Last month, the US State Department’s Legal Adviser Brian Egan highlighted one of the most significant legal issues on the horizon: US and coalition forces’ handling of large numbers of ISIL detainees as ISIL-controlled areas are liberated in Syria, Iraq, and elsewhere (See Kate Brannen’s Just Security post, “Beyond Gitmo: What is the US Going to Do About the Coming Wave of ISIL Detainees?”). A question is what happens when a US-aligned armed group engages in gross violations of international law in its treatment of detainees? Indeed, that was one the key questions that Charlie Savage posed to Egan at the event, and the Legal Adviser’s response was not specific enough to provide any real insight into where the government draws the legal lines in such cases.
This type of problem and the legal questions it raises are not limited to the United States, to the counter-ISIL fight, or to detainee operations. These issues will trouble States as they rely on local partner forces across different conflicts.
We provide a legal risk analysis: evaluating the prospect that a State will run afoul of its international law obligations when it assists a non-State armed group’s engaging in internationally wrongful conduct. For legal experts the specific question is: does the rule in Article 16 of the Articles of State Responsibility apply to State aid and assistance to organized armed groups (OAGs)? (We explain for others what we mean by that query below.)
Our bottom line up front: A State runs a high legal risk of violating international law in aiding or assisting an OAG in the group’s commission of internationally wrongful acts.
I. Bracketing other legal rules that may render a State and its officials liable
In the main part of our analysis, we do not address three other legal risks that might also be relevant to a State’s support for an OAG:
1. Aiding and abetting liability under international criminal law
Note 1. Aiding and abetting war crimes clearly applies to assistance given to a non-State actor.
Note 2. Aiding and abetting liability applies if members of the OAG engage in a war crime or other international crime. It does not necessarily apply if the OAG’s actions are unlawful but fall below the line of international crime.
Note 3. Aiding and abetting liability does not require the assisting officials share the same purpose as the OAG members—knowledge that their support will facilitate the OAG’s offences suffices.
Note 4. Aiding and abetting liability may require that the State’s support to the OAG has a “substantial effect” on the commission of the crime (watch this space: Beth Van Schaack and Alex Whiting will write soon on this element of aiding and abetting liability for the purpose of evaluating State assistance). This may be one of the sharpest distinctions between aiding and abetting liability and complicity under the law of State Responsibility: in the latter framework, an assisting State may violate international law for a modest level of support to the internationally wrongful acts. More on that below (see III(3)).
2. Primary rules (and due diligence) in human rights law and law of armed conflict (LOAC)
Subject matter specific legal regimes contain what are called “primary rules,” which impose direct obligations on States not to support deprivations of human rights and humanitarian protections carried out by OAGs. These rules apply even if the OAG does not itself have a direct obligation under international law. Courts often conclude that these primary rules include “due diligence” obligations that require State officials to ensure they do not support such wrongful acts, and, under some circumstances, that States act to protect individuals from deprivations by third parties.
Note 1: Under the Geneva Conventions, the ICRC has interpreted this primary rule very expansively to require a State to help ensure respect for the Conventions even by belligerents and armed conflicts with no connection to the State. The US and other governments appear to have rejected that broad interpretation (see here and n. 32 here). However, due diligence obligations under the Geneva Conventions might be understood to apply in a narrower set of cases, for example, when a State affirmatively encourages or supports an OAG in the group’s commission of clear LOAC violations. For US purposes, especially if one takes into account prior official statements favoring relatively broad primary obligations under (Article 1) of the Geneva Conventions, the government might be held to the primary obligations in this relatively narrow set of circumstances.
Note 2: Under human rights law, two factors may appear to reduce significantly or eliminate the legal risks: whether human rights law applies extraterritorially and, even if it does, whether human rights obligations apply to operations in which the delivery of lethal force lacks “authority and control” over an individual. But that appearance of significantly reducing or eliminating the risk is somewhat deceptive. The first issue—extraterritorial reach of human rights law—is more a question about jurisdictional provisions of treaties not about customary international human rights law (see Ryan’s Just Security post, “The United States’ Long (and Proud) Tradition in Support of the Extraterritorial Application of International Human Rights Law.”). Second, some operational situations will not raise the vexing issue of “authority of control” associated, for example, with airstrikes, but will instead involve much closer and clearer legal issues where control is exercised in the custody of detainees.
