inclusion of countermeasures in ARSIWA or pushed for stronger and tighter restrictions. The point was put most forcefully by Brazil when it noted that to recognize “the legitimacy of countermeasures meant denying the equality of States and actually heightened their inequality.”²¹ The compromise eventually reached in the International Law Commission (ILC), and now included in ARSIWA, was found satisfactory by the majority of states in the UN General Assembly (UNGA).

Even if the regime on countermeasures in ARSIWA is generally accepted to reflect customary law, some doubts remain in respect of certain aspects, most notably the procedural conditions of notification and negotiation.²² Since the adoption of ARSIWA in 2001, states have tended to endorse the substance of its provisions, even though they have not always mentioned them expressly when taking or addressing countermeasures.²³ Thus, states have relied upon these rules in their diplomatic practice,²⁴ and in litigation.²⁵ The ARSIWA regime on countermeasures has also, for the most part, been endorsed by states in statements on international law in cyberspace.²⁶

Two aspects of the regulation of countermeasures are relevant for the analysis in the present Article: the characterization of these measures as justifications, namely defenses that render conduct permissible, and the question of who is entitled to take countermeasures. Each of these is addressed in turn.

1. Countermeasures as Justifications

Countermeasures constitute justifications, that is, they are defenses that render permissible conduct that would otherwise be unlawful. This characterization is widely accepted by states.²⁷ Justifications do not set aside the affected obligation(s): they have no effect on the continued existence of the relevant primary rule.²⁸ The primary rule affected under justification remains in force throughout: even during the period when the state is permitted to forego compliance with it in the light of the justifying circumstances. This is particularly clear in the case of countermeasures: ARSIWA Article 49(2) states that countermeasures are “limited to the non-performance for the time being” of obligations of the invoking state.²⁹ There are two features of the concept of justifications that are relevant for our purposes.

First, justifications involve conduct that is prima facie wrong: namely, that is inconsistent with an existing rule of the legal order. The prima facie wrong is at a minimum a procedural one,³⁰ in line with the International Court of Justice's (ICJ) approach in the Nicaragua case.³¹ That is to say, the conduct is wrongful at first blush, or on first impression, until all else—including the justification—is considered. But, all things considered, the conduct is held to be permissible. Even so, the point remains that, even after all things have been considered, and the conduct is found not to be a wrong after all, it is conduct inconsistent with an existing rule of the legal order. As we discuss in more detail later, this is why countermeasures are often described as “intrinsically wrongful acts.”³²

Second, justifications can entail permissible, rather than necessarily commendable, conduct. That is, it is not necessary to benefit from a justification that an agent's conduct is in some way right or praiseworthy, although it may be. Justifications can include conduct that is often described as being merely tolerable³³—conduct that, while permissible, is still viewed unfavorably by the legal order.³⁴ As we discuss in more detail in Section III.D, this is a feature of countermeasures widely emphasized in state practice and scholarship. It is in this sense that Monica Hakimi describes countermeasures as an institution of “unfriendly unilateralism” that must be tolerated despite their “unsavory attributes, because the legal order's formal enforcement processes are commonly weak or absent.”³⁵

2. Who Can Take Countermeasures?

According to ARSIWA Article 49(1), countermeasures can be taken by “[a]n injured State.” Injured states are those defined in ARSIWA Article 42. Under this provision, states are considered to be injured by a breach of international law in three circumstances, which turn on the character of the obligation breached.³⁶ These are: first, in the case of bilateral obligations, when the breach affects an obligation that is owed to that state individually;³⁷ second, in the case of obligations erga omnes or erga omnes partes, when the breach specially affects a state to which the obligation is owed;³⁸ and third, in the case of interdependent obligations, where the breach is of “such a character as radically to change the position of all the other States to which the obligation is owed with respect to the further performance of the obligation.”³⁹ Each of these state(s), as an injured state, is entitled “to resort to all means of redress contemplated in the” Articles on State Responsibility, including countermeasures.⁴⁰

Whether any other states, beyond those which are injured, are entitled to take countermeasures has been a contentious question for some time. In practice, states often take measures in response to perceived serious violations of international law. Throughout 2022 and 2023, for example, a large number of states adopted restrictive measures against Russia in response to the latter's aggression against Ukraine.⁴¹ So long as any such measures, even if unfriendly and antagonistic, do not infringe the rights of their target state, they are permissible. The more difficult question is whether all states (to whom an obligation is owed but who are not injured) can adopt measures inconsistent with the rights of a wrong-doing state, where that state's prior violation affects an obligation owed erga omnes (or erga omnes partes). After all, in line with ARSIWA Article 48, all states to which the obligation is owed are recognized to have a legal interest in compliance with it and, as such, to have legal standing to invoke the responsibility of the wrongdoer.⁴²

Which states are entitled to take countermeasures under international law is a key question for our analysis. As we will argue, where a state is entitled to take countermeasures against a wrong-doing state, it may lawfully collaborate with another state that is injured by the same wrongdoer and is, as such, also entitled to take countermeasures against it. But matters are less simple where the collaborating state is not so entitled. It is thus useful to consider in some more depth the position of states other than an injured state, and whether international law has come to recognize their entitlement to resort to countermeasures in response to violations of erga omnes (partes) obligations owed to them. After all, collaboration in the taking of countermeasures most likely arises in these circumstances—if only by virtue of the fact that multiple states will be reacting to the same breach. The next two sections will address the position taken by the ILC in ARSIWA as a starting point, and the practice since then, with a view to evaluating whether customary international law currently permits states other than an injured state to take countermeasures in the event of breach of erga omnes (partes) obligations. That is, it will ask whether countermeasures in the general interest are permitted under customary law.⁴³

a. The Position in ARSIWA

The Articles on State Responsibility make no provision for countermeasures in the general interest, though the matter was considered in the final two years of the ILC's work. In light of controversial views both within the Commission and the UNGA Sixth Committee,⁴⁴ the ILC resolved to leave the question open for future development. Article 54 of ARSIWA is accordingly a savings clause,⁴⁵ the Commentary to which notes the developing practice on countermeasures in the general interest but ultimately concludes that at that time—2001—there was “no clearly recognized entitlement of States referred to in article 48 to take countermeasures in the collective interest.”⁴⁶

The Commission's choice of a savings clause was driven by two main considerations. First, there was the view that the practice existing at that time was insufficient to meet the requirements of generality, consistency, and uniformity necessary to evidence a rule of customary law. This is clear from the Commentary to ARSIWA, which refers to this practice as “controversial,”⁴⁷ “limited and rather embryonic,”⁴⁸ and “sparse and involv[ing] a limited number of States.”⁴⁹ Second, there was the general tenor of states’ comments in the UNGA Sixth Committee. No fewer than thirty-five states addressed countermeasures in the general interest,⁵⁰ and of these six spoke in favor,⁵¹ thirteen against,⁵² and ten were cautious about this development.⁵³ In some cases, states’ views were in any event hard to interpret.⁵⁴ These were, to be sure, comparatively small numbers. But they were to be read against a background in which a considerable number of states had expressed concern about, or even opposed, the inclusion of countermeasures in the Articles on State Responsibility at all.⁵⁵ Perhaps, given the novelty of other key aspects of the Articles proposed by the Commission—not least of which was the very notion that states other than an injured state had legal standing to invoke responsibility—the choice may have also been driven by the desire to ensure that states in the Sixth Committee would accept the text as a whole.⁵⁶

b. Developments Since 2001

Academic commentary shortly after the adoption of ARSIWA was already critical of the ILC's stance. This commentary engaged in a deep and detailed reassessment of the practice considered by the ILC, noting that many of these measures were inconsistent with obligations owed by the states taking them toward the target state, and found additional instances of practice. These scholars thus concluded that the practice was not limited, embryonic, or sparse, as the ILC had suggested. On the contrary, it was said to be widespread and consistent: it was a settled practice.⁵⁷ Moreover, for these scholars, the Sixth Committee's reception of draft Article 54 (2000) was not as negative as suggested by Special Rapporteur James Crawford's Fourth, and final, report on state responsibility. In their view, states had been much more open and receptive to the notion of countermeasures in the general interest.⁵⁸ These scholars all concluded that international law currently permits this type of countermeasure.⁵⁹ But not all commentators share this assessment.⁶⁰

Since the adoption of the Articles in 2001, states have continued to resort to multi-party responses to perceived violations of erga omnes (partes) obligations.⁶¹ Martin Dawidowicz listed at least five instances of multi-party reactions to perceived violations of erga omnes (partes) obligations in the period 2001–2014.⁶² Further sanctions regimes were adopted, among others, against Venezuela since 2017,⁶³ against Myanmar in 2020, and against Russia and Belarus in 2022 and 2023.⁶⁴ The practice is, by now, considerable—quantitatively at least.

