Taking One for the Team: Legal Consequences of Misconduct by Partners

As Nordic states and their allies embark on ever more challenging military operations abroad, the complexity of the applicable rules – and legal consequences they may have – is growing rapidly. In addition to the specific terms of the mandate and the underlying principles of the UN Charter, there are two regimes that govern the conduct of military and other security forces across the war-peace spectrum: the law of armed conflict (LOAC) and international human rights law (IHRL). The exact application of these regimes and their interaction remain matters of debate. Yet, the most vexing question is: what happens when it is not one state’s own forces, but those of a partner that run afoul the LOAC and IHRL standards? This article maps and disentangles the relevant international rules and regimes under which responsibility may arise for states due to abuse committed by their partners in military operations abroad. It considers the general rules of state responsibility – in particular, the rule on state complicity – and the relevant substantive rules under the LOAC and IHRL as well as the Arms Trade Treaty and the UN Human Rights Due Diligence Policy. By and large, states have to abstain from any form of assistance that could contribute to the commission of IHRL and LOAC violations by their partners. Under certain circumstances, they even have to actively and continuously assess the risk of such misconduct occurring and take mitigating measures. Nevertheless, these obligations do not necessarily translate into an effective system of accountability.


Introduction
In June 2018, the High Court of Eastern Denmark rendered its judgement in the so-called Operation Green Desert case and ordered the Danish government to compensate a group of Iraqi nationals that had sued the Ministry of Defense over ill-treatment during their detention in Iraq in 2004. 1 Around 350 Danish forces had assisted Iraqi and British forces in a counter-insurgency operation. 30 Iraqi insurgents had been captured in the course of that operation and transferred to a detention facility operated by the Iraqi police, where most of them were subjected to torture or other forms of inhuman treatment. The court accepted the government's view that Danish forces had not been involved in the apprehension and handling of the detainees. Nevertheless, it found the Ministry of Defence liable as a matter of national tort law. According to the court, Danish forces "should have known" that those captured in the operation were at risk of being subjected to inhuman treatment in Iraqi custody. It is to be seen whether the judgement will be upheld on appeal before the Danish Supreme Court.
What is surprising is the lack of references to international law throughout the entire judgement (Henriksen 2018), while public international law is certainly not silent on the matter. What exactly are the rules and regimes governing the provision of military assistance between states in the face of possible abuse by one or the other? And how do they apply to the different mission scenarios? Answering those questions -as this article sets out to do 2 -is not only pertinent to the facts of the Operation Green Desert case, i.e. mistreatment

MILITARY STUDIES
and torture of persons in the custody of partners, both of which are clearly outlawed by the law of armed conflict (LOAC) and international human rights law (IHRL). In fact, the analysis is also relevant for unlawful use of force (including long-range targeting) in violation of the applicable rules under both regimes. As the case may be, such violations can arise in shared detention facilities, following transfers of detainees to partners or on-base training or during mentoring or joint operations in the field. There are also situations where assisting states are physically removed from the actual scene; for instance, when they provide air-toair refuelling and air support or engage in intelligence sharing, arms delivery or other forms of assistance.

General Rules of State Responsibility
The so-called Articles of Responsibility of States for Internationally Wrongful Acts (ARSIWA) of the International Law Commission (ILC) form a starting point for the inquiry. 3 States are primarily responsible for their own acts and omissions. It is, therefore, important to consider whether possible abuse by a state can also be attributed to other states participating in the same operation. In short, the rules of attribution, enshrined in ARSIWA, link the actions of natural persons, such as soldiers, to legal entities, such as states. Generally, attribution seems uncontroversial for armed forces in relation to their home states. In international military operations, however, it may provide formidable challenges. For a case of attribution to another state, the soldiers perpetrating the alleged violations must usually act under the direction and control of that state. 4 Yet, in view of the loose command-and-control arrangements common in most operations, this standard will only rarely be met. 5 The picture is, however, more complex when the partner is a non-state armed group (e.g. Kurdish forces in Syria), for which attribution is often based on more relaxed control standards. 6 Most importantly, international responsibility does also arise when the partner's misconduct is not directly attributable, but where a state "aids or assists" another state in the commission of a wrongful act -also commonly known as "complicity". Article 16 of the ARSIWA (ILC 2001) states: 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: (a) that State does so with knowledge of the circumstances of the internationally wrongful act; and (b) the act would be internationally wrongful if committed by that State.
