Introduction

International law is in the midst of a widely proclaimed paradigm shift. On the one hand, the international legal system continues to be understood in “classical” terms—with the state as the principal (if not exclusive) actor and the international legal regime organized around the interests of hermetically-sealed territorial units.[1] These interests are conceptualized in terms of sovereign equality, territorial integrity, and non-intervention and lend themselves to a discourse of exclusive and absolute state right or authority over particular territories. The competing—or “new”—paradigm better reflects the realities of modern globalization (including the ease with which people, goods, weapons, and even criminality move across borders), and is significantly more pluralistic in its recognition of relevant actors and “human-centric” in its conception of the international legal order. Within this paradigm, the legal regime is (or at least ought to be) deployed principally in defence of human rights (including, of particular relevance for present purposes, rights to physical integrity) and human security. As a result, sovereignty is increasingly conceptualized in relation to these interests—defined by reference to the obligations associated with international human rights law (IHRL). Professors Criddle and Fox-Decent’s conceptualization of this new paradigm in fiduciary terms portrays with much elegance the trends of at least the latter half of the twentieth century.[2]

This is not, however, a straightforward story of one paradigm completely displacing another. As with all in-progress paradigm shifts, the normative pull of each paradigm is not entirely decisive, and even within a single “paradigm” there can be serious tensions. Of interest for present purposes, respect for human rights is often understood to be in tension with a robust approach to protecting human security (both within a single state’s territory and across territorial boundaries). International law struggles with managing any real or imagined tension between these potentially competing interests.

Principles like those which form the basis of the fiduciary theory of sovereignty—such as non-instrumentalization and non-domination—may suggest an approach to balancing these competing interests,[3] but perhaps not always with the specificity and detail required of particular legal contexts. The “fiduciary” balance also does not necessarily reflect the lex lata. For instance, Criddle and Fox-Decent argue that a fiduciary approach to sovereignty suggests that a “least harmful means” requirement be incorporated into the “proportionality” calculus of international humanitarian law (IHL),[4] as discussed further below. This argument has also been advanced on more positive law grounds, but is not widely accepted by states.[5] Criddle and Fox-Decent further suggest that states should refrain from using force against non-state actors (NSAs) abroad if such force would pose a disproportionate threat to international peace or security.[6] This understanding of the fiduciary approach to the jus ad bellum is of particular interest insofar as it seems to balance, and treat as equivalent, a state’s fiduciary obligations to its own population (in particular to protect that population from armed attacks emanating from foreign territory), and its fiduciary obligations to the international community as a whole (as distinguished from its fiduciary obligations to the individuals within the territorial state from which the armed attacks are emanating and in which it is using defensive force). While this approach to the jus ad bellum, in its emphasis on a state’s fiduciary duties to its own population, reflects the lex lata,[7] its emphasis on other-regarding obligations requires further nuance.

This article seeks to explore an alternative route to balancing these competing interests in a very particular context (set out in Part I below)—one which draws on positive international law. The approach adopted in this article resonates with the fiduciary theory of sovereignty insofar as the theory is interpretive, but looks to positive law to do the prescriptive heavy lifting with a view to (at least some) specificity and effectiveness in defining states’ obligations when their conduct affects competing interests.

I.  La Problématique and the Inadequacies of the Fiduciary Theory

This article explores some of the tensions within (and across) international law paradigms—taking the quintessential threat to both human rights and human security, in the form of armed conflict, as its case study. The outbreak of an internal armed conflict (or non-international armed conflict (NIAC)) creates new risks for the local population,[8] but also for the populations of third-party states—most particularly by undermining civic order and giving armed actors the physical space and opportunity to launch armed attacks against the populations of third-party states. The NIAC, however, often significantly impairs the territorial state’s capacity to discharge its “fiduciary” duties to its own population and to the broader international community. Understood in terms of the fiduciary model, this increasing mismatch between what is required of the territorial “fiduciary” and its ability to comply with its obligations highlights the importance of the “fiduciary responsibilities” of third-party states.

This article will focus on states which provide arms, funding, or other support to those engaged in a foreign NIAC[9]—in cases where the intervening state’s support is in professed discharge of the human rights obligations owed to its own population. In particular, a state has obligations to both respect and ensure respect for the physical integrity rights of those subject to its jurisdiction.[10] Where those rights are threatened from abroad, a state is therefore bound by human rights law to take feasible measures to protect the rights of its domestic population. The particular case addressed here is intervention in a foreign NIAC, where that intervention is a necessary and proportionate response to a threat (defined in terms of an “armed attack” by reference to Article 51 of the UN Charter[11]). While supporting NSAGs or foreign governments in a NIAC abroad may, consistent with a state’s human rights (and indeed fiduciary) obligations to its own population, protect that population from armed attacks emanating from abroad,[12] the implications for the interests of the various groups within the territorial state are complex and potentially grave.

Criddle and Fox-Decent focus on the fiduciary responsibility to establish “secure and equal freedom”.[13] In a NIAC, arming the government or a NSAG potentially helps achieve such freedom, for example by increasing the state or group’s ability to defend themselves/others, and to achieve military success, which may bring the conflict to an end and/or advance the freedom of at least part of the population. This will particularly be the case where the support is for a party engaged in an armed conflict against an oppressive government or NSAG. But making parties to a conflict more effective (and more dangerous) in their war-waging might also have the exact opposite consequences for other parts of the population (or indeed for all within the reach of the armed conflict if it exacerbates the hostilities). Responsibility for balancing the interests of groups within the local population on the fiduciary model is usually, in the first instance, borne by the territorial state, but the NIAC often means that the government of the territorial state is incapacitated and unable to meet its responsibilities—indeed it may be the source of the danger to at least part of its population.

As to the fiduciary responsibilities of an intervening state—these require balancing its own national security interests (which gives rise to the decision to arm or otherwise support governments or NSAGs in a foreign NIAC in exercise of the right of self-defence) against the very different interests of the territorial state’s local population (which themselves call for complicated balancing among different groups). And consistent with the fiduciary model’s recognition that states owe duties directly to individuals outside their territory, positive international law is (or can increasingly be) understood as imposing obligations on intervening states, owed directly to the territorial state’s affected population in a NIAC. These obligations can apply whether the support is provided to the territorial government or to NSAGs.

Specifically, we argue that international law, in particular IHL (informed by emerging interpretations of relevant treaty obligations), lex lata developments in respect of IHRL, and the law of state responsibility and arms control, require intervening states to prevent, or at least not to increase the risk of, serious IHL violations by the beneficiaries of their support. IHL is the basis for identifying the minimum protection to which all groups within a population subject to armed conflict are entitled in principle, and intervening states are required by international law to do everything feasible to ensure that that protection is not undermined through their intervention and support of armed actors.