Two observations may nevertheless be made, which ought to be borne in mind when assessing whether the practice meets the requirement of generality.⁶⁵ First, the legality or illegality of the measures is often difficult to establish in the abstract, thus calling into question the relevance of at least some of this practice for the purposes of establishing a customary rule concerning countermeasures in the general interest.⁶⁶ For example, asset freezes, a very common measure in the contemporary sanctions toolkit, are often said to be prima facie wrongful acts,⁶⁷ as coercive interferences with state property⁶⁸ and/or breaches of state immunity.⁶⁹ But such assessments are not unanimous. Asset freezes do not seem to meet the Nicaragua threshold for coercion contrary to the principle of non-intervention,⁷⁰ and to the extent that they are implemented by executive action there is a question as to whether they are covered by immunities.⁷¹ Moreover, whether a measure is legal or not may vary from state to state, in light—in particular—of the treaty relations they may have toward the target state (which may either prohibit the measure, or provide for treaty-based specific responses to breach). Trade restrictions could, for instance, be justified under the General Agreement on Tariffs and Trade security exception,⁷² and other measures could be justified by other rules of international law—such as material breach of treaty or fundamental change of circumstances.⁷³

Second, despite increasing numbers, the practice remains predominantly Western. States in other regions, including Africa and the Middle East, have engaged in such measures too, though seemingly less frequently than Western states.⁷⁴ Aside from some Caribbean members of the Commonwealth,⁷⁵ states in Central and South America have, until recently, not participated in certain sanctions regimes against other states.⁷⁶ Latin America has a strong tradition of rejecting and opposing this kind of intervention, though it is noteworthy that there were some, perhaps timid, restrictive measures imposed against Venezuelan governmental officials in 2019.⁷⁷

The question of opinio juris also remains vexed. As a starting point, it is fair to say that evidence of opinio juris is limited. States rarely invoke countermeasures to justify measures or other sanctions imposed against perceived violators of erga omnes (partes) obligations. Indeed, those studies that have concluded that customary international law permits countermeasures in the general interest have tended to infer opinio juris from the practice.⁷⁸ Thus, for Dawidowicz, this inference is justified because the rule in question is a permissive one, the practice is consistent and regular and evidences no “opinio non juris,” and the notion of “third-party countermeasures” (in his terminology) is the only way to explain the legality of the relevant conduct.⁷⁹ But such an approach may be in tension with the ILC's Conclusions on the Identification of Customary International Law, widely endorsed by states in the UNGA Sixth Committee, pursuant to which ‘no simple inference of acceptance as law’ can be made from the practice in question.⁸⁰ For indeed, the general proposition must surely be that, as held by the ICJ in Nicaragua, legal views should not be ascribed to states, which they do not themselves formulate.⁸¹ To argue that the only way legally to explain the practice is by means of the concept of countermeasures overlooks the fact that a state might have assessed its conduct to be permissible in light of its existent commitments and accepted the risk that it might be unlawful. Likewise, the (relative) absence of protest to these measures by other states may evidence opinio juris only in specific circumstances.⁸²

The available evidence of states’ legal views on countermeasures in the general interest remains mixed. During the ILC work, as already noted, at least six states expressed support for the notion of countermeasures in the general interest—at the very least, as progressive development.⁸³ A further five could be said to have implicitly supported this development, as they expressed satisfaction with the proposals made by the Commission on countermeasures as a whole.⁸⁴ At the same time, other states have clearly rejected this notion. No fewer than thirteen states were opposed to draft Article 54 (2000) in the Sixth Committee, and ten expressed caution.⁸⁵

As to other evidence, states have expressed views on the permissibility of these measures in other contexts. At one end of the spectrum are states which continue to reject the notion of countermeasures in the general interest. These include Canada,⁸⁶ China,⁸⁷ and France.⁸⁸ At the other end are states which have expressly endorsed the customary status of this entitlement: Costa Rica, ⁸⁹ Estonia,⁹⁰ Ireland,⁹¹ Italy,⁹² and Poland.⁹³ In between, are those states that have endorsed the possibility in aspirational terms, like New Zealand,⁹⁴ and those that have continued to express doubts as to their legality or noted ongoing disagreements among states on the matter, like Algeria,⁹⁵ Denmark,⁹⁶ Netherlands,⁹⁷ Russia,⁹⁸ and (seemingly) the United Kingdom.⁹⁹ Beyond this, expressions of support or objection in the abstract for this concept are infrequent.

It is also noteworthy that states that regularly adopt sanctions against perceived violators of erga omnes (partes) obligations, such as the United States and the United Kingdom, have so far not explicitly and clearly supported the permissibility of countermeasures in the general interest—either in the Sixth Committee, during the ILC's work on state responsibility, or in their legal justifications for the restrictive measures and sanctions regimes. In relation to Russia in particular, the United States has been reported as endorsing countermeasures adopted “by states ‘injured’ and ‘specially affected’ by [the Russian] aggression” against Ukraine.¹⁰⁰ It has included in this category “allies of Ukraine who have bankrolled its economy and military during the war.”¹⁰¹ This approach would transform the United States, and certain other allies of Ukraine, from being a “State other than the injured State,” in the sense of ARSIWA Article 48, into injured states, in the sense of ARSIWA Article 42.¹⁰² This move would place them in the category of states that are entitled to take countermeasures under current international law, bypassing the debate on countermeasures in the general interest.

In sum, assessments of customary law are notoriously difficult, and the present context raises distinctive methodological problems. As Fernando Lusa Bordin has observed, few customary rules “even those long viewed as established, can survive the brutal scrutiny of the magnifying glass.”¹⁰³ While there is no doubt that evidence of both practice and opinio juris is necessary to identify a rule of customary law, this two-element approach, endorsed in the ILC Conclusions with the support of states, must be applied flexibly.¹⁰⁴ Even if assessed flexibly, however, the best that can be said is that the law remains uncertain.

B. Collaboration in the Taking of Countermeasures

States regularly collaborate in the pursuance of shared, or collective, aims. Such cooperation is often desirable and, in many instances, necessary: some challenges facing the international community, such as climate change and health emergencies, can only be addressed through coordination and cooperation. States also regularly collaborate in the adoption of sanctions regimes against perceived violators of erga omnes (partes) obligations. For example, G7 states have closely cooperated in the imposition of sanctions against Russia in response to its aggression of Ukraine.¹⁰⁵

As we noted in the introduction, some states have recently addressed collaboration in the taking of countermeasures. If the proposed plan to distribute Russian assets goes ahead, it will likely involve collaboration amongst a set of allied states in the taking of a countermeasure: as these states have acknowledged, the confiscation of such assets would need to be justified by countermeasures if it is to be permissible.¹⁰⁶ Similarly, in recent statements on the application of international law to cyberspace, a number of states have referred to the possibility of collaboration in countermeasures to respond to the wrongful cyberactivity by other states. Thus, Canada suggests that “[a]ssistance can be provided on request of an injured State, for example where the injured State does not possess all the technical or legal expertise to respond to internationally wrongful cyber acts.”¹⁰⁷ Costa Rica proposes that countermeasures may be taken by what they call third states “in response to violations of obligations of an erga omnes nature or upon request by the injured State.”¹⁰⁸ And Estonia has been a long-standing supporter of the idea that “[a]llies matter also in cyberspace”—that “states which are not directly injured may apply countermeasures to support the state directly affected by the malicious cyber operation.”¹⁰⁹ Finally, in December of 2023, the EU Anti-coercion Instrument came into force, which allows the EU Council to decide on responses to perceived economic coercion against the Union or (one of) its members, including by way of countermeasures.¹¹⁰

This renewed emphasis on collective action appears to be instigated by two concerns, old and new. As to the old, there has been a longstanding worry that the material inequalities among states are such that not only are weaker states more likely to be targeted by countermeasures, they are also less likely to be in a position to take countermeasures themselves, and to do so effectively.¹¹¹ For example, there are likely to be fewer foreign state-owned assets in smaller and less developed economies that could be subjected to freezing as a means to induce compliance. Smaller states are also less likely to possess the technological capability to respond with cyber-measures to wrongful acts, including malicious cyberoperations, by another state. For this reason, the question of collaboration really matters, for it provides a mechanism to offset, if only partially and contingently, the impact of material inequalities among states on the enforcement of their rights. Otherwise, these states may suffer a compounding of injustice: the initial wrongful act, and the material inability to avail themselves of an important enforcement mechanism ostensibly available, equally, to all states. As to the new, or at least a concern that is increasing, the escalating polarization in international politics make the question of state collaboration, particularly in regional blocs, even more likely—which may lead, too, to increased support and cooperation also in reacting to perceived wrongs.

Leaving aside the recent examples just mentioned, collaboration in the context of countermeasures has rarely been addressed as such. Within the ILC, as noted by Martti Koskenniemi, the default assumption was that “[t]here was no general right to assist the injured State” in taking countermeasures.¹¹² Collaboration was, however, proposed as a requirement for the taking of countermeasures in the general interest. The ILC's proposal to states on this question required both a request from the injured state, where there was one, and cooperation between states participating in countermeasures in the general interest.¹¹³ These proposals, as noted above, did not make it into the final draft. Beyond this, there is no evidence of the issue having been discussed within the Commission or in the Sixth Committee of the UNGA. There has been similarly limited engagement by the literature on this topic, and then rarely in depth. Vague references may be found in writings of the late nineteenth and the first part of the twentieth centuries.¹¹⁴ Since then, only a handful of writers considered collaboration in this field in the 1970s and 1980s. Michael Akehurst and D.N. Hutchinson's accounts anticipate the notion of countermeasures in the general interest: indeed, crucial for their respective accounts was the identification of an entitlement, for the collaborating state, to take countermeasures.¹¹⁵ And while Elisabeth Zoller clearly addressed collaboration where the prior breach affected an individual right of the injured state, her considerations were limited to policy questions: it seemed right and natural, she thought, that victims should rely on their allies in enforcing their rights.¹¹⁶ The legal regulation of collaboration in this field thus remains unclear.

These recent developments call for an in-depth analysis of the question of collaboration in the taking of countermeasures. The developments have, indeed, come from different states and as a reaction to different contexts: the Russian invasion of Ukraine, the cyber-activity of states, and the increased use of trade-related measures in a polarized world. They signal broader trends in international relations, and not just contingent responses to particular threats. At the same time, states have addressed the question in very broad terms: the specific manner of cooperation has not been clarified. And yet, what form cooperation takes may affect the permissibility of the conduct of those collaborating in the taking of countermeasures.