The aid or assistance in question does not have to be "essential" for the commission of the wrongful act, but at least a "significant" contribution is required. This excludes minor and peripheral contributions without a clear nexus (ILC 2001: Art. 16, para. 5). More controversially, however, the commentary introduces an additional requirement: "the aid or assistance must be given with a view to facilitating the commission of that act, and must actually do so" (ILC 2001: Art. 16, para. 3, emphasis added). This intent requirement goes far beyond the knowledge explicitly mentioned and thus excludes many situations in which states provide significant assistance towards the commission of wrongful acts. For Aust (2011: 237-38) and Crawford (2013: 408), the intent requirement should be seen in the light of reactions by states and their necessary support for the relatively new rule on complicity. It can also be explained with the chilling effect that a simple requirement based on knowledge alone would have on cooperation among states (Moynihan 2016: 22). Such considerations do not apply in relation to the assistance of non-state actors like armed groups, which is why complicity in unlawful acts by such groups is not subject to an intent requirement (Goodman and Lanovoy 2016). 7 3 ILC 2001. For the equivalent rules of responsibility for international organisations: ILC 2011. 4 ILC 2001: Art. 6, para. 3 ('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', emphasis added); Boutin 2017. 5 The assessment would change if a broader control standard were to be applied, as suggested by Dannenbaum 2010, pp. 157-192 (' control most likely to be effective in preventing the wrong in question') and Mothers of Srebrenica 2014: para. 4.46. 6 Henriksen 2017, pp. 134-136 (discussing the different control standards ' effective control' or ' overall control' in the Nicaragua and Genocide cases before the International Court of Justice and the Tadic case before the International Criminal Tribunal for the former Yugoslavia). Note that the European Court of Human Rights (ECtHR) even went as far as considering ' decisive influence' and crucial 'military, economic, financial and political support' sufficient for attributing the acts of the Transnistrian secessionist authorities in Moldova to Russia (Ilaşcu 2004: paras. 392-394). This finding may, however, be specific to Russia's special role in the post-Soviet break-away regions and not necessarily applicable to the relationship between states involved in the anti-ISIS coalition and Kurdish forces in Syria. 7 Note that the text of Art. 16 is limited to state-to-state cooperation. Yet, customary law provides for a broader complicity rule, covering also assistance to non-state armed groups and international organisations.
This requirement appears to be met at least when the assisting state knows with certainty that its assistance contributes to the misconduct in question. By contrast, constructive knowledge (i.e. that the assisting state should have known) and failure to make enquiries will arguably not be enough for responsibility to arise (Moynihan 2016: 22;Goodman and Lanovoy 2016).
To be considered complicit in the violations of another state, the assisting state must have the same (or corresponding) obligations. In general, states are bound by very similar rules regarding the treatment of persons as well as targeting and the use of force under the LOAC and IHRL. 8 However, certain differences remain and may affect the way in which different states operate effectively together in the course of an international military operation. Due to this double-obligation requirement, no case of complicity could possibly arise. Yet, such cases of legal interoperability may create particular challenges under the substantive regimes themselves, which will be considered further below. 9 What is more, Article 41 of the ARSIWA requires states to cooperate in order to bring serious breaches of peremptory norms to an end and not to "render aid or assistance in maintaining that situation" (ILC 2001: Art. 41, emphasis added). It does not include a knowledge (or intent) requirement, assuming that it is inconceivable that such violations go unnoticed (ILC 2006: para. 11). Moreover, Article 41 only applies after the commission of a serious breach. It could, however, be argued that it nevertheless imposes more far-reaching restrictions than the complicity rule enshrined in Article 16 whenever serious breaches of peremptory norms are at stake (Aust 2011: 422;Moynihan 2016: 23). The ban of torture and the most fundamental LOAC targeting rules are strong candidates for the inclusion in the list of peremptory norms (jus cogens). 10 However, Article 41 refers to serious breaches -defined as "gross and systematic" violations -of peremptory norms. This clearly implies an even narrower scope for possible mission scenarios (ILC 2006: Art. 40, para. 7). Finally, the legal status of Article 41 ARSIWA remains unclear. In fact, doubts have been raised whether it constitutes a norm of customary law and whether it will ever attain that status (Aust 2011: 343;Moynihan 2016: 22). As a consequence, Article 41 does not seem to provide a robust basis under which states can be said to incur international responsibility for possible LOAC and IHRL violations committed by their partners in international military operations.