Note 3: In another respect, human rights law imposes more far-reaching obligations on States and increases their legal exposure compared to the general standards set forth in the principles of State responsibility. As Shaheed Fatima explains in a Just Security post, a claim against a State alleging that it violated its human rights obligations by assisting another State “would not require [a] court to concretely determine the lawfulness of the other State’s conduct,” while “by contrast, [the rule for State responsibility in assisting another State] would necessarily require a court to decide whether the other State was responsible for an internationally wrongful act.” For example, a State will automatically engage its international responsibility under human rights law by transferring a detainee to a place where there is only a “risk” of torture or inhumane treatment.
3. Attribution under the Articles of State Responsibility
One of the ways in which a State may be liable for the illegal conduct of an OAG is if the group is under the complete dependence of the State (i.e. operates as a de facto organ of that State) – a very high threshold (Nicaragua case (Nicaragua v US), paras 109-110; Bosnia Genocide (Bosnia and Herzegovina v Serbia and Montenegro) case, para 392). Alternatively, where the OAG is left with some degree of autonomy from the State, the illegal conduct of such OAG can be attributed to the State if it was performed on the instructions, under the direction or control of the State (Article 8 of the Articles of State Responsibility). The question for our purpose is whether that exhausts the principles of State responsibility for how a State to run afoul of international law in its aiding and assisting an OAG’s conduct. (For legal experts: we also bracket Article 11 of the Articles of State Responsibility). By way of analogy, do the principles of State responsibility include a form of “contributory liability” for State assistance to an OAG?
II. State complicity in aiding and assisting the conduct of OAGs
Aside from the rules of attribution, and regardless of whether the State exercised control over the wrongful conduct committed by an OAG, a State will presumably violate international law in aiding or assisting an OAG in the group’s commission of an internationally wrongful act.
It is commonplace to find that the actions of an OAG are not directly attributable to a State, but that a State may nonetheless be responsible for aid and assistance to that OAG in the group’s commission of an illegal act. In short, that’s what happened in the ICJ judgment in the Nicaragua case. The Court held that the relationship between the US and the contra rebel forces did not satisfy the effective control test for direct attribution (para 115). Nevertheless, the Court found that the US was under “an obligation not to encourage persons or groups” to act in violation of common Article 3 of the Geneva Conventions. The Court added, “such an obligation does not derive only from the Conventions themselves, but from the general principles of humanitarian law to which the Conventions merely give specific expression” (para 220). The Court found that the US support to the contras breached the State’s obligation under customary international law not to intervene in the affairs of another State (para 292(3)) as well as its obligation not to encourage IHL violations (para 292(9)). Likewise, in Armed Activities on the Territory of the Congo (DRC v Uganda), the Court held that “even if the evidence does not suggest that [the Congo Liberation Movement (MLC)] conduct is attributable to Uganda, the training and military support given by Uganda to ALC, the military wing of MLC, violates certain obligations of international law” (para 161), including the principles of non-intervention and the prohibition of the use of force. We do not mean to confuse the primary rule in humanitarian law and on the use of force with the general principles of State responsibility. However, it is notably a relatively small step from the ICJ’s logic to general principles of state responsibility for assistance to groups like the contras or MLC.
1. Applying Article 16 to States aiding or assisting OAGs (theory)
The text of Article 16 of the Articles of State Responsibility applies only to State aid and assistance to another State. Does the absence of State aid and assistance to non-State actors in that provision mean that such relationships are not covered by principles of State responsibility under customary international law? The short answer is no.
First, the ILC articles are not, and were not intended to be, exhaustive. Hence, the subsequent effort by the ILC to draft a separate set of articles on the responsibility of international organizations. That raises another point for understanding the limited ambition of Article 16. Note that Article 16 does not include State aid and assistance to an international organization (another form of a non-State actor). By contrast, the analogous provision of the more recent ILC draft articles on the responsibility of international organizations does cover an international organization’s aid and assistance to a State or to another international organization. It would be a mistake to conclude from Article 16’s silence on aid to an international organization that States can lawfully provide aid to an international organization for the purpose of assisting the organization in the commission of international law violations. The same is true for State aid and assistance to other non-State actors—OAGs included.