At this stage, it is worth noting that Criddle and Fox-Decent suggest that IHRL is the minimum yardstick against which both the territorial state’s and intervening state’s behaviour (particularly in cases of humanitarian intervention and asymmetrical self-defence, each as discussed further below) is measured in NIAC situations.[14] This is presumably based on an argument that IHRL is not wholly displaced by IHL in situations of armed conflict—a principle which has indeed been clearly established by the International Court of Justice (ICJ), the regional human rights courts, and other treaty bodies.[15]

While it is without a doubt the case that IHL as a regime does not displace IHRL as a regime, it is also the case that certain IHRL norms are interpreted in light of IHL in situations of armed conflict (e.g. the right to life). In respect of the right to life, the lex specialis IHL norm displaces the IHRL norm, even though IHL does not displace the IHRL regime as a whole.[16] We therefore do not proceed on the basis that IHRL provides the minimum yardstick by which conduct in armed conflicts is measured—at least not in respect of physical integrity rights. Having said which, IHL and IHRL norms very often overlap significantly—particularly with respect to humane treatment.

Criddle and Fox-Decent also address the intervening state’s obligations to the territorial state’s population. They do so in respect of both humanitarian intervention (HI)[17] and asymmetrical self-defence, and characterize the intervening state in both cases as a surrogate sovereign.[18] In respect of HI, Criddle and Fox-Decent posit that intervening states are under very particular fiduciary obligations in virtue of their intervention on behalf of the territorial state’s population. In such cases, Criddle and Fox-Decent argue that the fiduciary is effectively acting in the place of the territorial sovereign, and must therefore use coercive power for the benefit of the territorial state’s population.[19]

While a state might be supporting NSAGs in a foreign NIAC for humanitarian purposes, and thereby assume the fiduciary obligations of a “surrogate sovereign”,[20] this article explores the situation of support for armed actors in a foreign NIAC when that support is dictated primarily by the intervening (and supporting) state’s own security (framed in terms of Article 51 of the UN Charter[21]). Even accepting that, in the HI context, an intervening state’s relationship with the population of the territorial state is similar to the relationship between a state and its own domestic population during an armed conflict, this will not be the case in the event of intervention abroad (direct or indirect) for the purposes of protecting security at home (asymmetrical self-defence). This latter case involves competing fiduciary obligations for the intervening state to its local population and to the population of the territorial state. The intervening state is not intervening in the interests of the territorial state’s population; rather it is intervening in its own population’s interests, even if it bears obligations to the territorial state’s population. Characterizing the intervening state, in the context of asymmetrical self-defence, as a surrogate sovereign is unrealistic. Doing so suggests a level of obligation not sustained by positive law and unlikely to be effective in practice. It glosses over the tension between the human rights and security interests of the intervening (and supporting) state’s local population, and the human rights and security interests of the territorial state’s population.

Criddle and Fox-Decent, in positing that intervening states in cases of asymmetrical self-defence are surrogate sovereigns, and arguing that intervening states are subject to human rights obligations in respect of the population of the territorial state, fail to acknowledge or engage with the competing human rights and security interests at play. The analysis in this article takes a different approach, although it certainly intersects with elements of the fiduciary theory. In particular, our starting point is that states cannot comply with their human rights obligation to protect their domestic population against armed attack at the expense of the fundamental protections to which the territorial state’s population is entitled. Having said so, those fundamental protections and the intervening state’s obligations in respect of them are defined in reference to the applicable body of law: IHL. IHL recognizes that rights to physical integrity during an armed conflict are not absolute—and that such rights will be balanced against military advantage through the well-known principle of proportionality.[22] IHL, together with general rules of international law, also imposes responsibilities on states not party to the conflict, to support the protection of the rights which its “substantive” norms and principles recognize. This combination of IHL norms inherently balances the interests of the intervening state’s domestic population (defined in terms of the military advantage of the operations it is supporting by way of its own defence) against the rights to physical integrity of the local population subject to the armed conflict.

II. Towards “Fiduciary” Obligations of Intervening States Via the Lex Lata

In this Part, we look to positive law to further define the IHL-related obligations of an intervening state in the context of asymmetrical self-defence. These IHL obligations can, of course, be understood or explained as giving effect to and defining key elements of the “fiduciary duties” of intervening states. But we argue that they are derived from positive law. While the fiduciary theory is perhaps a useful framework for understanding these obligations (at least in part), it is not the source of the obligations. Further, although it correctly draws attention to the imperative for the full range of interests involved to be recognized, it does not address the complicated balancing of interests which is inherent in the situation explored in Part I. The proposed development of IHL norms addressed below demonstrates the potential for established international law-making processes to respond to the underlying concerns addressed by the fiduciary theory, while assuring that international law is effective in governing the behaviour of intervening states by acknowledging and indeed respecting their role in shaping that law.

A. Common Article 1 of the Geneva Conventions

The starting point for an account of positive law obligations on supporting/intervening states in this field lies in the four Geneva Conventions of 1949 and Additional Protocol I of 1977 (the Conventions). Each of these includes a Common Article 1 (CA1), by which the parties “undertake to respect and to ensure respect for the present Convention in all circumstances.”

1.    The Obligation to Respect and Ensure Respect in General

Considering the two commitments encapsulated in CA1 in turn, the obligation to respect requires that the parties themselves comply with the Conventions. This means that conduct which is attributable to a state party must be in conformity with that state’s obligations under the substantive provisions of the relevant Convention. For example, a state engaged in conflict must respect those provisions which embody the requirements of the principle of proportionality when it launches an attack. Presuming that the meaning of the word “respect” does not change within a single sentence, it follows that the obligation to ensure respect requires that the state ensure other entities with legal obligations under the Conventions comply with their own obligations.[23]

The meaning of the word “ensure” requires further elaboration here. In a range of multilateral humanitarian treaties, many of the same era as the Conventions, requirements that states “secure” or “ensure” rights, or “prevent” abuses are not interpreted as imposing an absolute obligation to successfully bring about or prevent the relevant result.[24] Like those obligations, the obligation to ensure respect should be interpreted as one of diligence, requiring reasonable efforts to bring about IHL compliance by the actual parties to the conflict.