In this light, in Part III below we distinguish five ways that states might collaborate in the taking of countermeasures. We use “collaborate” in a non-technical sense, as an umbrella term to cover five ways states may interact in this context. These concern first, independent but coordinated action; second, the secondment of an agent or organ to another state; third, joint conduct; fourth, the provision of assistance to another state's countermeasure; and fifth, the idea that a state may take a countermeasure at the request and on behalf of an injured state. We leave aside certain other ways that states may become involved in some way in the countermeasures of another state.¹¹⁷ Finally, it may be noted that a given factual situation may need to be assessed according to more than one category.

As we will argue, the rules regulating these different forms of collaboration vary and, as such, they result in different conclusions on the legality of the collaborating state's participation in the countermeasures of an injured state. Where the collaborating state possesses an entitlement of its own to take countermeasures against the targeted state, it can permissibly cooperate with another state in an equivalent position in taking countermeasures against their common wrongdoer. The ongoing uncertainty about the permissibility of countermeasures in the general interest, as well as the potential for collaboration in situations where there is no erga omnes (partes) breach in play at all, points to the need to understand the permissibility of collaboration in situations where the collaborating state does not possess an entitlement to take countermeasures. The majority of cases of multi-party responses to wrongdoing are likely to fall into this category. And as we will show, where the collaborating state does not possess an entitlement to take countermeasures, the permissibility of its conduct varies depending on the form of collaboration at issue. Whether this variation is desirable and justifiable is a question to which we turn in Part V.

III. Collaboration Where the Collaborating State Has No Individual Entitlement to Take a Countermeasure

This Part addresses five forms of collaboration in the taking of countermeasures. It does so where the collaborating state—the state on which we are focused—has no individual entitlement to take a countermeasure against the targeted state, but the injured state does.

A. Independent but Coordinated Action

The first form of collaboration arises where two (or more) states act individually while coordinating their response. Consider, for instance, a situation where two states independently confiscate and distribute the central bank assets of another state held in each of their jurisdictions, or where two states independently suspend landing rights of the national carrier of another state. This independent action may be taken pursuant to either an ad hoc or institutionalized common position or policy. However, there is no sense here that the independent but coordinated action entails a joint operation. Each state's acts stand on their own.

This situation is not complex in legal terms. It is straightforwardly the case that each state commits an act that is, independently, contrary to the rights of the targeted state. In line with the principle of independent responsibility, which underpins the Articles on State Responsibility each state is individually and independently responsible for its own act¹¹⁸—for example, its own confiscation of the assets, its own suspension of overflight rights. And this is the case even if there was some coordination between them. In this scenario, then, each state will have committed a wrongful act against the targeted state and, for this reason, each will require their own justification for their putatively wrongful conduct. Given that the collaborating state does not have one, and subject to the discussion to come on the possibility of proxy countermeasures, it acts wrongfully.

B. Secondment

The second form of collaboration is that of secondment. Here, the collaborating state seconds an official or a unit to the injured state. In ARSIWA, the Commentary notes that this might “include a section of the health service or some other unit placed under the orders of another country to assist in overcoming an epidemic or natural disaster, or judges appointed in particular cases to act as judicial organs of another State.”¹¹⁹

In understanding the legal implications of such a secondment, ARSIWA Article 6 sets out the following rule:

The conduct of an organ placed at the disposal of a State by another State shall be considered an act of the former State under international law if the organ is acting in the exercise of elements of the governmental authority of the State at whose disposal it is placed.¹²⁰

These are situations where, as noted by the Commentary, “in performing the functions entrusted to it by the beneficiary State, the organ must also act in conjunction with the machinery of that State and under its exclusive direction and control, rather than on instructions from the sending State.”¹²¹ These are narrow situations, where the organ of one state is placed at the disposal, and acts under the authority and for the benefit, of another state.¹²²

In legal terms, the effect of the “loan” of the state organ is that the conduct of that organ is attributable to the receiving state, and this state only. The conduct is not attributable to the sending state. When the Privy Council acts as court of final appeal of, say, Jamaica, its conduct is attributable to Jamaica and not to the United Kingdom.¹²³ Or when French or Spanish judges sit in the courts of Andorra, their conduct is attributable to Andorra and not to either France or Spain.¹²⁴ Similarly, consider a situation where the collaborating state fully seconds a cyber unit of its Defense Ministry to the injured state to assist it with the protection of its cyber infrastructure. Assuming that the unit acts under the exclusive direction and control of the injured state, conduct of the unit during the secondment would be attributable only to the injured state, and not to the seconding state.

This, then, appears to be a permissible form of collaboration in terms of our framework, though to describe it as such might be somewhat misleading. As set out above, under ARSIWA Article 6, the conduct of the seconded unit becomes attributable to the injured state. This state—as an injured state—has an entitlement to take a countermeasure, and so any such measure taken by its own organs or by the seconded unit will be covered by the defense. As to the seconding state, if the conditions of Article 6 are met, the key point is that there is no conduct by that state: the conduct of the seconded unit is not its conduct to justify. While in fact this may be a manifestation of cooperation among states, in a technical legal sense this description is not accurate—for the collaborating state is not acting at all.¹²⁵

In conclusion, two caveats may be noted. First, secondment of this type—that is, which meets the high threshold of Article 6—is not common, particularly where the secondment involves participating in actions of the receiving state that are prejudicial to another state. Indeed, practice shows that the “loan” of military units or personnel is generally not made in a way that satisfies the high threshold of Article 6: sending states rarely relinquish all control over those units or personnel.¹²⁶ Second, assuming such a threshold was met, although the rule set out in Article 6 is generally accepted, it does remain to be seen whether it would hold up in practice outside of the relatively non-conflictual examples given in the ILC's Commentary. Where a powerful state seconds a unit to a less powerful state that undertakes an operation contrary to the interests of another state, would that other state accept the complete non-attributability of that conduct to the powerful state? That remains an open question.

C. Joint Conduct in International Law

The third form of collaboration arises where two or more states undertake joint action or a joint operation. Consider a situation where two states acting together seek to leverage a vulnerability in a software program used by the wrong-doing state to gain unauthorized access to the networks of its cyber unit, whereupon they compromise the networks’ integrity.¹²⁷ Or consider two states undertaking a joint ocean patrol which seizes a vessel belonging to the wrong-doing state.¹²⁸

In thinking about joint responsibility in legal terms, there are two central possibilities.¹²⁹ First, states may act through a common organ.¹³⁰ In this respect, Stefan Talmon gives the example of the Coalition Provisional Authority in the occupation of Iraq between April 2003 and June 2004.¹³¹ In its core form, this scenario will entail the establishment of an integrated body composed of agents of each state.¹³² Each state is then responsible for the conduct of the common organ.¹³³ These are likely to be relatively exceptional cases.

Second, states can also act jointly in relation to specific conduct or a specific operation, without the creation of a common organ. As a statement of principle, the Commentary to ARSIWA Article 47 provides that:

[T]wo or more States might combine in carrying out together an internationally wrongful act in circumstances where they may be regarded as acting jointly in respect of the entire operation.¹³⁴

This implicitly assumes that the conduct of the operation as a whole is attributable to both states. As a number of scholars have noted, the elements of joint responsibility, in this sense, remain uncertain.¹³⁵ There is, indeed, difficulty, in understanding the boundaries of this form of responsibility.

To begin with, it is fair to assume that any such doctrine of joint attribution of conduct will be narrow, particularly when compared to relatively capacious doctrines of joint perpetration in domestic criminal or civil law.¹³⁶ As André Nollkaemper and Dov Jacobs argue, this may be explained by reference to the sovereign independence of states,¹³⁷ and its application in the law of state responsibility: the principle of independent responsibility.¹³⁸ Moreover, practice to date provides little clarity, and it is difficult to say much in the abstract, without reference to the specific primary rule at stake.¹³⁹ Scholars, too, disagree on the scope of this category.¹⁴⁰ For present purposes, we propose that at least where two states make essential and concerted contributions to a specific operation, the conduct of the entire operation may be attributed to both of them.¹⁴¹

In these instances, the attribution of the conduct to both the injured state and the collaborating state leads to different outcomes as between them. For the injured state, the wrongfulness of its conduct is precluded on the basis of its entitlement to take a countermeasure. Of course, all conditions for the taking of countermeasures must be met. For the collaborating state, however, the outcome is different. It would need its own justification for its conduct vis-à-vis the targeted state—the conduct that was attributable to both states. Given we assume here that the collaborating state has no individual entitlement to take a countermeasure, it has no way to justify its conduct. This state's conduct remains wrongful and a breach of its international obligations to the targeted state.¹⁴²

D. Assistance to Another State's Countermeasure

The fourth form of collaboration arises where the collaborating state assists the injured state in taking a countermeasure against the targeted state. Canada, for instance, proposed in its 2022 statement on the application of international law to cyberoperations that:

Assistance can be provided on request of an injured State, for example where the injured State does not possess all the technical or legal expertise to respond to internationally wrongful cyber acts. However, decisions as to possible responses remain solely with the injured State.¹⁴³

In this scenario, then, the collaborating state does not itself take a countermeasure—neither individually but pursuant to some coordination, nor jointly as part of an essential and concerted contribution to a joint operation. It merely assists the injured state in taking that measure for itself. The issue under consideration is thus not whether the collaborating state is responsible for the violation of the obligation infringed by the countermeasure as such. Only the injured state would be prima facie responsible for that breach, which it could justify as a countermeasure. The issue is whether by assisting the injured state in taking that measure, the collaborating state itself commits a wrongful act.