Second, during the drafting of the ILC Articles of State Responsibility, the degree to which non-State actors had direct international obligations was far more unsettled than it is today. Traditionally only subjects of international law (i.e., States) could generally breach international law. This is important because the rule in Article 16 requires both the assisting State and the recipient not to commit an internationally wrongful act, arguably even if the source of the underlying obligation is different.
During their respective tenures as the principal drafters of the Articles on State Responsibility, Special Rapporteurs Robert Ago and James Crawford did not conclusively discard the possibility of State “complicity” in the acts of a non-State actor serving as an additional basis of responsibility. Ago recognized that “[c]ases in which a State incurs international responsibility for the act of a subject of international law other than a State (e.g. an international organization or an insurrectional movement), although intellectually conceivable, are not covered, because there are no known cases in which this has actually happened and such cases are unlikely to occur in the future.” In earlier proceedings, Ago also stated:
“The study of international practice could, however, show that the acts of private individuals are never taken into account in determining the international responsibility of the State unless they are accompanied by certain actions or omissions of organs of the State. This should not automatically lead us to exclude the possibility of attributing the action of an individual to the State. Indeed, it could be so attributed, but only in cases where it is specifically characterized by a measure of participation or complicity on the part of State organs. … As has already been shown, when the State endorses the act of an individual it is the State itself which acts, both through the individual and through the organs which are ‘accomplices’; the idea of such complicity would thus be quite conceivable. … The action of an individual would be the basis of the internationally wrongful conduct of the State, and the State would violate an international obligation through the action of an individual in which certain organs were merely accomplices.”
Crawford, for his part, may have been hesitant to advance a theory that OAGs had direct international obligations, a condition for the application of Article 16 in its final form. Indeed, even as late as 2005, Crawford, in discussing Article 16, expressed a narrow conception of non-State actors ever functioning as subjects of international law capable of breaching direct legal obligations. He said:
“Individuals and corporations may of course act independently, and when they do, outside the very limited field of international criminal responsibility (which as we have seen is presently limited to individuals, not corporations) they can only be held accountable in national courts and to a large extent only under national law.”
Despite Crawford’s views, it has become clear over time that OAGs have direct obligations in the areas that concern us here—for example, under international human rights law, especially when the OAG exercises territorial control, and under the law of armed conflict.
[We note that jurists would not need to inquire only into the modern developments on that subject. Check out Chapter 2 of Vlad’s recent book which examines the concept of “implied complicity,” applied by tribunals in the late 19th and early 20th century, whereby a State could engage a responsibility for the act of a non-State actor.]
2. Applying Article 16 to States aiding or assisting OAGs (practice)
The rapid developments of international law, in particular taking place after the Cold War, show that OAGs are bound by an increasing number of international obligations under customary international law, including the law of armed conflict and human rights law. These developments not only answer the question whether an OAG can commit a breach of international law, but also prompt an inquiry into whether a State may be held legally responsible for aiding or assisting an OAG that commits violations of international law.
For the purposes of legal risk analysis, one should foreground the strong signal sent by the International Court of Justice, which other international bodies, national courts, and governments are likely to follow. In the Bosnia Genocide case, the ICJ applied Article 16 by analogy to a case of assistance from Serbia to the Bosnian Serb’s armed forces (paras 419-21). Although the ICJ was deciding whether Serbia was complicit under the Genocide Convention, the Court referred to Article 16 for guidance—on the basis that complicity in genocide is “similar to a category found among the customary rules constituting the law of State responsibility.” The Court acknowledged that the strict terms of Article 16 do not apply to aid to non-State actors but nevertheless appeared to consider the basic rule in Article 16 as an independent basis for complicity, that is, separate and apart from issues relating to effective control for purposes of attribution.
Notably, in its Order for Provisional Measures, the Court had also directed Serbia to “ensure that any military, paramilitary or irregular armed units which may be directed or supported by it, as well as any organizations and persons which may be subject to its control, direction or influence” refrain from illicit acts (emphasis added).