Law of Armed Conflict
The LOAC contains some substantive provisions that limit the way in which states can cooperate militarily with one another. For instance, prisoners of war may only be transferred to states that are willing and able to ensure the required treatment. 11 State practice seems to have extended this underlying non-refoulement principle to the transfer of other detainees, including in non-international armed conflicts (Droege 2008; International Committee of the Red Cross [ICRC] 2016: paras. 708-716). The real challenge in applying this standard in multinational military operations is often the identification of the detaining power prior to the transfer; in other words, who detained and held the detainee before the transfer took place? Moreover, the LOAC requires occupying powers "to protect the inhabitants of the occupied territory against acts of violence, and not to tolerate such violence by any third party" (Democratic Republic of the Congo v. Uganda 2005: para. 178), which also includes partners in the field. To this date, the brief occupation of Iraq in [2003][2004] has been the only recent case of a multinational occupation. Whether there were other states (besides the US and the UK) that qualified as occupying powers in Iraq at the time and in which exact areas they had to restore and ensure public order and safety are questions that have not been answered conclusively.
These few rules only apply in relation to a party to an armed conflict or an occupying power, which require either a certain level of fighting or a hostile military presence in foreign lands (ICRC 2006: paras. 217-451). This means that certain forms of remote assistance are likely beyond their reach. Yet, a state may also become a party to an ongoing conflict due to the support it provides to one side, without itself being involved in actual fighting. This is particularly relevant in a non-international armed conflict where a state assists the host state (e.g. Mali or Somalia) or a coalition of states (e.g. anti-ISIS alliance) against a non-state armed group. Such support may range from air-to-air refuelling of combat aircraft, providing arms and ammunition as well as actionable intelligence and target information to training and mentoring in the field and handling captured enemy forces (Ferraro 2013: 583-587). 12 Yet, the obligations under the LOAC go much further: Common Article 1 of the Geneva Conventions (1949) requires states to "ensure the respect" by other states and non-state entities involved in an armed conflict (ICRC 2016: para. 12). This includes a negative obligation to refrain from encouraging or aiding and assisting in LOAC violations. Beyond any form of assistance, states have also a positive obligation to take feasible measures in order to influence the parties to the conflict towards full compliance with the LOAC. This is an obligation of means; in other words, states are free to choose the measures necessary to prevent violations in case of a "foreseeable risk" and to stop ongoing violations. Moreover, the fact that a state cooperates closely with a party to the conflict "places it in a unique position to influence the behaviour of those forces" towards full compliance with the LOAC (ICRC 2016: paras. 158-73). While there continues to be some controversy as to its exact scope, 13 it appears reasonable to side with the ICRC that the duty to ensure respect by others does not only apply to states that are a party to the conflict in question but instead to all states. 14

International Human Rights Law
For Nordic countries, it is the European Convention on Human Rights (ECHR) that constitutes the most relevant human rights treaty, not least because of its highly effective European Court of Human Rights. The ECHR provides for negative obligations to refrain from any arbitrary interference with a right; and positive obligations to protect such rights from abuse by third parties. In doing so, the ECHR corresponds largely to its universal counterpart: the International Covenant on Civil and Political Rights (ICCPR). While this ensures a comprehensive framework of protection at home, the picture is more complex when states act abroad, such as in international military operations. The extraterritorial application of human rights treaties remains particularly challenging due to so-called "jurisdiction clauses" that the ECHR and most others contain. 15 Moreover, customary IHRL does not provide an alternative, because it is highly unlikely that states have accepted more far-reaching positive obligations than under treaty law (Milanović 2011: 3;Kretzmer 2005: 185).