If CA1 were to impose an obligation on all states to take such measures, whether or not they have any connection to the conflict, this would accord strongly with the fiduciary model’s insistence that all states are subject to fiduciary obligations owed to humanity as a whole.[25] Each state would, on such a view, be subject to some level of obligation in respect of the treatment of all humans affected by conflict, anywhere.[26]

Notwithstanding the reasonable clarity of the language, however, the extent to which any legal obligation is imposed by the undertaking to ensure respect (especially as regards other states or NSAGs) has proved controversial. The drafting history and subsequent practice are subject to competing interpretations.[27] An obligation on all states to ensure respect for IHL by the parties to a conflict, even tempered by the limitation to diligence measures, is considered by many implausibly onerous and unlikely to be practically effective.[28]

2.    The Obligation of Supporting States to Ensure Respect

Against this background, a number of commentators suggest that focusing on the duties of states which actively support one side or another in an armed conflict provides a more realistic compromise position.[29] Such an approach is not necessarily any less consistent with the fiduciary model than the broader interpretation; the difference is merely in the precise nature of the duties to which the state’s fiduciary role gives rise. Rather than arguing for a global obligation on all states to ensure respect by the parties to a conflict, on this view the obligation falls primarily or exclusively on those states which have to some extent involved themselves in the conflict indirectly.[30] Obligations which focus more narrowly on such supporting states may represent a position more likely to gain acceptance among states generally.[31]

More specifically, such states might be required either to use the influence provided by their support to encourage the recipients to respect IHL, or to cease providing support on account of the risk that the recipients will use it to breach IHL.[32] The first of these could be regarded as either an obligation triggered by the provision of support, or (to those who favour the broadest interpretations of the obligation to ensure respect) simply an intensification of the general obligation of diligence. By providing support, and thus being able to suspend or terminate that support in the future, a state is likely to gain influence over the recipients and have greater capacity to ensure that they respect IHL in their conduct of hostilities. Accordingly, if an obligation of diligence to ensure respect exists, it will generally require more of states which provide support[33] than of those which do not (of whom on some views it requires little or nothing at all[34]).

If, however, the risk of support facilitating violations of IHL exceeds a certain level, the duty may instead be to withhold that support.[35] Once triggered, this would be an obligation of result rather than of diligence, since clearly it is within the supporting state’s power to cease providing support. The difficulty lies in identifying how much risk is sufficient to trigger such a prohibition, and how far states are obliged to proactively investigate that risk prior to transfer of support.[36] Further, the quantification of “risk” potentially involves considering a number of different factors, such as the probability of violations, the severity of their consequences, and the importance of the norm which is at risk of violation. Further questions could then be raised as to whether, for example, a relatively low threshold of probability would apply regarding breaches of peremptory norms, and a higher threshold for other breaches.

These ambiguities are especially problematic given the competing interests at stake, which make it essential to identify when the risk of IHL breaches becomes so high as to override the considerations in favour of providing support. In cases of asymmetrical self-defence, supporting states may well consider continued support to be essential for their own security. They may also plead the legitimate security interests of the recipients of the support; the opportunities for asserting justification of the “lesser evil” in this field are significant. A particular difficulty potentially arises from the interaction of the duty to withhold with the duty to influence, as either might undermine the effectiveness of the other. A state may argue that only by providing support can it maintain the influence to moderate the behaviour of the recipients, and that overall compliance with IHL will be better, if not perfect, if support continues.[37] Proving such counter-factual assertions incorrect (or indeed correct) in any given conflict will often be difficult, to say the least.

Thus while CA1 provides a plausible basis for fiduciary obligations on outside states, and particularly those providing support, the brief textual provision requires considerable further concretization to provide a basis for effective constraint on states’ discretion in this area. This is particularly so in relation to balancing the risks associated with support against the range of competing considerations (many of which can also be regarded as reflecting fiduciary duties) which supporting states may invoke. Focusing primarily on the obligation to withhold support, the following subparts will consider the means by which positive law may assist in addressing central questions over the content of the duty in question, namely: the extent of the obligation to assess risk in advance; the threshold of risk at which support must be withdrawn; and the extent of “balancing” with competing interests which may be permitted.

B. The Non-Encouragement Obligation in Nicaragua

In its famous Nicaragua decision, the ICJ understood the obligation to “respect and ensure respect” under CA1 of the Geneva Conventions (as discussed in Part II. A above) in terms of a prohibition on “encouraging” breaches of IHL.[38] In the circumstances of the case, the obligation was one bearing on an intervening state offering military and material support to an armed group participating in a foreign NIAC, where such support was offered ostensibly by way of exercising rights of collective self-defence under Article 51 of the UN Charter. In effect, Nicaragua can be characterized as a case about asymmetrical self-defence, even though there were of course serious doubts as to the good faith of the American claim to have been acting in collective self-defence of El Salvador through its support of the contras. This obligation of “non-encouragement” is a negative obligation, distinguishable from a broader positive obligation to prevent breaches of IHL on that basis.

In evaluating whether a psychological operations manual published by the U.S. and disseminated to the contras was in breach of this obligation of non-encouragement, the Court took particular account of “whether [the] encouragement was offered to persons in circumstances where the commission of [breaches of IHL] was likely or foreseeable.”[39] As a result, CA1-breaching “encouragement” by an intervening state does not require actual knowledge of IHL breaches committed by the beneficiary of support, nor does it require tracing the material support from the supporting state to the commission of any such breaches. Instead, in assessing foreseeability, the Court noted that “those responsible for the issue of the manual were aware of, at the least, allegations that the behaviour of the contras in the field was not consistent with humanitarian law.”[40]

      Awareness of allegations as a threshold for assessing foreseeability and compliance with an obligation of non-encouragement suggests a further obligation bearing on the intervening (and supporting) state—one of informing itself of potential dangers (in the form of IHL non-compliance) arising out of its support. A failure to do so, where non-compliance is knowable on the basis of allegations, will result in responsibility. Indeed, an obligation to inform oneself, where the capacity to comply with an obligation hangs on available information, is a common feature of both IHL and general international law more broadly. For instance, in respect of the obligation to take precautionary measures—in the precise form of doing “everything feasible to verify that the objectives to be attacked are neither civilians nor civilian objects and are not subject to special protection but are military objectives”[41]—states are required to diligently develop an information-gathering capacity and to use that capacity diligently.[42] A similar obligation to keep informed can be derived from the International Law Commission’s (ILC) Draft Articles on the Prevention of Transboundary Harm.[43] Where an activity poses a risk to third-party states, the ILC Draft Articles require states to inform themselves of the dangers and risks arising therefrom.[44] These particular instantiations of an obligation to keep informed echo the general international law obligation states have to prevent their territory from being used to harm the interests of other states, in that states are bound to inform themselves of threats emanating from their territories and to diligently respond to reasonably discoverable threats.[45]

The Corfu Channel and ILC Transboundary Harm obligations for a state to inform itself can be understood as deriving from conceptions of the state as having exclusive and absolute control over its territory (from which risks to third-party states emanate). But this obligation is obviously not restricted to circumstances of control over territory in its IHL incarnations, and extends to control over the source of a threat to third-party interests. This is evident from both the obligation to take precautionary measures discussed above, and the Court’s reading of the obligation of non-encouragement derived from CA1 of the Geneva Conventions in Nicaragua, which is understood in terms of a state’s obligation to inform itself in respect of the IHL compliance of armed group beneficiaries of its support abroad. The Court applied the obligation to make inquiries to situations beyond the territorial control of the supporting state, but adhered to the principle that matters within the state’s exclusive control (for instance the existence or scope of its support) should not be allowed to develop into a threat to foreign interests.