The possibility of triggering a State’s responsibility for providing assistance to a non-State actor has also found support in the jurisprudence of regional human rights bodies. The European Court of Human Rights has, for example, found “military, economic, financial and political support” from a State to a non-State actor to be sufficient for the purposes of engaging the State’s responsibility (Ilascu v Moldova and Russia, Judgment, ECtHR Reports 2004-VII 179, paras 377-94). The Inter-American Court has even adopted the view that “the acts of private individuals … can be attributed to the State and, therefore, entail its responsibility in accordance with international Law, for which it is sufficient to prove that there has been support or tolerance by the public authorities in the breach of the right” (see e.g. Case of the ‘Mapiripán Massacre’ v Colombia, Merits, Reparations and Costs [2005] IACtHR Ser C No 134, paras 107-11 and 120) (emphasis added). In the context of rendition cases, it is also notable that regional human rights courts when determining State complicity often refer to the Articles on State Responsibility rather than expressly stating that they are devising special rules of attribution of conduct which would only apply in respect of human rights law violations (see e.g. El-Masri v The Former Yugoslav Republic of Macedonia, Judgment [2012] ECtHR No 39630/09 (2013) EHRR 25; Al Nashiri v Poland, Judgment [2014] ECtHR App No 28761/11 (2015) EHRR 16; Husayn (Abu Zubaydah) v Poland, Judgment [2014] ECtHR App No 7511/13 (2015) 60 EHRR 16). A caveat is important here – It is not always clear on which legal basis the State is being held responsible (i.e. a breach of a specific primary rule, on the basis of an attribution standard lower than that of effective control, or on the basis of Article 16) but the fact that courts have referred to Article 16 among the relevant applicable law is telling.
[As an aside: notably, in a Just Security post, Monica Hakimi explained, “international courts have also applied [the obligation to protect] extraterritorially—in particular, to states that substantially support rebel groups in other states.” (See also her law review article, in which she writes, “a state’s extraterritorial duties to protect might be triggered if it substantially supports the third party that perpetrates atrocities.”) Those particular obligations may be primary rules specific to the human rights regime in which they operate. Nonetheless, they once again show how little it would take for international courts to consider a similar principle more generally exists under the customary international law of State responsibility.]
States have also occasionally referred to the same principle in Article 16 as the basis for responsibility of a State in connection with aid or assistance rendered to OAGs. Austria, for instance, invoked Article 16 and stated: “Should supplied arms be used by armed opposition groups in Syria in the commission of internationally wrongful acts, the States who had supplied these arms and had knowledge of these acts would incur responsibility for their aid or assistance in the commission of such acts.” Derek Jinks documented the development of this conception of State responsibility for the conduct of OAGs following September 11th. In an article entitled, “State Responsibility for the Acts of Private Armed Groups,” he wrote:
“[T]he US arguably sought to impute al Qaeda’s conduct to Afghanistan simply because the Taliban had harbored and supported the group, irrespective of whether the state exercised “effective control” (or “overall control”) over the group. Although this line of argument is not new for the United States, the claim enjoyed much broader international support in the wake of September 11. As discussed more fully in Part I, the UN Security Council, NATO, and the OAS expressly or tacitly endorsed the US position. Moreover, many distinguished commentators have expressed some measure of support for this type of claim.”
In sum, the exposure of States to legal risk for aiding an OAG in the commission of an internationally wrongful act is high.