This scenario is both the most practically relevant and the most legally and theoretically complex. It is difficult to find public practice on this form of collaboration among states, and there has so far been limited engagement in the scholarly literature.¹⁴⁴ We start this Section by setting out possible rules of international law which the collaborating state's assistance might breach, before discussing how these apply where a state assists in another state's countermeasure.

1. Aid and Assistance in International Law

There are two kinds of rules, in particular, that might regulate a situation in which a state assists another state in the taking of a countermeasure: certain specific primary rules and the general rule reflected in ARSIWA Article 16.¹⁴⁵ The first arises where the very conduct that constitutes the assistance infringes a primary rule of international law that the collaborating state owes to the targeted state.¹⁴⁶ It is impossible to set out the range of examples where this might be the case, but a hypothetical will illustrate the point. It could be, for instance, that the collaborating state and the targeted state are parties to a bilateral treaty pursuant to which they share intelligence and cooperate on technical matters of defense, and under which disclosure of information to third parties is prohibited.¹⁴⁷ If the collaborating state shares with the injured state information about the targeted state's cyber vulnerabilities obtained through that cooperation arrangement under the bilateral treaty, this would straightforwardly breach the primary norm.¹⁴⁸

The second is the general rule reflected in ARSIWA Article 16, found to be customary in the Bosnian Genocide case.¹⁴⁹ Article 16 provides:

A State which aids or assists another State in the commission of an internationally wrongful act by the latter is internationally responsible for doing so if:

1. (a) that State does so with knowledge of the circumstances of the internationally wrongful act; and

2. (b) the act would be internationally wrongful if committed by that State.

Four elements of the rule in Article 16 are worth stressing. First, the assisting state must provide its assistance with knowledge of the circumstances of the wrongful act committed by the assisted state.¹⁵⁰ That is, the assisting state (through its organs) must be aware that its assistance is facilitating a wrong by the assisted state.¹⁵¹ Second, the rule is not concerned with general cooperation between states—there must be a clear link between the assistance and the recipient state's wrongful conduct.¹⁵² Third, the rule entails what is known as the “double obligation” requirement: if the assisting state had undertaken the relevant wrongful conduct of the assisted state, it would have been wrongful for it too.¹⁵³ Finally, under the rule, the assisting state is responsible for its own acts of assistance,¹⁵⁴ and not for the conduct of the assisted state.

2. A Provisional View on the Lawfulness of Assistance

The question, then, is whether the collaborating state acts wrongfully in assisting the injured state's countermeasure, where, to reiterate, the collaborating state has no entitlement of its own to take a countermeasure. As a starting point, it is necessary to distinguish situations where there is an applicable specific primary rule, as mentioned above, from the application of the general complicity rule in ARSIWA Article 16. As to the former, the example given above entailed the collaborating state's act of assistance infringing a bilateral treaty between it and the targeted state relating to the non-disclosure of technical information to third parties. Where the collaborating state shares such information with the injured state to facilitate its countermeasure against the targeted state, it infringes its obligations under the treaty.¹⁵⁵ Moreover, given that it has no individual entitlement to take a countermeasure, it has no defense for doing so. Unless the injured state's own defense can somehow be vested in or transferred to the collaborating state, a matter we discuss in the next Subsection, the collaborating state would act wrongfully.

The more widely applicable and pressing question is whether the collaborating state, in facilitating the injured state's countermeasure, would breach the general rule on assistance in ARSIWA Article 16. By way of introduction, two points may be stressed. First, there appears to be little (public) practice on this question and, since the ILC did not raise this issue during its work on state responsibility,¹⁵⁶ it is difficult to say how states (would) frame the permissibility of their collaboration in these circumstances in legal terms. Second, there appear to be no clear and obvious solutions to what domestic scholarship sometimes calls the “problem of accessorial liability,”¹⁵⁷ namely to the question of the (im)permissibility of assistance in the justified, but prima facie wrongful, act of a principal. This is evident in the variety of solutions available in domestic legal orders: systems of domestic criminal law¹⁵⁸ and private law¹⁵⁹ resolve this problem in a range of ways, and their choices usually depend on the weighing and balancing of system-specific factors.¹⁶⁰

With these points in mind, it makes sense to start with a preliminary, and intuitively appealing, view. To reiterate, Article 16 applies where a state “aids or assists another State in the commission of an internationally wrongful act.” That is, Article 16 is a complicity rule, one which turns on the existence of the principal's wrong. Thus, of course, if the injured state's countermeasure does not fulfill the relevant conditions for the lawfulness of countermeasures, then the collaborating state will be responsible for its wrongful assistance, assuming the other elements of Article 16 are fulfilled.¹⁶¹ It will have assisted in the commission of a wrongful act by the assisted state. Crucially, though, if the assisted state's countermeasure is lawful, it would seem to follow that there is no wrong for the assisting state to be complicit in.¹⁶² On this approach, assuming there is no other specific primary rule in play, it would be lawful for states to assist the injured state in taking a countermeasure.

This intuitive response to the problem of assistance to a countermeasure may be supported by reference to a commonly held view in criminal law doctrine and scholarship. In that context, scholars have long discussed the applicability of a principal's defenses to secondary parties or accessories.¹⁶³ The issue is closely linked to the debate on the distinction between justifications and excuses. One common view is that if a principal's conduct is justified, then accessories may assist them and benefit from their defense.¹⁶⁴ Douglas Husak illustrates this position with the following example: “Suppose that [the principal] acts in self-defense in repelling an unlawful aggressor. Surely [the accessory] has a defense if he assists [the principal] in his efforts.”¹⁶⁵ By contrast, where the principal's conduct is merely excused, it is suggested that any accessory will not benefit from the excuse.¹⁶⁶ Underpinning this doctrinal position is the assumption that where the principal's conduct is justified, their act is lawful, and assisting in a lawful act is not wrongful.¹⁶⁷ Here, the fact that justifications center on features or characteristics of the act (rather than on features and characteristics of the actor), and render that conduct lawful within the legal order, produces a “universalizing” effect such that they can also be relied on by accomplices.¹⁶⁸

On this approach, the fact that countermeasures are generally understood as providing a justification to the injured state would mean that the assisting state commits no wrong. The injured state's justification would have a universalizing effect, providing cover also for the conduct of the assisting state. This approach is certainly plausible, and has found support as a position of principle in the international law scholarship that has addressed this problem.¹⁶⁹

3. Assisting Tolerable Conduct and the Question of Agent-Relative Justifications

In this Section, we draw attention to two factors that provoke further reflection on this provisional conclusion. They stem from the traditional caution and concern of many states about the institution of countermeasures in international law, and the tightly regulated regime intended to provide safeguards against their abuse.

First, many states were ambivalent about the entire institution of countermeasures during the ILC's work, several even resisted their inclusion in ARSIWA at all.¹⁷⁰ These states tended to warn against the dangers inherent in the unilateral and—usually—unchecked character of countermeasures. For example, Brazil thought them “distasteful,”¹⁷¹ Mexico that they tended to exacerbate instead of solve disputes,¹⁷² and South Africa that they should be “marginalized.”¹⁷³ Even states that were more accepting of the notion, tended to emphasize the exceptionality and undesirability of countermeasures. For Argentina, they “could only be tolerated under international law as an extreme remedy to be taken only in exceptional cases,”¹⁷⁴ and for Colombia “[r]esort to countermeasures was not a right under international law, but a barely tolerated practice in exceptional cases covered by partially developed customary law.”¹⁷⁵ More recently, Brazil's legal position on international law in cyberspace recalled “that several states have criticized countermeasures because they would be prone to abuses, especially due to the material inequality of states.”¹⁷⁶ Scholars, too, have emphasized this ambivalence. Hakimi, quoted earlier, frames countermeasures as an institution of “unfriendly unilateralism” and, in language reminiscent of that of Argentina and Colombia, argues that countermeasures must be “tolerated” by international law “because the legal order's formal enforcement processes are commonly weak or absent.”¹⁷⁷ A common definition highlights their character as “intrinsically wrongful acts.”¹⁷⁸ These observations point, essentially, to one of the key features of countermeasures we emphasized earlier: that they involve conduct that is prima facie wrongful and that, even if all-things-considered permissible, remains inconsistent with an existing rule of the legal order.¹⁷⁹

Second, states’ traditional unwillingness to accept the permissibility of countermeasures in the general interest—as we discussed above—suggests that an entitlement to take countermeasures arises from a particular relationship with the injury and/or the obligation breached by the wrong-doing state. During the ILC's work, states could not agree that any state beyond the injured state, even one to which the obligation breached is owed, could react by way of countermeasures. These discussions did not revolve around our question of assistance to a countermeasure. But they reveal an important feature of the right to take countermeasures: namely, that it is personal to the injured state (or Article 48 state, if that is the case) by virtue of its particular connection to the injury and/or the breach in question.

In our view, these points reveal two important features of countermeasures—their tolerable and personal character—which may have a bearing on whether it ought to be permissible for one state to assist another's countermeasure. Moreover, these features track two insights in philosophical work that challenge the intuitive view that justifications are universalizable, and thus necessarily lead to the legality of the accessory's acts of assistance. The first concerns what exactly is entailed in the idea of justification¹⁸⁰ and, in particular, how countermeasures fit into this category. The second concerns the idea that countermeasures, although a justification, entail what philosophers refer to as an agent-relative justification.