Injecting weapons and material support into a NIAC, even when doing so is for the purposes of protecting security at home, very obviously implicates third-party interests. The logic of the Court’s approach in Corfu Channel and the ILC’s approach to managing the risks of transboundary harm is therefore easily transposed to the context of IHL and asymmetrical self-defence, and sounds clearly in a reading of Nicaragua which obliges states to inform themselves in respect of the IHL compliance of the beneficiaries of their military support.

C. IHRL Obligation of Non-Refoulement

IHRL has long recognized that increasing the risk that another actor will commit a serious breach of international law is in itself wrongful. In particular, states have a primary obligation not to contribute to the real risk that an individual will be subject to torture (or other serious violations of IHRL) abroad—which takes the form of the obligation of non-refoulement. The Convention against Torture sets out the obligation not to expel, return (refouler) or extradite a person to another state where there are substantial grounds for believing that she would be in danger of being subjected to torture expressly.[46] International and regional human rights treaties do not expressly impose a non-refoulement obligation, and the obligation is instead derived from the prohibition of torture and other cruel or inhumane treatment[47] or the positive obligation to provide (for instance) for due process.[48] For example, in its seminal decision in Soering, the European Court of Human Rights held it would

hardly be compatible with the underlying values of the Convention … were a Contracting State knowingly to surrender a fugitive to another State where there were substantial grounds for believing that he would be in danger of being subjected to torture … Extradition in such circumstances, while not explicitly referred to in the brief wording of Article 3, would plainly be contrary to the spirit and intendment of the Article …[49]

These readings of the prohibition of torture are of course entirely in keeping with a fiduciary theory of sovereignty and its principles of non-instrumentalization and non-domination. In particular, the non-refoulement obligation recognizes that a state’s obligations to ensure respect for fundamental human rights extend to conduct which creates a risk to those rights beyond its borders. Reading obligations not to facilitate into prohibitions, where these protect values of fundamental importance to the international legal order, is an important mechanism for ensuring a broader collective responsibility for the rule of law in international society.

Just as in the context of the regional human rights instruments, where the non-refoulement obligation is derived from a related prohibition bearing on states, international law increasingly recognizes that a prohibition on facilitation, where there is a real risk that IHL breaches will be committed (including through material support of actors whose record of non-compliance with IHL obligations is known or discoverable), can be derived from the prohibition bearing on states not to commit IHL breaches themselves. As discussed in Part III. A above, the implicit obligation not to facilitate breaches which are the subject of a negative obligation is in part derived from the general obligation to “ensure respect” of norms which protect human dignity and physical integrity, and the scope and nature of this obligation in the IHL context might helpfully be informed by the nature of the non-refoulement analysis.

Of interest for present purposes, the non-refoulement model in the torture context addresses the “balancing” of competing interests—in that states are prohibited from extraditing a person who faces a real risk of torture even if that person poses a significant security risk to the requested state.[50] This absolute prohibition (triggered by a real risk threshold) responds to the jus cogens nature of the norm prohibiting torture—no other interests are deemed sufficiently important to be balanced against the absolute nature of the prohibition. A state does not get to trade torture for its own security. From the perspective of intervening states which are supporting armed actors in foreign NIACs with a view to protecting security at home, this would certainly suggest that such support should be absolutely prohibited in cases where there is a real risk that the support will be used to commit jus cogens breaches of IHL. For instance, militarily or materially supporting armed actors in a NIAC abroad would be strictly prohibited if such support created a real risk that those actors would target civilians. The prohibition on directly targeting civilians is an elemental feature of the international legal framework governing the conduct of hostilities and is undoubtedly of jus cogens status. It cannot be balanced against any competing interests. Coupled with a state’s obligation to inform itself in respect of the conduct of the beneficiaries of its support (discussed in Part II. B above),[51] this is a powerful obligation in support of IHL jus cogens norms as a minimum standard of protection to which the territorial state’s population is entitled as a matter of international law. Of note, the absolute nature of the prohibition, and the impermissibility of balancing competing interests in gauging respect therefore, is entirely derived from positive law sources. While the fiduciary theory—as an interpretive theory—frames this obligation in a particular way, it does none of the prescriptive work. It may nevertheless be a useful lens through which to explore the prohibition.

D. The General Framework of State Responsibility

1.    The Prohibition on Complicity in Breaches of International Law

Article 16 of the ILC Articles on State Responsibility[52] prohibits states from being knowingly complicit in the internationally-wrongful act of another state,[53] and requires that both the wrongdoing state and assisting state should be bound by the same primary rule.[54] In its Bosnia Genocide Case decision, the ICJ extended responsibility for complicity set out in Article 16 to the conduct of NSAGs.[55] In extending complicity thusly, the principle that both the wrongdoing actor and the assisting (and complicit) state be bound by the same primary rule remains the same. In the IHL context, given the customary nature of the body of law, and its binding nature vis à vis NSAs, the obligations imposed on states and NSAGs alike are coextensive.

Complicity is a possible basis for responsibility in respect of state support to IHL-breaching armed actors, but it is not based on foreseeability (as is the CA1 obligation insofar as the Court was concerned in Nicaragua) or real risk (insofar as the non-refoulement model in the torture context can be understood to apply to such support). Instead, knowing assistance is required—Article 16 (or at least the Commentaries thereto) all but requires that the wrongdoing actor and the supporting state have a shared intention to commit the internationally wrongful act. This will very rarely be the case in the context under consideration—most particularly where the support to armed actors participating in a NIAC abroad is provided by the intervening state with a view to protecting its security interests at home. Article 16 might therefore be considered a minimum obligation of states, but does not respond to the balancing or competing interests, and tolerance for risk in respect thereof, at issue here. By contrast, the negative obligation under CA1,[56] as a specific primary obligation directed to these questions, can be interpreted as establishing broader responsibility for aiding breaches of international law, without the requirement of “intent”.[57]

2.    The Obligation to Co-Operate to Bring Serious Breaches of Jus Cogens Norms to an End

Just as Article 16 is amenable to application in the context of NSAs, in particular where the obligation in question is one binding on both state and NSAs alike, so too should Articles 40 and 41 of the ILC’s Articles on State Responsibility.[58] Read together, Articles 40 and 41 of the ILC Articles on State Responsibility require states to co-operate, to bring to an end through lawful means, any gross or systematic breaches of jus cogens norms. These obligations are thus a perfect fit for the fiduciary theory of sovereignty insofar as states are treated as fiduciaries of humanity in the broadest sense—they are charged individually and collectively with upholding norms which are foundational to the international legal order. The obligation pertains irrespective of whether the state is individually affected by the breach or not.