3. Level of support
As mentioned in Part I, under aiding and abetting criminal liability, the support provided may need to have a “substantial effect” on the commission of the crime by members of an OAG. The principle of State responsibility in Article 16 functions very differently. The principle operates, in part, like a primary rule insofar as it holds that the provision of aid or assistance toward the commission of an internationally wrongful act is wrongful in and of itself. Regardless of whether one conceives of the rule that particular way, it is clear that a modest contribution to the wrongful conduct could suffice. It is worth quoting the ILC Commentary to Article 16 in full for this proposition:
“In accordance with article 16, the assisting State is responsible for its own act in deliberately assisting another State to breach an international obligation by which they are both bound. It is not responsible, as such, for the act of the assisted State. In some cases this may be a distinction without a difference: where the assistance is a necessary element in the wrongful act in absence of which it could not have occurred, the injury suffered can be concurrently attributed to the assisting and the acting State. In other cases, however, the difference may be very material: the assistance may have been only an incidental factor in the commission of the primary act, and may have contributed only to a minor degree, if at all, to the injury suffered. By assisting another State to commit an internationally wrongful act, a State should not necessarily be held to indemnify the victim for all the consequences of the act, but only for those which, in accordance with the principles stated in Part Two of the articles, flow from its own conduct.” (emphasis added)
Many commentators take an earlier passage in the Commentary to suggest that aid or assistance must have contributed significantly to an internationally wrongful act. A closer reading (and in light of the above quote) may suggest that such a threshold of support is a sufficient but not a necessary condition. Specifically, the Commentary states: “There is no requirement that the aid or assistance should have been essential to the performance of the internationally wrongful act; it is sufficient if it contributed significantly to that act” (ILC Commentary to Article 16, para 5). As Ryan and Miles Jackson wrote in a Just Security post:
“Some scholars argue that the Commentary requires that the assistance contribute significantly to the principal wrong. The Commentary, however, states only that a significant contribution is ‘sufficient,’ not that it is necessary. When the Commentary turns to the degree to which an assisting State may be held responsible for its acts, the ILC contemplates situations in which ‘the assistance may have been only an incidental factor in the commission of the primary act, and may have contributed only to a minor degree, if at all, to the injury suffered.’ We assess that the latter threshold is likely too low as a matter of customary international law.”
For what it’s worth, the commentary to the earlier draft of this provision noted that “aid or assistance must have the effect of making it materially easier [the commission of an internationally wrongful act]” (ILC Commentary to former draft Article 27, para 17) (emphasis added). Detailed analysis of comments by States in the process of elaboration of the Articles of State Responsibility suggest that States were focused on limiting the scope of this form of responsibility to instances where assistance is clearly linked to the wrongful act. (For more on the nexus requirement, see the recent Chatham House report and Just Security’s mini forum with Chatham House on the topic; see also Chapter 5(II)(A) of Vlad’s book).
Crucially then, any form of aid or assistance provided to an internationally wrongful act, with the knowledge of the circumstances of that act, is capable of raising the specter of responsibility of the aiding or assisting State. It will be difficult for the State to deny its responsibility on the sole ground that its knowing aid or assistance clearly linked to the commission of an internationally wrongful act did not contribute significantly to it.
4. Knowledge/ purpose element
Finally, one of “the most vexing issues” in analyzing the principle of State responsibility in Article 16 is whether the rule requires an assisting State to have an intent to facilitate the actions of the recipient State—and what might be meant by “intent” (see Oona Hathaway and Ryan Goodman’s discussion of Brian Finucane’s article; see also Chapter 5(III) of Vlad’s book ). Importantly, this purpose element may not be the same in the case of State assistance to a non-State actor. According to the recent Chatham House report, a key justification for including a high threshold for intent under Article 16 is to avoid “a chilling effect on cooperation between states” (para 63). That reasoning works because State-to-State cooperation is generally considered a public good in international relations. Notably, that justification doesn’t apply so easily to State-OAG cooperation. In part the reason for a potentially stricter mental element for Article 16 responsibility comes from an assumption that in its dealings a State is entitled to rely on the good faith commitment of its counterpart (e.g. Lac Lanoux (France v Spain) (1957) 12 RIAA 281, 313; Tacna-Arica Question (Chile/Peru) (1925) 2 RIAA 921, 930; see also Chapter 5(III) of Vlad’s book). This assumption of good faith does not and should not apply where the “counterpart” is an OAG and not a State. In practice, this means that the State has to run many more checks before delivering its assistance to OAG and ensuring that the aid will not contribute to international law violations. In sum, there is less reason to think that the purpose element exists at all or to the same degree when it comes to State aid and assistance to OAGs. At least a clear-eyed, legal risk analysis would need to be well informed of this possible distinction.
* The authors are grateful to the following experts for their advice on early drafts: Helmut Aust, Andrew Clapham, Shaheed Fatima QC, Monica Hakimi, and Miles Jackson.