As to the first, we noted already that the category of justification is often understood as entailing permissible rather than commendable conduct, and that permissible conduct can include both commendable or praiseworthy behavior as well as acts that are merely tolerated by the legal system.¹⁸¹ After all, to put it more broadly, the law can “take an unfavourable view of actions that it ultimately permits.”¹⁸²

A distinction between permissible conduct that is commendable and permissible conduct that is tolerable is not unknown to international law.¹⁸³ For instance, in his declaration in the Kosovo Advisory Opinion, Judge Bruno Simma took issue with the ICJ's conceptual approach to the question of (im)permissibility of unilateral declarations of independence under international law. The UNGA's resolution asked the Court whether the unilateral declaration of independence by Kosovo's Provisional Institutions of Self-Government was “in accordance with” international law.¹⁸⁴ The Court responded that the declaration had not violated international law. The Court's approach to this question, said Judge Simma, revealed a binary of permission/prohibition that he thought was inconsistent with the contemporary international legal order, and which omitted “the possible degrees of non-prohibition, ranging from ‘tolerated’ to ‘permissible’ to ‘desirable.’”¹⁸⁵ The Court, in his view, could have accepted that international law was neutral or silent with respect to certain acts, or even that it might simply “tolerate” those acts. For Judge Simma, “toleration” reflected a legal category between permission and prohibition: a concept of “non-prohibition” or “not illegality,” rather than “permissibility” and “legality.”¹⁸⁶ But the idea need not go that far, and we need not, and do not, adopt it here. It is possible to make room for certain “shades” within the category of legally permissible conduct by accepting that permissibility does not always entail good or right conduct (“desirable,” in Simma's categories), but that it may include also tolerable acts.¹⁸⁷

This point—the understanding that justified conduct may include both commendable and tolerable conduct—has implications for thinking about the role of those who assist. As Husak argues:

If [the principal's] conduct were commendable, it is plausible to suppose that the law should not discourage, and might actually encourage, assistance from others, notwithstanding the fact that [the principal's] conduct apparently violates a criminal law. But these results do not follow if [the principal's] conduct is merely tolerable. The law need not encourage, and might actively discourage, assistance with conduct that it is willing to permit.¹⁸⁸

A similar insight is revealed by the distinction between agent-neutral and agent-relative justifications, which tracks the second feature mentioned above. The distinction is generally accepted among moral philosophers, although much remains contested.¹⁸⁹ In this context, an agent-neutral justification is one that applies to all persons who are placed in the particular situation. This is typically the case with lesser-evils or consequentialist justifications, as illustrated in the famous trolley problem. If we accept that it is permissible to pull the lever to kill one and save five, such conduct would be permissible for any person who could pull the lever and saw the situation in the same way.¹⁹⁰ Or consider a situation where, in order to save a town from the spread of a fire, an individual control-burned a private dwelling. All else being equal, any person in the position of that individual would act permissibly if they set fire to the dwelling. But law and morality also allow individuals, in certain circumstances, to prioritize their own interests over those of others, even where their acts do not produce a net benefit or reduce overall harm. This includes acts that express an actor's partiality toward themselves, or toward those close to them with whom they have a special relation. To use an example from Larry Alexander and Michael Moore, a parent is “thought to be permitted (at the least) to save his own child even at the cost of not saving two other children to whom he has no special relation.”¹⁹¹ Here, an agent-relative justification allows the parent “to give extra weight to [their] own legitimate interests in [their] moral calculus.”¹⁹²

As with the distinction between commendable and tolerable conduct, the distinction between agent-neutral and agent-relative justifications may also have implications for the position of those who assist. Kimberley Kessler Ferzan, for instance, suggests that agent-relative justifications do not “permit third-party assistance,”¹⁹³ as do Alexander and Moore.¹⁹⁴ Similarly, Husak gives the example of a mother who diverts a train which is speeding toward her infant daughter, thus killing two other innocent children.¹⁹⁵ Although it may be the case that the mother's conduct is permissible on the basis of an agent-relative justification—that is, the parental relationship—it does not follow that others who are strangers should be allowed to assist the mother in prioritizing her own interests over the general good.¹⁹⁶ In fact, Husak puts it more strongly still, proposing that it “makes little sense for the law to allow or encourage any [third party], who bears no special relation to [the principal], to assist her.”¹⁹⁷

4. Conclusion on Assistance

Practice provides no clear answer to the problem of accomplice liability in relation to countermeasures in international law.¹⁹⁸ Two points pull in different directions. In one direction, there is the intuitive view that because countermeasures are justified—because the principal state's act is all-things-considered permissible—this should have a universalizing effect. As such, a state that assists in the countermeasure of another state would act permissibly. Pulling in the opposite direction is, however, the fact that the international legal order takes a cautious approach to countermeasures, one that suggests such measures are tolerable, rather than commendable, and that they arise from the particular relation between the injured state, the injury, and the breach. This point puts pressure on the intuitive response: for it provides reasons to restrict those behaviors, including by limiting the participation of third parties. We return in Part V to considerations of policy that may bear on this question.

E. Countermeasure at the Request and on Behalf of Another State – Proxy Countermeasures

The fifth form of collaboration arises where the collaborating state purports to take a countermeasure at the request and on behalf of an injured state. In practical terms, consider a situation where the targeted (wrong-doing) state violates a bilateral obligation it owes to the injured state: it fails, for example, to ensure the inviolability of the injured state's diplomatic premises. Alternatively, the targeted state might be violating the injured state's territorial sovereignty through an exorbitant exercise of enforcement jurisdiction within the injured state's territory. The injured state would be entitled to take countermeasures against it. Could the injured state request a collaborating state to take a countermeasure on its behalf?¹⁹⁹

That a state may take such a measure in this situation is, perhaps, the most radical proposal to emerge from recent discussion on the application of international law in cyberspace.²⁰⁰ Thus, to give an example, Costa Rica's recent statement proposes that “countermeasures may be taken by the injured State, i.e., the State specifically affected by the breach, as well as third States in response to violations of obligations of an erga omnes nature or upon request by the injured State.”²⁰¹ The proposal is radical because it suggests that a state that is not entitled to take countermeasures may rely on the entitlement of the injured state by virtue of the latter's request. To put it more bluntly, where a state is prohibited to, for example, freeze the assets of the target state on its own, it may permissibly do so if, and because, it is requested to do so by a state which can so act.

The original nature of this proposal is sometimes obscured by the terminology that has become accepted in the cyber context: states and scholars alike have discussed this form of collaboration under the label of “collective countermeasures.”²⁰² This term, however, does not distinguish situations where the collaborating state taking the countermeasure (1) has and (2) does not have an independent entitlement to take such a measure. Indeed, the term “collective countermeasures” has previously been used to describe what we have termed “countermeasures in the general interest”: that is, it is often taken to refer to the entitlement of states other than an injured state to take countermeasures in response to the breach of an erga omnes (partes) obligation owed to them. It is for this reason that we think a new term is needed to refer to the scenario of a collaborating state that, without an independent entitlement, takes a countermeasure at the request, and on behalf, of the injured state. For, as we will argue, whether the collaborating state possesses an entitlement of its own to take countermeasures currently determines the legality of its collaboration in the countermeasures of another in this scenario. We will refer to this latter collaboration scenario—where the collaborating state has no entitlement of its own—as one of “proxy countermeasures.”

In this Section, we start by showing how the possibility that a state might request another to take countermeasures on its behalf was considered by the ILC. The crucial point, in this respect, is that the ILC's consideration exclusively arose in the context of assessing the entitlement of Article 48 states to resort to countermeasures in the event of a breach of an erga omnes (partes) obligation that was owed to them. It did not arise where the obligation breached was not owed to the requested state. Thereafter, we review the available practice, limited as it is. As will be seen, no clear basis emerges from these sources in support of the concept of proxy countermeasures.²⁰³ We will then consider additional arguments that might be taken to support the permissibility of the collaborating state acting at the request and on behalf of the injured state in the absence of an independent entitlement to do so: two analogies, and the idea of rights vesting. The analogies are, however, faulty, and international law does not recognize any doctrine of transfer or vesting of a defense in this context.

1. The Idea of Request in the Work of the ILC

The idea that countermeasures could be taken by a state on behalf and at the request of another state emerged in the work of the ILC, during its consideration of countermeasures in the general interest.²⁰⁴ The Commission considered that, where there existed a state that was specially affected (and so, injured) by the breach of the erga omnes (partes) obligation,²⁰⁵ other states to which the obligation was owed should not be allowed to act without a request by this state. This was because states that were not injured “should not be able, in effect, to intervene in a dispute by taking countermeasures if the principal parties wish to resolve it by other means.”²⁰⁶ As discussed in Part II, the proposed draft Article 54 did not survive the cool reception given to it by states in the Sixth Committee. In its final form, Article 54 is simply a savings clause that makes room for further development. The Commentary to this provision, however, retains some echo of the distinction when it notes, after a brief survey of relevant practice, that “in those cases where there was, identifiably, a State primarily injured by the breach in question, other States have acted at the request and on behalf of that State.”²⁰⁷ Whether this is now a requirement of positive law is not clear.²⁰⁸

The key takeaway from this discussion is that, while the ILC considered the question of a state requesting another to take a countermeasure on its behalf, it only did so in the context of countermeasures in the general interest. That is, it was working from the assumption that the obligation that had been breached was erga omnes, and was owed to all of the states involved: the injured state (where there was one) and the requested state. This is crucial, because it makes clear that the entitlement of the requested state to take countermeasures in these circumstances arose from the character of the obligation breached as erga omnes (partes), and its legal interest in compliance with that obligation, and not from the request by the injured state. The requirement that the injured state consent, or request, the taking of the countermeasure by the other state was intended to limit or condition the exercise of the latter state's right. The request had a limiting function.