By and large, a state is only expected to ensure the rights of persons abroad that find themselves in the physical custody of or in an area controlled by that state (Al-Skeini 2011: paras. 133-140). Where this is the case, the state has to take all necessary measures to protect them from possible abuse by others, including their partners. For instance, Danish soldiers acting abroad may not transfer detainees to the host state authorities or to other contingents if there is a real risk of torture or inhuman treatment (i.e. non-refoulement principle). 16 Yet, premised on the exercise of exclusive control, the ECHR cannot be applied so easily to international military operations, where states often share control over an area or facility with other states, 17 or where it is difficult to determine the exact time and circumstances under which a person is held in custody by a state. Note that an additional "public powers" model (Al-Skeini 2011: para. 135) applies in situations where the high control standards (over territory or persons) are not met due to the involvement of different actors (e.g. host state or other partners). Yet, this model covers only acts directly attributable to the state in question, not omissions; in other words, it does not apply to a state's failures to prevent abuses by others, including its partners. Hence, no jurisdiction and thus no positive obligations could possibly arise in relation to misconduct by others under the public powers model. This means that many situations of possible abuse by partners fall outside the scope of the ECHR and other human rights treaties, because the states in question exercise no control and provide only very remote forms of assistance, e.g. intelligence sharing, arms delivery or logistical assistance. This is why the rule on state complicity is of enormous importance in order to fill some of the responsibility gaps. In other words, even if a state's IHRL obligations are not directly applicable, the state may nevertheless incur international responsibility for willingly contributing significant assistance to the commission of IHRL violations by a partner state.

New Instruments
Adopted in 2013, the Arms Trade Treaty (ATT) provides an additional layer of responsibility. 18 The ATT entered into force in 2014; so far, it has 105 state parties, including the Nordic states and most OECD countries. 19 Among others, it prohibits the transfer of conventional arms, ammunition or parts thereof, if they are likely to be used in the commission of serious LOAC and IHRL violations. 20 To that end, it requires states to assess the potential risk that such arms and items could be used to commit or facilitate serious violations of the LOAC and IHRL. Where such risks exist, states must consider specific mitigating measures, including confidence-building measures or joint programmes. If the risk persists, the transfer may not take place. 21 In sum, the ATT covers an important part of the assistance that states may provide to partners in international military operations and beyond. As it extends only to conventional arms and items, the ATT does not restrict the provision of facilities (e.g. landing strips or ports), intelligence and targeting information or other forms of support that may potentially contribute to the commission of LOAC and IHRL violations by partners.
In March 2013, the UN Secretary-General issued the Human Rights Due Diligence Policy (HRDDP). 22 While not legally binding stricto sensu (Aust 2015), it has been specifically mentioned as a requirement in some UN Security Council resolutions providing the mandate for UN peacekeeping missions. 23 It is of particular relevance to national troops (e.g. from Denmark) deployed to UN peace operations that cooperate closely with host state forces, such as in Mali. Along the same lines as the ATT, the HRDDP requires an assessment of the potential risks in providing support. If there is -despite mitigating measures -a risk of grave violations of the LOAC and IHRL, the support must be withheld. 24

Risk Assessment and Mitigation Measures
To some extent, states have far-reaching obligations to protect and not to assist in relation to misconduct by their partners, deriving from the LOAC and IHRL itself. This framework is significantly complemented by the ATT and the HRDDP as well as the rule on state complicity, provided that the required intent requirement is met. As a general standard deriving from those different legal regimes, states have to abstain from any form of assistance that could contribute to the commission of IHRL and LOAC violations by their partners. Under certain circumstances, 25 they even have to actively and continuously assess the risk of such misconduct occurring based on the partner's previous compliance record, including the failure to hold perpetrators accountable. Thus, Denmark and other Nordic states are advised to make such risk assessments a standard procedure for any form of military cooperation that they may engage in. In particular, risk assessments may help identify particularly problematic partners and detect structural deficiencies on the part of the partner, which can be addressed by tailored mitigation measures, e.g. capacity-building and training (Finucan 2016: 427;Moynihan 2016: 42-43;ICRC 2016: para. 181). Ideally, instruction of the applicable legal rules under IHRL and the LOAC should be at the core of such measures, including tactical training with weapons. It is noteworthy that training and other activities towards better compliance with international law do not fall 18 For a comprehensive commentary of the ATT, see Clapham et al. 2016. 19 The US is only a signatory and notified on 18 July 2019 that it no longer intends to become a party to the ATT. For further information: https://treaties.un.org/Pages/ViewDetails.aspx?src=TREATY&mtdsg_no=XXVI-8&chapter=26&clang=_en. 20 Art. 6 (3) ATT, emphasis added. 21 Art. 7 ATT. 22 UNSG 2013. See also : Aust 2015. 23 For instance, UNSC 2017: para. 26 ('Requests that MINUSMA take fully into account the need to protect civilians and mitigate risk to civilians, including, in particular, women, children and displaced persons and civilian objects in the performance of its mandate as defined in paragraphs 16 and 17 above, where undertaken jointly with the Malian Defence and Security Forces, in strict compliance with the Human Rights Due Diligence Policy', emphasis added). 24 Para. 16 ('must not engage in the provision of support') and para. 17. 25 Such obligations derive from the duty to ensure respect by others under the LOAC as well as from the ATT and HRDDP. Due to the limited extraterritorial scope of IHRL, its risk assessment duties only apply to situations involving the prior exercise of jurisdiction abroad.