In its commentary, the ILC notes that Article 41 may be a progressive development, but that it is intended to strengthen existing mechanisms under international law.[59] Indeed, if Articles 40 and 41 have any role in suggesting the scope of an intervening state’s obligations in the face of IHL breaches by the beneficiaries of its support, it is to colour or further define the obligations derived from other sources. In particular, given its application to jus cogens norms, the obligation of non-co-operation is supportive of the approach suggested by analogy from non-refoulement obligations explored in Part II. C above.

E.  Obligations Deriving from the Arms Trade Treaty

The field of arms trade regulation has seen particular efforts to define more clearly state obligations to ensure respect for IHL. The UN Arms Trade Treaty of 2013 (ATT), which makes specific reference to the duty to respect and ensure respect for IHL, requires parties to assess the risk of their arms exports being misused before authorizing them.[60] Exports are prohibited by the ATT if the exporting state has knowledge that they would be used for war crimes, a restriction which given its high threshold in terms of the probability and gravity of the risk may add little to the existing rules of responsibility on aiding violations discussed in Part II. D above.[61] However exports are also prohibited if there is an “overriding risk” of their use to “commit or facilitate a serious violation” of IHL.[62]

This adds a considerably more specific basis for the requirement to test the lawfulness of providing arms by reference to the risk of IHL breaches, as compared to the very general text of CA1. It also explicitly requires exporting states to carry out a risk assessment in this regard.[63] The meaning of the word “overriding”, however, appears to leave in place the possibility of appeals to competing interests, including what might be regarded as other fiduciary duties. The ATT specifically acknowledges “the legitimate interests of States to acquire conventional arms to exercise their right to self-defence,” and the possibility that an arms transfer could contribute to peace and security, suggesting that these considerations could in some situations “override” some level of risk of IHL breaches.[64] While it may be rare that such considerations could justify transfers to NSAGs, the possibility of arguing that such a transfer will contribute to the “security” of certain populations (against threats from other NSAGs or the territorial government) is not necessarily excluded. Considerable discretion to invoke competing fiduciary duties as justifying arms transfers remains.

Comparison to the EU’s regional norm on arms exports is informative in this regard—under the EU Common Position, states may not export arms where there is a “clear risk” of serious violations of IHL.[65] That standard does not make allowance for balancing the risk against competing considerations in favour of transfers (although in practice the determination of when the threshold of “clear risk” is met has left considerable space for disagreement[66]). In theory, the discretion on the part of the fiduciary as to how it may balance its various duties is significantly curtailed by imposing this precautionary standard, regardless of the state’s assessment of the possible benefits of the transfer.

By contrast, although the ATT provides a clearer legal basis for requiring states to assess the consequences of their actions in this area, the prospects for actually obliging them to withhold support from participants in armed conflict on this basis are heavily dependent on how the “overriding risk” test is interpreted. The ATT is also lacking in enforcement mechanisms, has not been ratified by a number of the most significant arms exporting states,[67] and of course applies to only one form of support (the provision of arms). The effect is, on the one hand, a clear acknowledgment by ratifying states of their fiduciary responsibilities, but, on the other hand, an insistence on preserving a broad discretion as to how they seek to discharge those responsibilities and balance competing interests.

Conclusion

Past and ongoing efforts to develop state obligations to ensure respect for IHL in conflicts outside their territory can readily be understood in the terms suggested by Criddle and Fox-Decent’s fiduciary model. An increasing acceptance that states are legally obliged to have regard for the interests of individuals outside their territory can be identified in the development of these obligations by states, by the ICRC and ILC, and by the ICJ. The fiduciary model can provide theoretical support for those who seek the further development of those obligations.

But our consideration of the obligation to ensure respect for IHL in the context of foreign NIACs has identified some of the challenges associated with the fiduciary model. The difficulty lies less in establishing the existence of fiduciary responsibilities on states than in making the constraints they impose meaningful and effective. In some cases, the interests in question will point to a reasonably clear result as to what the fiduciary role of states requires or precludes. This is the case for encouraging other actors to violate IHL, or providing support to such actors with actual knowledge or intention that doing so will aid violations. Excluding such conduct from the ambit of states’ permitted fiduciary discretion follows logically from their agreement that there are no considerations which could justify them breaching the core norms of IHL themselves.

Conduct which falls short of active commission or knowing complicity in IHL breaches, however, raises more significant difficulties. The possibility that the competing fiduciary responsibilities owed by the state (for instance to protect its own population, as opposed to its role as fiduciary of humanity more broadly) may justify providing support to armed actors participating in a NIAC abroad, where there is some degree of risk that supported actors will breach IHL, renders the effective regulation of the discretion inherent in the fiduciary model particularly challenging. In that regard, the initial positive law basis furnished by CA1 has been progressively elaborated and supplemented to provide much more of the detail required for effective regulation, such as obligations to be informed, the permissibility or otherwise of balancing other interests against the risk of IHL breaches, and the differentiated treatment of risks to jus cogens compliance compared with those risks relating to other IHL breaches.

Until states agree more concrete and specific rules (as they have done in the ATT and even more so in the EU Common Position) as to how they may balance the competing interests of their various beneficiaries, significant challenges to effective regulation will remain. Progress to date has been slow, reflecting an international system in respect of which, in practice, it still largely falls to the fiduciaries to define the duties which they owe to their beneficiaries.

This is not necessarily a defect in the fiduciary model. It may be that leaving the most difficult decisions to states (whether case by case or by adopting general rules), while prohibiting the most clearly abusive behaviour, is exactly the model of constraint and discretion which a fiduciary role implies. The evolutions described above demonstrate states’ willingness to develop their obligations in this area in a manner which accounts for the complicated balances involved in protecting those at home while respecting the rights and “equal freedom” of those abroad—even if the results have been slow, limited, and uneven. They also demonstrate, particularly through the contributions of the ICJ and the ICRC, that the development of fiduciary obligations in this context is not entirely dependent on the active engagement of states. Although the results are far from perfect, the flexibility inherent in the doctrines of sources and interpretation of international law have allowed for significant development of positive law duties to ensure respect for IHL—duties which resonate in the fiduciary theory even if not derived therefrom.

                                   

[1]     For a discussion of the evolving relationship between international law and territoriality, see Daniel Bethlehem, “The End of Geography: The Changing Nature of the International System and the Challenge to International Law” (2014) 25:1 Eur J Intl L 9.