There is thus no support for the idea of what we call proxy countermeasures in the ILC's work.²⁰⁹ In the conceptual framework developed by the Commission to regulate countermeasures by states other than the injured state, the idea of a request was intended to condition or limit the entitlement of these states in cases where the breach specially affected one (or more) state(s). The crucial point is that any such concept was only understood to be applicable where the requested state itself had a legal interest in compliance with the obligation breached by the wrong-doing state. Nothing in the ILC's work supports the idea that a request to act could create a legal entitlement where there was not otherwise one.

2. The Lack of Basis in State Practice

Aside from the statements of a handful of states in their legal positions on international law in cyberspace, the legal value of which is still uncertain,²¹⁰ it is difficult to find practice of states taking a countermeasure on behalf of another where they have no independent entitlement to take one themselves, or wider statements of support for such a proposition.²¹¹ Indeed, this has been put more strongly. Writing in 2018, Isabella Brunner argued that “[t]here currently seems to be no state practice or opinio juris supporting the notion that [such] countermeasures without any legal interest would be in compliance with international law.”²¹² In this Section, we discuss two possible examples—responses to the Iran Hostage Crisis and a claim relating to looted Albanian Gold after World War II. In the end, it is doubtful that either provides any real support for the proposition.

The first arises in the aftermath of the Iran Hostage Crisis in 1979/1980 in relation to Iran's serious breaches of diplomatic law. Obligations arising under diplomatic law are ordinarily understood as bilateralizable, even when arising under a multilateral treaty.²¹³ In response to the occupation of the United States’ Embassy, a statement of the nine foreign ministers of the member states of the European Economic Community announced that:

The Foreign Ministers of the Nine, deeply concerned that a continuation of this situation may endanger international peace and security, have decided to seek immediate legislation where necessary in their national Parliaments to impose sanctions against Iran in accordance with the Security Council Resolution on Iran, dated 10th January 1980, which was vetoed and in accordance with the tenets of international law.²¹⁴

The measures included a ban on any new export or services contracts with persons or organizations in Iran, reduction of embassy staff in Tehran and the number of Iranian diplomats accredited in each country, the reintroduction of travel visas, and arms-sales related bans,²¹⁵ as well as the suspension of all contracts concluded after the taking of the hostages.²¹⁶ At least some of the Nine subsequently did impose sanctions on Iran.²¹⁷

It is difficult to say whether this practice supports the idea of proxy countermeasures, but the better view is that it does not. First, it is not altogether clear that the proposed and/or implemented measures would have infringed or did infringe obligations owed by these states to Iran under international law—that is, it is not clear that the measures imposed would need to be legally justified in the first place.²¹⁸ Second, the underlying wrongful acts by Iran entailed breaches of different kinds of obligations—both ordinary bilateralizable obligations under diplomatic law but also the taking of hostages and the deprivations of liberty that are likely breaches of obligations owed erga omnes, and potentially the use of force.²¹⁹ Third, and relatedly, the statement of the nine foreign ministers is focused on the release of the hostages, but also notes the collective interest in “international peace and security” when explaining the imposition of sanctions.²²⁰ For these reasons, this example provides at best equivocal, and probably doubtful, support for the assertion of an entitlement to take countermeasures absent any legal interest in the other state's compliance with the underlying obligation.

The second example arises out of the events at issue in the Monetary Gold case before the ICJ. At issue was a quantity of gold, looted by Germany from Rome in 1943, later recovered, and then found to belong to Albania.²²¹ Under the provisions of the Paris Agreement on Reparation of 1946, the United Kingdom, the United States, and France composed a Tripartite Commission tasked with returning the gold to its original owners. On the basis that Albania had failed to pay the compensation ordered in its favor in the Corfu Channel case, the United Kingdom sought to have that judgment satisfied from the gold. The three states composing the Tripartite Commission agreed, subject to certain conditions, including whether Italy could provide proof that it had better title to the gold. In his assessment of this practice, Akehurst proposes that it might be understood as a “clear case of reprisals by third States,”²²² given the real argument was that “Albania's illegal refusal to pay the damages awarded in the Corfu Channel case to the United Kingdom justified not only the United Kingdom, but also France and the United States, in committing what would normally have been an illegal act.”²²³

On reflection, however, this practice does not provide support for a general doctrine of proxy countermeasures. First, the basis of the United Kingdom's claim is extremely narrow—it is specifically about actions in response to the non-settlement of a compensation award by the ICJ. This is evident in Sir Gerald Fitzmaurice's pleadings before the Court:

Not a penny of those damages . . . has in fact been paid. . . . This has been a matter of great regret to my country . . . [but] must also, on wider grounds, be a matter of importance to all members of the family of nations—whose relations are governed by international law—that the judgments of the highest international tribunal, as indeed of all tribunals, should be respected and carried out.²²⁴

Even if the argument were accepted, which remains a matter of dispute, this is a claim about conduct in response to non-compliance with an international judgment specifically. Besides, as a matter of policy, where there is such a judgment the risks of abuse by third states—an important factor that we will return to in Part V—are diminished.²²⁵ Second, the argument also relied heavily on shared or common interests:

It cannot fail to be prejudicial to the international community and to the rule of law in international relations if the judgments of international tribunals, and particularly of such a tribunal as the present Court, are contravened or disregarded. It would be right to say . . . that not only must such an occurrence be a matter of concern to all members of the international community, but also that all countries are, if not bound, at any rate entitled to take all such reasonable and legitimate steps as may be open to them to prevent such an occurrence, and either individually or by common action to do what they can to ensure that judgments, particularly of this court, are duly implemented and carried out.²²⁶

This passage clearly foreshadowed later developments—led, in part, by Fitzmaurice himself during his tenure as ILC special rapporteur on the law of treaties—concerning multilateral obligations and erga omnes obligations in international law.²²⁷ Whether such a collective interest could (and can, today) be said to exist is a matter of debate.²²⁸ But at the very least this emphasis casts doubt on the precedential value of this case for proxy countermeasures. On the contrary, if this provides evidence of any state practice at all, it is arguably in favor of countermeasures in the general interest. It may be worth noting, too, that after the United Kingdom could not find any Albanian assets within its jurisdiction, it did not request any of its allies to confiscate assets available within theirs—as Fitzmaurice had suggested, on behalf of the United Kingdom, would be possible.

In sum, these two instances, though interesting, do not provide support for a general claim about the permissibility of proxy countermeasures.

3. Two Faulty Analogies: Collective Self-Defense and Non-state Actors

There are two analogies that might be explored in this context that may ground the idea of proxy countermeasures. The first is the analogy with collective self-defense; the second is with the way that states may request private actors to act on their behalf.

To start with the first—collective self-defense—the analogy goes like this. If states other than the injured state(s) are permitted to use armed force at the request of an injured state, why should such states not be permitted to take countermeasures at the request of an injured state?²²⁹ But this analogy is faulty. It is faulty because the prohibition on the use of force is widely understood to generate obligations erga omnes.²³⁰ And, indeed, it was understood in this way by the ILC, when raising collective self-defense in discussing the possibility of countermeasures by states other than the injured state.²³¹ As such, all states have a legal interest in compliance with the obligation and the right of collective self-defense may thus be rationalized as a specific manifestation of states’ legal interest in compliance with that obligation.²³² The request by the injured state in the collective self-defense context is playing a limiting role, rather than an expansive one.²³³ The request, that is, does not create the entitlement to act in self-defense: it limits that entitlement.

The second analogy focuses, instead, on the fact that states may (and do) request private actors to act on their behalf. For instance, no one would doubt that an injured state could engage a private corporation—say, a technology company—to hack into the targeted state's networks as a countermeasure, of course subject to all of the ordinary conditions. If the injured state can engage a technology company to act on its behalf, why can it not engage a collaborating state to do the same?

To frame the problem this way, however, overlooks a crucial difference: that the collaborating state is bound to respect the targeted state's sovereignty (and any other applicable rules), an obligation which may be implicated by the operation.²³⁴ The private entity is not bound to respect the targeted state's sovereignty, so its own responsibility, as a general matter, is not really at stake at all so far as international law is concerned. However, under international law, its acts can be attributed to the injured state. The private conduct could be attributed under ARSIWA Article 5, insofar as the entity was exercising elements of the injured state's governmental authority. It could also be attributed under ARSIWA Article 8, insofar as the act would be instructed, directed, or controlled by the state. As to the putative wrongfulness of that conduct, the injured state's justification bites here, precluding the wrongfulness of the conduct. Thus, the potential violation of the targeted state's sovereignty by means of the cyberoperation would be attributed to the injured state, which could then rely on its entitlement to take a countermeasure to justify its prima facie breach of the targeted state's rights. But the same is not true for a collaborating state: for, unlike the private entity, it is itself bound to respect the targeted state's sovereignty. It would remain responsible for its conduct—the operation—in breach of its own international obligations toward the targeted state. Where the private entity can hide, as it were, behind the veil of the injured state's sovereignty (and thus entitlement), the collaborating state cannot do the same. The analogy with a request to a non-state actor does not work.

4. Agency or the Transfer of a Justification?

Finally, it may be asked whether there is some other way in which the collaborating state might act lawfully when taking a countermeasure at the request and on behalf of an injured state. An initial difficulty here is that there is as yet little assessment of how this idea of a proxy countermeasure is to be understood in legal terms rather than as a simple description of the phenomenon. As noted above, one possibility is the exceptional case of secondment recognized in ARSIWA Article 6.²³⁵ Here, by placing its organ at the exclusive disposal of the injured state, the collaborating state drops out of the picture. The relevant conduct is the injured state's conduct, on which its own justification bites. To reiterate, however, as we noted above it is not clear that this rule would hold in conflictual scenarios.