under "support" for the purpose of the above-mentioned HRDDP (UNSG 2013: para. 9). In other words, such activities can continue even when other forms of support must be withheld due to the risk of possible violations. When assisting their partners, states should condition their support to compliant behaviour (Finucan 2016: 427-429;Moynihan 2016: 41-42;ICRC 2016: para. 181). For instance, they should only provide weapons, logistical assistance, intelligence on targets or air support, if the partner agrees to comply with the required targeting and use-of-force standards. Likewise, assisting states may decline to participate in joint cordonand-search operations or to transfer detainees to their partner until reliable assurances have been given, regarding the treatment of the detainees. The European Court of Human Rights has drawn up a detailed list for assessing the reliability of diplomatic assurances, 26 which includes, among others: the involvement of the central government, and the effectiveness of the ban on torture and inhumane treatment at the domestic level as well as the diplomatic assurance's disclosure, specificity and legal power to bind the receiving state.
Conditionality will only work with effective monitoring (Finucan 2016: 429;Moynihan 2016: 43-44). This can take the form of visits by the ICRC or other humanitarian organisations, e.g. where the post-transfer treatment of detainees is in doubt. Monitoring may also involve personnel of the assisting state itself, either on an ad-hoc basis or as embedded units for a longer period of time.
To ensure better compliance, states can also take a more leading role on the ground, e.g. in the form of joint planning of operations (ICRC 2016: para. 181). Yet, where such enhanced presence involves a significant degree of command authority over the partner, there is an increased risk of attribution, whereby the assisting state would be held directly responsible for any violations by its partner (Finucan 2016: 430). Hence, states have to consider carefully whether it is better to get more involved or to withdraw from the operation and thus cease all forms of assistance. In such cases, states should focus on securing evidence for possible future investigations, either by themselves or other actors.
While developing and maintaining complex risk assessment procedures is the right way forward, they also come at a high cost. Indeed, especially smaller nations like those in Scandinavia have to make a disproportionate amount of resources available. Therefore, it makes sense to start pooling them with like-minded allies and developing joint risk assessment procedures.

Accountability and Litigation
As has been shown in the foregoing sections, international law does not provide one single solution for responsibility arising from misconduct by partners. Rather, there is a plethora of rules and regimes that possible situations may fall under, with many of them reinforcing each other and helping to fill possible gaps. Nevertheless, this comprehensive set of rules and obligations does not necessarily translate into an effective system of accountability. Indeed, the degree to which states are exposed to legal actions for breaches of their obligations is very much left to their own discretion. The success of possible proceedings before national courts depends largely on the way the domestic legal system treats liability claims against the state, especially when they are based on non-incorporated treaty provisions or customary law. As the above-mentioned Operation Green Desert (2018) case clearly shows, courts may choose to examine cases almost exclusively from a national law perspective. Two recent cases from Germany 27 and the United Kingdom 28 concerning assistance for military operations in Yemen follow a similar pattern. Most importantly, state complicity in a breach of international law does not in itself give rise to an individual claim on the part of the victim. The same dilemma prevails when states fail to meet their due diligence obligations under Common Article 1, the ATT or the HRDDP. 29 The nature of the rules at stake would require state-to-state litigation before international courts, e.g. before the International Court of Justice. Such options are often very limited; even where such avenues are open, the interested states may refrain from using them for merely political reasons.