[2]     We cannot at present bring ourselves to characterize the twenty-first century as continuing such trends—not least given the rise of populist and nationalist narratives, which are skewing the balance between respect for human rights and security in favour of actual or apparent security within once-upon-a-time human rights oriented democracies.

[3]     Criddle and Fox-Decent argue that, while a state’s principal fiduciary obligation is to provide “a regime of secure and equal freedom for [its] own people,” sovereign authority defined in terms of “fiduciaries of humanity” requires states to exercise power over foreign nationals in keeping with their standing as subjects of the international legal order and rights bearers. See Evan J Criddle & Evan Fox-Decent, Fiduciaries of Humanity: How International Law Constitutes Authority (New York: Oxford University Press, 2016) at 171 [emphasis added]. Criddle and Fox-Decent’s account of sovereignty in this respect builds on Benvenisti’s insight that states must “take other-regarding considerations seriously into account” (ibid at 171–72).

[4]     Ibid at 178.

[5]     See Ryan Goodman, “The Power to Kill or Capture Enemy Combatants” (2013) 24:3 Eur J Intl L 819; Michael N Schmitt, “Wound, Capture, or Kill: A Reply to Ryan Goodman’s ‘The Power to Kill or Capture Enemy Combatants’” (2013) 24:3 Eur J Intl L 855; Ryan Goodman, “The Power to Kill or Capture Enemy Combatants: A Rejoinder to Michael N Schmitt” (2013) 24:3 Eur J Intl L 863.

[6]     Supra note 3 at 188–89

[7]     See Kimberley N Trapp, “Actor-Pluralism, the ‘Turn to Responsibility’ and the Jus Ad Bellum: ‘Unwilling or Unable’ in Context” (2015) 2:2 J Use of Force & Intl L 199.

[8]     The term “territorial state” is used throughout this article to denote the state in whose territory a NIAC is taking place, whether it is a conflict between the territorial state’s government and a non-State armed group (NSAG) or between NSAGs.

[9]     States providing support to armed actors in a foreign NIAC will be referred to throughout as “intervening states”.

[10]    See e.g. International Covenant on Civil and Political Rights, 19 December 1966, 999 UNTS 171, arts 2, 6, 7 (entered into force 23 March 1976) [ICCPR].

[11]    Charter of the United Nations, 26 June 1945, 1 UNTS XVI, arts 2(4), 51 (entered into force 24 October 1945) [UN Charter]. Art 51 is an exception to the Art 2(4) UN Charter prohibition on the use of force, and provides that:

Nothing in the present Charter shall impair the inherent right of individual or collective self-defence if an armed attack occurs against a Member of the United Nations, until the Security Council has taken measures necessary to maintain international peace and security. Measures taken by Members in the exercise of this right of self-defence shall be immediately reported to the Security Council and shall not in any way affect the authority and responsibility of the Security Council under the present Charter to take at any time such action as it deems necessary in order to maintain or restore international peace and security.

[12]    Whether and in what conditions such measures represent an effective means of achieving such protection is a debate beyond the scope of this paper.

[13]    Supra note 3 at 3, 18, 23.

[14]    Supra note 3 at 190, 206.

[15]    See Legality of the Threat or Use of Nuclear Weapons, Advisory Opinion, [1996] ICJ Rep 226 at para 25; UNOHCHR, Human Rights Committee, General Comment No 31: The Nature of the General Legal Obligation Imposed on States Parties to the Covenant, UN Doc CCPR/C/21/Rev.1/Add. 13, 26 May 2004 at para 11; African Commission on Human and Peoples’ Rights, “General Comment No 3 on the African Charter on Human and Peoples’ Rights: The Right to Life (Article 4)” (November 2015) at para 13; Hassan v United Kingdom [GC], No 29750/09 [2014] VI ECHR 1 at para 104; Coard v United States (1999) Inter-Am Comm HR, No 109/99 at para 42, Annual Report of the Inter-American Commission on Human Rights: 1999, OEA/Ser.L/V/II.106.

[16]    See the sources cited in supra note 15. Although some of these authorities relate to IACs, the lex specialis approach has also been employed in relation to NIACs—see e.g. Abella v Argentina (1997) Inter-Am Comm HR, No 55/97, OEA/Ser.L/V/II.106, Annual Report of the Inter-American Commission on Human Rights: 1997, OEA/Ser.L/V/II.98—and whether on this basis or through derogation, where permitted, it is likely that IHL will play the predominant role in setting standards for the conduct of hostilities in NIACs.

[17]    Supra note 3 at 205–06.

[18]    Ibid at 185ff.

[19]    This approach to HI dictated by the fiduciary theory is also reflected in positive law approaches to humanitarian intervention. See Kimberley N Trapp, “Unauthorized Military Interventions for the Public Good: A Response to Harold Koh” (2017) 111 AJIL Unbound 292. This does not mean, however, as Criddle & Fox-Decent argue, supra note 13 and accompanying text, that IHRL is the governing legal regime in respect of such interventions. IHL obviously applies in that HI will certainly meet the thresholds for an armed conflict, and it is widely recognized that IHL provides a lex specialis basis for defining the scope of the right to life in armed conflict: see supra notes 15–16. In addition, limitations on the territorial scope of IHRL treaty obligations pose particular obstacles when the intervention takes the form of support to an NSAG rather than, for example, direct military intervention or occupation. Support might however bring individuals within the supporting state’s “jurisdiction” and thus the scope of its IHRL treaty obligations in some more extreme cases. See e.g. Ilaşcu v Moldova & Russia [GC], No 48787/99, [2004] VII ECHR 179 at paras 377–94.

[20]    This was the case in Libya before SC Res 1973, UNSC, UN Doc S/RES/1973 (2011) 1 (authorising “all necessary measures” to protect civilians at para 4).

[21]    Supra note 11.

[22]    Under IHL, military attacks are characterized as disproportionate if they “may be expected to cause incidental loss of civilian life, injury to civilians, damage to civilian objects, or a combination thereof, which would be excessive in relation to the concrete and direct military advantage anticipated.” Protocol Additional to the Geneva Conventions of 12 August 1949, and relating to the Protection of Victims of International Armed Conflicts (Protocol I), 8 June 1977, 1125 UNTS 3, art 57(2)(a)(iii) (entered into force 7 December 1978) [API].

[23]    Such other entities being other states which are party to the relevant Convention and, under Common Art3 of the Geneva Conventions of 1949, NSAGs.

[24]    See European Convention for the Protection of Human Rights and Fundamental Freedoms, 4 November 1950, ETS 5, art 1 (entered into force 3 September 1953); ICCPR, supra note 10, art 2; Convention on the Prevention and Punishment of the Crime of Genocide, 9 December 1948, 78 UNTS 277, art 1 (entered into force 12 January 1951).

[25]    See Criddle & Fox-Decent, supra note 3 at 30.