Beyond this exceptional case, we might consider two other possibilities: the first, a wider analogy to the law of agency, and the second, the transfer or vesting of a legal power. As to the first, Angelo Sereni, writing in 1940, surveyed a range of practice under which states empowered other states to act for them. These instances included general authority to act in international relations, as well as more specific authority to act in particular fields, such as economic relations or diplomatic protection.²³⁶Although Sereni thought that there were no sufficiently determinate rules of international law governing agency, he suggested that no “objection has been raised in principle by any member of the international community to the resort to agency in international relations.”²³⁷

In legal terms, as with the domestic law of agency, the relevant conduct of the agent is treated by the legal order as the conduct of the principal.²³⁸ In our situation, the principal's conduct would be that of the injured state, which raises the possibility that the injured State's own justification would render the conduct in question lawful. The difficulty with the argument is that it fails to properly consider the position of third states—that is, third states to the agency agreement. In our case, this is the targeted state, who is the target of the countermeasure taken by the collaborating state (putatively) acting as an agent of the injured state. Sereni argued, in the first place, that notification of the arrangement to third states was required. Second, he suggested that it would then be up to third parties so notified to “proceed to its formal recognition”’ if they so wished.²³⁹ This seems to us to be right as a matter of principle. If this is true, agency would not apply in the hostile situations we are examining: the targeted state would need to be notified of the arrangement as to the taking of the countermeasure by the collaborating states, and have no objections to it.

The second possibility is whether international law recognizes a doctrine by which the injured state could vest, or transfer, the exercise of its entitlement to take countermeasures onto the collaborating state.²⁴⁰ In legal terms, it would not be the case, as in the previous paragraph, that the conduct of the collaborating state is treated as conduct of the injured state, but rather that the justification would be transferred from one to the other. Here, some care is needed. It is true that legal orders recognize a variety of doctrines pursuant to which one party can transfer or vest a right or power in another. However, the question here is much more specific—and concerns the putative transfer of a defense such that the collaborating state has a justification for infringing its own obligations vis-à-vis the targeted state. Put in this narrower way, it is hard to find any support in even roughly analogous situations in domestic law. Although a full survey of all domestic orders is not possible, it is doubtful that in contract law, for instance, such a possibility would exist. It would entail that, absent specific agreement to the contrary, a breach of a contractual obligation owed by X to Y, which might justify some remedial action by Y, could entitle Z not to comply with its own, independent, obligations to X just because Y requests that they not comply. Indeed, much as in international law, it is doubtful that Z would even have standing to bring a claim in relation to the breach.²⁴¹

F. Conclusions on Collaboration

Drawing together this discussion, the following conclusions may be drawn. For independent but coordinated action, as well as joint action, the collaborating state is responsible for the relevant conduct. As such, given it has no entitlement of its own to take a countermeasure, it acts unlawfully. Similarly, international law does not recognize the possibility of what we have called proxy countermeasures—where a state takes such a measure at the request and on behalf of an injured state. That leaves two scenarios—secondment and assistance. As to secondment, as we argue above, under the rule in Article 6 this would be a permissible form of collaboration, even if that term is slightly misleading—the legal order treats the relevant acts as acts of the injured state only. This, however, would be a rare case, and doubts arise as to whether it would really hold up in conflictual situations such as this. For these reasons, assistance becomes a crucial question. As argued above, it is plausible to simply take the view that a collaborating state that assists another state's countermeasure acts lawfully, for what it is doing is simply assisting in the commission of an act whose wrongfulness is precluded. However, certain essential features of the institution of countermeasures—their tolerable rather than desirable quality and their personal character—put pressure on this conclusion.

IV. Collaboration Where the Collaborating State Does Have an Individual Entitlement to Take a Countermeasure

For completeness, this Part addresses collaboration in the taking of countermeasures where both—or all—of the collaborating states have their own entitlement to take a countermeasure. It will do so briefly, as for the most part it does not raise particular problems. In all five instances of collaboration, assuming other requirements are met, the collaborating states act lawfully.

First, as to independent but coordinated action, each state remains responsible for their own conduct—the distribution of the confiscated assets, the suspension of overflight rights, and so on. Each state's own justification precludes the wrongfulness of its own conduct.

Second, as to secondment, the situation is to be analyzed slightly differently. To reiterate the rule set out in ARSIWA Article 6:

The conduct of an organ placed at the disposal of a State by another State shall be considered an act of the former State under international law if the organ is acting in the exercise of elements of the governmental authority of the State at whose disposal it is placed.²⁴²

In Part III above, we discussed the requirements of this rule, its likely use in practice, and certain doubts as to whether it would really operate as intended in conflictual scenarios. But for present purposes, those questions can be left aside. Whether the conduct of the seconded unit is attributed to the receiving state—as set out in the rule—or not, the entitlement of both or all collaborating states to take a countermeasure renders that conduct justifiable.

Third, as to joint action, the conduct is attributed to each of the states participating in the joint operation. Part III, above, explored the boundaries of joint responsibility of this kind under international law. For present purposes, the key point is that each of the states to whom the conduct is attributed possesses its own independent justification for its conduct. Each state's entitlement to take a countermeasure will provide a legal basis for the permissibility of its measure.

Fourth, there is the question of assistance by one state to another state in the taking of countermeasures. In Part III, above, we set out certain rules that regulate interstate assistance. For the purposes of this Part, the key point is that the assisting state's individual entitlement to take a countermeasure can justify its putatively wrongful acts of assistance. Aside from the limits imposed in ARSIWA Article 50 and the requirement of proportionality,²⁴³ and absent specific rules to the contrary, states can take measures that affect any obligations owed to the targeted state. Any potential breach of obligations not to assist another state in committing wrongful acts would be justified by the collaborating state's own entitlement to take countermeasures: it can, as a countermeasure, choose to disregard these particular obligations. In a less doctrinal sense, this conclusion also follows from the collaborating state's own entitlement to respond to the targeted state's breach by taking a countermeasure of the kind taken by the assisted state. To spell that out, if the collaborating state could have lawfully undertaken that conduct, and it could lawfully have undertaken that conduct jointly with the other state, it must be the case that it can assist the other state too. This is a situation where permissibility of the lesser—assisting another state—is rightly implied by the greater—the collaborating state's independent entitlement to take a countermeasure.²⁴⁴

Fifth, and finally, there is the question of collaboration at the request and on behalf of another state. Here, the terminology of “at the request and on behalf of” another state does not appear accurately capture the legal relationships at play. To reiterate, we have assumed for the purposes of this Part that the collaborating state has its own entitlement to take a countermeasure against the targeted state. In a colloquial sense, the state may be doing so at the request and on behalf of the other state, but legally what it is doing is exercising its own independent entitlement to take a countermeasure to uphold a collective interest.²⁴⁵ Any such request is not doing any legal work.

V. Policy and Potential Developments

A. Recapitulation

Where states have an independent entitlement to take countermeasures, they can permissibly collaborate with an injured state (or with other states so entitled) in taking countermeasures. The scope of who is so entitled is thus the central question. As we set out in Part II, despite the increasing weight of practice, the lawfulness of countermeasures in the general interest remains a difficult question.

Where states are not entitled to take countermeasures, the permissibility of collaboration varies depending on the form that it takes. For the most part, in our analysis of the relevant rules, states that are not entitled to take a countermeasure cannot collaborate with others in doing so. Specifically, independent but coordinated measures, joint measures, and what we have called proxy countermeasures are impermissible. Two forms of collaboration entail different answers: secondment and assistance. First, as to secondment, any such conduct of the seconded unit is treated as conduct of the receiving state, a state whose own justification to act thus applies. As we note above, technically speaking this is not really collaboration at all as formally the seconding state is not acting in a legal sense, and we are not certain the rule on secondment would be accepted in a conflictual scenario. Second, as to assistance, this is the most practically relevant and difficult scenario to analyze. We show above that while it is plausible to think assistance to another state's countermeasure is permissible, certain essential features of the institution of countermeasures point in the opposite direction.

An initial question is whether these overarching conclusions cohere with other aspects of the legal order. For the most part, we think they do. As a starting point, distinctions that follow from the different positions of individually entitled/non-individually entitled states make sense. Where the collaboration breaches an obligation of the collaborating state toward the target state, and the conduct is attributable to it, then it will be liable for its own act. That a collaborating state acts impermissibly in this case, where it has no entitlement to take a countermeasure of its own, is consistent with the principle of independent responsibility and its extension to defenses: like responsibility, defenses, too, are individual. One may query, however, the varying legality of different forms of collaboration for a non-entitled state—in particular given that the lines between forms of participation will often be a matter of degree. The line between joint action and secondment, for example, may be blurred in practice, and the permissibility of the collaborating state's conduct will vary depending on which side of the line it falls. Similarly, the analysis above places immense pressure on the line between assistance and joint conduct. Where the non-entitled state's collaboration remains within the realm of assistance, it is possible to argue that it acts lawfully; where it becomes joint conduct, it does not.

B. Evaluation

This leads to the question of whether the current position is desirable. As a starting point, it is worth emphasizing the context of these developments: one of increased polarity in international affairs and geopolitical conflicts between powerful states, and between these and smaller states. These developments have manifested themselves in, among others, instances and threats of aggression against smaller neighbors, cyberoperations between great powers and against weaker states, and the continued use of economic pressure and coercion as an instrument of foreign policy. In all these instances, it is usually smaller states who stand to lose the most, without any obviously effective remedies. In many situations, too, there may be spillover or collateral effects on a broader number of states.