The most effective accountability mechanisms for possible victims are provided by human rights treaties such as the ECHR, which come with their own courts or quasi-judicial bodies and allow for individual complaints. Where applicable, IHRL includes far-reaching duties to protect without the need to clarify whether the perpetrator of the abuse violated international law or was even bound by it. Yet, many situations are not covered; where the state in question is geographically removed from the abuse itself, its assistance or mere inaction will not trigger its positive IHRL obligations. This is why Jackson (2016) has called for a broadening of the concept of jurisdiction under the ECHR to also include cases of state complicity not otherwise covered. It is, however, unlikely that the court will heed this call in the near future. In fact, one would expect that the court should first address the responsibility of states for their own IHRL violations abroad, by extending at least the negative obligations to all forms of extraterritorial conduct.
Nevertheless, state officials should be particularly careful, as they may incur individual criminal liability for aiding and abetting in international crimes (Goodman 2016a;Jackson 2015: 69-85;Aksenova 2016: 103-111). To be precise, there is no need for sharing the purpose under international criminal law: knowledge that the support will facilitate the commission of war crimes, crimes against humanity or even genocide will usually suffice. 30 Most importantly, immunity generally enjoyed by state officials does not necessarily shield them from prosecution for these international core crimes. 31

Conclusion
International law lacks a single framework governing the responsibility of states in cases of misconduct by partners in military operations. In view of that, this article sought to outline and disentangle the relevant international rules and regimes. IHRL provides for far-reaching obligations to protect individuals from abuse by others: it does not only prohibit any form of assistance to the perpetrator but also the failure to prevent and stop such abuse. As shown above, however, these positive obligations do not readily apply to many scenarios in international military operations abroad, due to a lack of jurisdiction. Thus, the rule on state complicity is of enormous importance, as it helps to fill some of the responsibility gaps. In other words, even if a state's IHRL obligations are not directly applicable, that state may nevertheless incur responsibility for willingly contributing significant assistance to the commission of IHRL violations by a partner in the field.
By contrast, the LOAC provides for a tightly knit framework that captures the full spectrum of possible scenarios. Indeed, the duty to ensure respect by others includes a negative obligation to refrain from encouraging or aiding and assisting in LOAC violations; and a positive obligation to take feasible measures in order to induce partners to abide by the LOAC. State complicity is, therefore, of only little relevance in armed conflict situations. Both the ATT and the HRDDP are largely inspired by this duty to ensure respect by others but extend this framework -including its risk assessment requirement -to the field of IHRL, even where the latter is not applicable in relation to the transferring or assisting state. Nevertheless, this comprehensive set of rules and obligations does not necessarily translate into an effective system of accountability.
Beyond questions of accountability, there is a need for further research in view of increasingly remote forms of aid and assistance that states may provide to their military partners, e.g. through drone technology, exchange of intelligence and cyber operations. Yet, new technologies may also play an important role in operationalising risk assessments and mitigation actions. Indeed, they may be used for documenting possible abuse by partners and for recording relevant communication with them. What is more, enhanced surveillance technology is particularly useful for monitoring purposes. Ultimately, where interventions with the partner fail to stop the abuse or to hold the perpetrators accountable, digital documentation may be passed on as evidence for possible investigations and prosecutions in international fora. 29 Moreover, practice by national courts examining such cases shows a great degree of deference to the executive branch. 30 Under customary law, ' aiding and abetting' is broader than Art. 25 (3) of the Rome Statute (1998), using 'for the purpose of facilitating the commission of such a crime'. But ' aiding or abetting' requires a 'substantial effect' on the commission of the crime, which is a slightly higher threshold than under the rule of state complicity: Van Schaack and Whiting 2017. 31 For instance, heads of state, heads of governments and foreign ministers enjoy a so-called 'personal immunity', shielding them from possible prosecution abroad, regardless of the severity of the crimes. Such immunities do not, however, prevent prosecution by the ICC where the state in question has accepted its jurisdiction (per ratification or declaration) or where the situation has been referred to the ICC by the UN Security Council.