[26]    For recent endorsement of such an expansive view of the obligation to ensure respect, see the revised commentaries by the International Committee of the Red Cross (ICRC) on the First and Second Geneva Conventions: International Committee of the Red Cross, Commentary on the First Geneva Convention: Convention (I) for the Amelioration of the Condition of the Wounded and Sick in Armed Forces in the Field, Knut Dörmann et al, eds (Cambridge, UK: Cambridge University Press, 2016) [ICRC, First Convention Commentary]; International Committee of the Red Cross, Commentary on the Second Geneva Convention: Convention (II) for the Amelioration of the Condition of Wounded, Sick and Shipwrecked Members of Armed Forces at Sea, Knut Dörmann et al, eds (Cambridge, UK: Cambridge University Press, 2017). The revised commentaries interpret the obligation as requiring states to “do everything reasonably in their power to ensure respect for the Conventions by others that are Party to a conflict,” bifurcated into a positive obligation requiring proactive measures such as using their influence to prevent breaches, and a negative obligation prohibiting encouraging, aiding or assisting breaches. See ICRC, First Convention Commentary at paras 153–73.

[27]    For the classic critiques of broad interpretations of the obligation to ensure respect’s legal effect see Frits Kalshoven, “The Undertaking to Respect and Ensure Respect in All Circumstances: From Tiny Seed to Ripening Fruit” (1999) 2 YB Intl Human L 3, reprinted in Michael N Schmitt & Wolff Heintschel von Heinegg, The Development and Principles of International Humanitarian Law (London, UK: Routledge, 2016) 415; Carlo Focarelli, “Common Article 1 of the 1949 Geneva Conventions: A Soap Bubble?” (2010) 21:1 Eur J Intl L 125.

[28]    See Ryan Goodman, “Two US Positions on the Duty to Ensure Respect for the Geneva Conventions” (26 September 2016), online: Just Security <https://www.justsecurity.
org/33166/u-s-positions-duty-ensure-respect-geneva-conventions> [Goodman, “Two US Positions”].

[29]    See e.g. Oona A Hathaway et al, “Ensuring Responsibility: Common Article 1 and State Responsibility for Non-State Actors” (2017) 95:3 Tex L Rev 539; Monica Hakimi, “Toward a Legal Theory on the Responsibility to Protect” (2014) 39:2 Yale J Intl L 247 at 254–55, 271–73; Goodman, “Two US Positions”, supra note 28.

[30]    Such an approach appears to respond well to Judge Shahabuddeen’s suggestion that “it may be useful to consider whether there is merit in the argument that, by deciding to use force through an entity [by the provision of support], a state places itself under an obligation of due diligence to ensure that such use does not degenerate into such breaches [of IHL], as it can.” Prosecutor v Dusko Tadic, IT-94-1-A, Appeal Judgement (15 July 1999) at para 20 (International Criminal Tribunal for the former Yugoslavia, Appeals Chamber), Shahabuddeen J, online: ICTY <www.icty.org>.

[31]    For example, as Hathaway and Manfredi note, at the same time as rejecting the broadest interpretation of the undertaking to ensure respect, the Legal Adviser to the US Department of State in 2016 added that “[a]s a matter of international law, we would look to the law of State responsibility and our partners’ compliance with the law of armed conflict in assessing the lawfulness of our assistance to, and joint operations with, those military partners”. Oona Hathaway & Zachary Manfredi, “The State Department Advisor Signals a Middle Road on Common Article 1” (12 April 2016), online: Just Security <https://www.justsecurity.org/30560/state-department-adviser-signals-middle-road-common-article-1>.

[32]    Although distinct, both of these duties can plausibly be deduced from the undertaking to ensure respect. See Robin Geiß, “The Obligation to Respect and to Ensure Respect for the Conventions” in Andrew Claphamet al, eds, The 1949 Geneva Conventions: A Commentary (Oxford: Oxford University Press, 2015) at 119.

[33]    The obvious comparison here is to the ICJ’s analysis in Application of the Convention on the Prevention and Punishment of the Crime of Genocide (Bosnia and Herzegovina v Serbia and Montenegro), [2007] ICJ Rep 43 at paras 430, 434, 438 [Bosnia Genocide Case]. The support provided by the then Federal Republic of Yugoslavia (FRY) to the Bosnian Serbs provided the FRY with significant capacity to influence their conduct and thus resulted in a more intense obligation on the FRY to prevent the recipients of that support from committing genocide.

[34]    See Goodman, “Two US Positions”, supra note 28.

[35]    See e.g. the negative obligation outlined in the ICRC’s revised commentaries, as described in supra note 26.

[36]    For a discussion of the obligation to be informed, see infra notes 41–45 and accompanying text.

[37]    This appears to be the position of the US government in relation to its ongoing support to Saudi Arabia in the Yemen conflict, for example. See Dan Lamothe, “Mattis Defends US Efforts to Prevent Civilian Casualties in Yemen”, The Washington Post (29 December 2017), online: <https://www.washingtonpost.com>.

[38]    See Military and Paramilitary Activities in and against Nicaragua (Nicaragua v United States of America), [1986] ICJ Rep 14 at paras 220, 255 [Nicaragua].

[39]    Ibid at para 256.

[40]    Ibid [emphasis added]. The Court also noted that US encouragement to breach IHL, based on awareness of at least allegations of IHL non-compliant behaviour by the contras, was likely to be effective (ibid)—but does not appear to consider “effectiveness” (presumably judged before rather than after the fact) as a necessary element of the breach. The case also provides an example of the type of “lesser evil” argument mentioned in Part II.A above; it had been argued that the purpose of the manual was to “moderate” the conduct of the contras (ibid).

[41]    See API, supra note 22, art 57(2)(a)(i).

[42]    See International Committee of the Red Cross, “Protocol Additional to the Geneva Conventions of 12 August 1949, and Relating to the Protection of Victims of International Armed Conflicts (Protocol I), 8 June 1977: Commentary of 1987”) at para 2195, online: <www.icrc.org/ihl.nsf/WebList?ReadForm&id=470&t=com>; Kimberley Trapp, “Great Resources Mean Great Responsibility: A Framework of Analysis for Assessing Compliance with API Obligations in the Information Age”, in Dan Saxon, ed, International Humanitarian Law and the Changing Technology of War (Leiden: Martinus Nijhoff, 2013) 153.

[43]    International Law Commission, “Report of the International Law Commission on the Work of its Fifty-Third Session, 23 April–1 June and 2 July–10 August 2001” (UN Doc A/56/10) in Yearbook of the International Law Commission 2001, Vol 2, Part 2 (New York: United Nations, 2007) 1 at 146ff (A/CN.4/SER.A/2001/Add.1 (Part 2)).