Inevitably the question turns to what can be done to stop these harms or threats, or at the very least minimize their impact. In the case of aggression, the victim state is entitled to react in self-defense and can call on its allies to act in collective self-defense. But it may not be always in the allies’ interest to participate in self-defensive action. In addition, cyberactivity tends to remain “below the threshold”—that is, it falls below the threshold of an armed attack as a trigger for the lawful exercise of self-defense by the victim and its allies,²⁴⁶ and many forms of economic coercion are not usually understood to be prohibited.²⁴⁷ While these are not as jarring as armed aggression, both malicious cyberoperations and economic coercion can have devastating financial and humanitarian consequences in the targeted state.²⁴⁸ In all cases, due to material inequalities, the victims’ responses may be ineffective in inducing compliance by the wrongdoer, even when they involve the taking of measures that would be in breach of the rights of the target state. In the light of this context, states have begun to articulate and propose policies and mechanisms of joint response, so as to be able to collaborate with allies in more effectively counteracting these harms. It is thus no surprise that the strongest proponents of collaboration in the taking of countermeasures currently include Ukraine, the victim of aggression, and Estonia and Costa Rica, both of which have been victims of wide-ranging and damaging cyberoperations.²⁴⁹

This context, as we mentioned earlier, has brought to the surface an old worry about the institution of countermeasures. This is that countermeasures, and perhaps unilateral enforcement more generally,²⁵⁰ can entail a double disadvantage for smaller and developing states.²⁵¹ These are the states that, historically, were the primary victims of abuse of the institution of reprisals (the ancestor, as it were, of today's countermeasures) by powerful states. At the same time, these states are also less likely to be able to resort to countermeasures, and to do so effectively: material differences mean that these states are less likely to have the means, or to have the weight, to put sufficient pressure on larger and more economically powerful states to induce them to comply with their obligations. As Simma put it, international law enforcement reflects an “every-man-for-himself” paradigm which, while in keeping with state sovereignty and the principle of equality, “unveils, and even endorses, the crucial dependence of the bilateral enforceability of a State's international rights upon a favourable distribution of power.”²⁵²

The cyber-context provides a distinctive manifestation of this general point. In its recent statement on the law applicable to cyberspace, Canada, for example, noted that collaboration in the taking of countermeasures can be provided “where the injured State does not possess all the technical or legal expertise to respond to internationally wrongful cyber acts.”²⁵³ For Ireland, collaboration was particularly relevant in the cyber context given that some states “lack the technological capacity” to respond to cyberoperations alone.²⁵⁴ This point is similarly emphasized in the literature in support of cooperation in the field of cyber-countermeasures.²⁵⁵ Moreover, it is evident too in the discussions surrounding sanctions against Russia, and the desire to come to the assistance of Ukraine in achieving reparation: there are simply not sufficient Russian assets within Ukraine to meet the extent of the reparative obligation. Turning back to our collaborative scenarios, this all suggests a policy imperative to widen the set of situations where states can collaborate in the taking of countermeasures. Indeed, if the institution of countermeasures is the only equally available legal mechanism for the enforcement of international law, then why not facilitate it?²⁵⁶

Allowing allied participation, especially by powerful states, in the countermeasures of other states is, however, not without objections and risks. First, the expansion in the number of potential participants may also increase the likelihood of escalation of disputes and tensions—something inherent in self-help measures of this kind. In the context of increasing polarization in international relations, it is not implausible that powerful states could use harm and injury caused to others as a pretext to target each other, as well as less-powerful allied states. Instead of the proxy wars of the 1970s and 1980s, this may result in proxy-conflict of lesser intensity (at least, lesser kinetic intensity)—but one that is still harmful and damaging to the parties involved, and to the international legal order. Such risks may be heightened in the cyber-context, where attribution of the initial wrong may be difficult, where hasty reactions are often needed, and where covertness and secrecy are key to effectiveness. But these risks are pervasive across all the measures that states might undertake in this context: after all, countermeasures are usually taken in the absence of third-party determination of the existence of the prior wrong.

Second, although pitched as a way to correct the power imbalances that hamper smaller and developing states’ ability to resort to countermeasures, the reality is that the central difficulty of the distribution of power among states is not quite so easily contained, nor so easily resolved. On the contrary, certain solutions may end up further entrenching countermeasures as an enforcement mechanism for the few. In this respect, reliance on powerful states to collaborate in the taking of countermeasures means that, once again, it is these states that will end up taking the measures or, at the very least, that countermeasures cannot be taken without them. Put differently, to allow collaboration in these circumstances may be another way to license powerful states to disregard their obligations, or to facilitate the disregarding of such obligations by others.

Third, there is a risk of a too-idealistic picture of the reasons for which, and situations in which, powerful states would be willing to collaborate with another state in taking a countermeasure. In some situations, these reasons may have less to do with a desire to ensure that the injured state's rights be upheld, and more to do with their own interests.²⁵⁷ Writing in 2002 on what he called solidarity measures, Koskenniemi argued that given there is no suggestion of a duty to act—unlike in respect of a domestic police force—powerful states could create “a world order of their liking by choosing between violations they enforce and violations they do not, as well as deciding on the manner and intensity of their reaction.”²⁵⁸ Koskenniemi's point applies with as much force to questions of collaboration.

C. Moving Forward on Collaboration

As states continue to formulate their positions on countermeasures and react to other states’ breaches, the following considerations become key. First, there is a terminological issue. In the coming period, the issue of collaboration in the taking of countermeasures is likely to become more central in the operational practice of states, in their articulation of legal positions, and in the scholarly literature. In this respect, greater precision in describing what exactly is being done and/or proposed would assist with both doctrinal analysis and policy evaluation. This includes distinguishing between different forms of collaboration, but also the key issue of whether the term “collective countermeasures,” ordinarily used to refer to countermeasures in response to the breach of erga omnes (partes) obligations, extends to situations where the collaborating state has no independent entitlement to take a countermeasure. As we set out above, we propose the use of the term proxy countermeasures to draw attention to the distinctive issues at play in that situation.

Second, the analysis above shows that a fundamental question for assessing the legality of state collaboration in the taking of countermeasures remains the question of the permissibility of countermeasures in the general interest. As we discussed in Part II, practice continues to increase, even though there remains a question of its representativity. The real issue, however, is opinio juris. To the extent that states desire a permissible regime for collaboration in the taking of countermeasures, the first step would be to make clear their understanding of the lawfulness of countermeasures in the general interest, and the conditions for their use.

Third, and relatedly, to the extent countermeasures in the general interest are permissible, the next question becomes one of determining exactly which obligations are owed erga omnes or erga omnes partes. This question has received comparatively less attention in the practice and scholarship,²⁵⁹ but it is critical. In this respect, the core is generally accepted—aggression, genocide, slavery, and racial discrimination and self-determination of peoples.²⁶⁰ But what of the principles of sovereignty and non-intervention?²⁶¹ These would traditionally be understood as entailing bilateral relations between states, but this understanding could come under pressure to the extent states wish to respond collectively to breaches of these obligations by other states, especially in the context of cyber-space. To widen the pool of multilateral obligations is to widen the situations in which states could permissibly collaborate in response to a breach (subject, of course, to the acceptance of countermeasures in the general interest).²⁶²

Fourth, states may wish to clarify or develop the rules applicable to collaboration specifically, particularly insofar as a state without an entitlement to take countermeasures wishes to collaborate with an injured state. In this respect, we suggest that the idea of proxy countermeasures is not an appropriate approach. In addition to distinctive risks of abuse and escalation, this would also entail systemic incoherence in the legal order. It would grant to a state the right to take a countermeasure where it has no legal interest of its own in compliance with the underlying obligation, and would simply bypass the deep disagreement amongst states about the permissibility of countermeasures in the general interest and argument about which obligations are owed erga omnes. By contrast, it is in relation to assistance where there is room for confirming or developing the law. At present, as we argued earlier, it is not clear that assistance in these circumstances is permitted in relation to the rule in Article 16 ARSIWA. Moreover, we raised above certain considerations of principle that militate against the appropriateness of assistance to another state's countermeasure in these situations. But it is possible that the value of being able to come to another state's aid may ultimately outweigh those concerns.²⁶³ Crucially, this assistance could not cross the line into joint responsibility such that the assisting state would then need its own justification, or breach a discrete primary obligation of the assisting state.

Fifth, and finally, the rules we discuss above apply by default—they are part of the general law of responsibility. By treaty, and subject to certain overriding constraints, states may set up particular regimes regulating their relationships, including by providing for distinctive rules multilateralizing responsibility or processes for third-state enforcement. If it is the case that the general rules do not provide workable solutions for a particular subfield—as is sometimes suggested for aspects of responsibility in cyberspace—it is open to states to vary those general rules by specific agreement.²⁶⁴ This may include distinctive rules relating to collaboration in the taking of countermeasures.²⁶⁵ In the long term, that might be a better solution than seeking to accommodate all of the needs of any particular subfield within the relatively stable core of general international law.

VI. Conclusion

The question of collaboration in the taking of countermeasures is bound up with the wider debate about the promises and risks of unilateral enforcement measures in a decentralized legal order. There is a compelling demand for effective enforcement mechanisms, and a compelling need to restrain their abuse. These considerations have accompanied the institution of countermeasures for decades and were central to the ILC's codification efforts in this area. They are now re-emerging in a different political context, and by reference to different features of the regulation of countermeasures in international law. The new political reality of international affairs may demand that the balance arrived at in the ARSIWA regime of countermeasures in 2001 be revisited, possibly in the direction of unilateral enforcement, including through collaboration. Writing on the tail-end of the Cold War, a period of similar polarity and fracture in international affairs, Jonathan Charney noted that “[t]hird State remedies may appear at first glance to serve only the desirable goal of promoting rules of international law, but they may also produce negative side effects.”²⁶⁶ This warning remains as relevant today.