[44]    Art 3 of the ILC Draft Articles on the Prevention of Transboundary Harm requires “[t]he State of origin [to] take all appropriate measures to prevent significant transboundary harm or at any event to minimize the risk thereof” (ibid at 146). In the Commentaries, the ILC makes it clear that the “obligation extends to taking appropriate measures to identify activities which involve such a risk, and this obligation is of a continuing character” (ibid at 154). See also Art 7 (ibid at 146). This requirement, as specifically manifested in the obligation to carry out environmental impact assessments where activities to be undertaken in a state pose a potential transboundary risk, has subsequently been affirmed as a general international law obligation by the ICJ in Pulp Mills on the River Uruguay (Argentina v Uruguay), [2010] ICJ Rep 14 at para 204; Certain Activities Carried Out by Nicaragua in the Border Area (Costa Rica v Nicaragua) and Construction of a Road in Costa Rica along the San Juan River (Nicaragua v Costa Rica), [2015] ICJ Rep 665 at para 104. See also the Separate Opinion of Judge Donoghue, ibid at paras 2–9, 12–13 for further clarification regarding the customary nature of the obligation.

[45]    The Nicaragua-derived obligation hereunder discussion, the more general IHL obligation in respect of precautionary measures, and the ILC’s work on Transboundary harm, all draw on the decision in Corfu Channel Case (United Kingdom of Great Britain and Northern Ireland v Albania), [1949] ICJ Rep 4 at 22, in which the Court held that the laying of the minefields in the Channel “could not have been accomplished without the knowledge of the Albanian Government” [Corfu Channel].” In effect, Albania was held responsible for a failure to warn because it knew or ought to have known about the mine-laying. See also Kimberley N Trapp, State Responsibility for International Terrorism: Problems and Prospects (Oxford, Oxford University Press, 2011), at 66–70.

[46]    Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, 10 December 1984, 1465 UNTS 85, art 3 (entered into force 26 June 1987).

[47]    See Soering v United Kingdom (1989), 161 ECHR (Ser A) 1, 11 EHRR 439 [Soering].

[48]    See RB (Algeria) v Secretary of State for the Home Department, [2009] UKHL 10, [2010] AC 110 (Eng); Othman (Abu Qatada) v United Kingdom, No 8139/09, [2012] I ECHR 159.

[49]    Supra note 47 at para 88.

[50]    See e.g. Chahal v United Kingdom (1996), 23 EHRR 413.

[51]    An obligation comparable to that implicit is in a recent non-refoulement case. See Ilias and Ahmed v Hungary, No 47287/15 (referral to Grand Chamber) (14 March 2017) (holding that Hungary’s failure in various ways to obtain and consider available information as to the risk faced by the applicants should they be transferred amounted itself to a violation of the Art 3 non-refoulement obligation at paras 124–25).

[52]    See International Law Commission, supra note 43 at 27.

[53]    The question of characterization, in respect of a rule on complicity in international law, is somewhat difficult. While Art 16 complicity can be framed as a prohibition, this makes it a primary rule, when “complicity” might more accurately be characterized as a hybrid primary rule (the prohibition is implicit in responsibility flowing from complicity) and secondary rule (defining the consequences of assistance in another actor’s breach of international law). See generally Miles Jackson, Complicity in International Law (Oxford: Oxford University Press, 2015); Helmut Philipp Aust, Complicity and the Law of State Responsibility (Cambridge, UK: Cambridge University Press, 2011) at 188; James Crawford, “Second Report on State Responsibility”, (UN Doc A/CN.4/498/Add.1) in Yearbook of the International Law Commission 1999, Vol 2, Part 1, 3 at paras 164–65, 185–86 (UN Doc A/CN.4/SER.A/1999/Add.1 (Part 1)); Bernhard Graefrath, “‘Complicity in the Law of International Responsibility”‘ [1996] 2 RBDC 371.

[54]    See Crawford, supra note 53 at paras 181–84.

[55]    Bosnia Genocide Case, supra note 33 at para 420. The Court extends complicity under Art 16 to NSAGs through the primary norm prohibiting complicity in genocide set out in Art III(e) of the Genocide Convention, drawing on Art 16 by way of analogy.

[56]    See supra notes 24, 26, 35 and their accompanying text.

[57]    See Marco Sassòli, “State Responsibility for Violations of International Humanitarian Law” (2002) 84:6 Intl Rev Red Cross (2002) 401 at 412–13.

[58]    See International Law Commission, supra note 43 at 29.

[59]    Ibid (commentary to Art 41, para 3 at 115–16).

[60]    In this context arms exports are relatively broadly defined and would appear to cover non-commercial state transfers of arms to NSAGs, for example. See Stuart Casey-Maslen et al, The Arms Trade Treaty: A Commentary (Oxford: Oxford University Press, 2016), at 66, 246. This view is supported by all three of the states (Switzerland, New Zealand and Liechtenstein) which made interpretative declarations dealing with the issue (see infra note 61).

[61]    Arms Trade Treaty, 2 April 2013, 52 ILM 988 (entered into force 24 December 2014).

[62]    Ibid, art 7(3).

[63]    Ibid, art 7(1).

[64]    Ibid at Preamble, art 7(1)(a). See also Stuart Casey-Maslen et al, supra note 60 at 274–76; Professor Philippe Sands, Professor Andrew Clapham & Blinne Ní Ghrálaigh, “‘Legal Opinion: The Lawfulness of the Authorisation by the United Kingdom of Weapons and Related Items for Export to Saudi Arabia in the Context of Saudi Arabia’s Military Intervention in Yemen” (11 December 2015) at paras 5.47–49 online: Amnesty International <https://www.amnesty.org.uk/files/webfm/Documents/issues/final_legal_opinion_
saudi_arabia_18_december_2015_-_final.pdf>.

[65]    EC, Council Common Position 2008/944/CFSP of 8 December 2008 defining common rules governing control of exports of military technology and equipment, [2008] OJ L 335/99, art 2, criterion 2(c).

[66]    See e.g. Susanne Therese Hansen, “Taking Ambiguity Seriously: Explaining the Indeterminacy of the European Union Conventional Arms Export Control Regime” (2016) 22:1 European J Intl Relations 192; Laurence Lustgarten, “The European Union, the Member States and the Arms Trade: A Study in Law and Policy” (2013) 38 Eur L Rev 521.

[67]    Although it has been suggested by Tom Ruys that the prohibition on transfers which “would” be used for war crimes reflects a crystallization of the customary duty to ensure respect for IHL, and would thus bind all states. See Tom Ruys, “‘Of Arms, Funding and ‘Non-lethal assistance’: Issues Surrounding Third-State Intervention in the Syrian Civil War” (2014) 13:1 Chinese J Intl L 13 at 29.