ISIS Detentions in Syria and the Responsibility of Supporting States
While the world’s attention is focused on the hostilities in Gaza and Ukraine, the conflict against ISIS in Syria and Iraq continues. Although out of the headlines, coalition forces led by the United States continue to work with local partner forces to achieve the enduring defeat of ISIS and prevent the group’s resurgence as a credible fighting force.
Also far from the headlines is the fact that the coalition’s main partner in Syria, the Syrian Democratic Forces (SDF), continue to detain thousands of former ISIS fighters and their families. As a representative of the U.S. Department of State recently confirmed, “there are approximately 9,000 alleged ISIS fighters in detention in northeast Syria … [and] approximately 43,000 displaced persons in al-Hol and Roj camps … most of the individuals in these two camps are children.”
It is now over five years since the fall of Baghuz, when many of these individuals were first detained. For the majority, there is no prospect of release in sight. While some States are repatriating their nationals, others remain reluctant to do so. Meanwhile, the conditions in which these detainees and displaced persons live are appalling. The dire realities of life within the facilities are well documented, including by Amnesty International and the former UN special rapporteur on human rights and counterterrorism. Allegations include systemic torture, gender-based violence, inadequate water and sanitation, and arbitrary detention.
The actor that bears primary responsibility for these potential violations of international law is the detaining authority, namely the Autonomous Administration of North and East Syria (AANES), which includes the SDF. Yet, the AANES is a non-State actor that faces threats from all directions and is overwhelmed by the challenges associated with its detention activities (see the AANES’s response at Annex 1 to the Amnesty report). The AANES can only continue to hold such a huge number of detainees with the support of States.
This post considers States’ involvement in the ongoing detentions in northeast Syria and their potential responsibility for any associated violations of international law. It addresses the issue of attribution, then examines whether the law of armed conflict (LOAC) imposes any obligations directly on States regarding their support to non-State actors engaged in a non-international armed conflict.
State Involvement in SDF Detentions
Some members of the Global Coalition to defeat ISIS have provided direct support to the SDF’s detention operations. Their assistance includes constructing detention facilities, training, equipping, and paying prison guards (p. 17-18), and providing generators, lights, and security equipment to secure the perimeter of al-Hol camp (p. 25). The United States has also conducted joint operations with the SDF, resulting in the detention of additional ISIS fighters who are then placed into these same overcrowded and unsanitary detention facilities (p. 57).
The States involved are undoubtedly aware of the conditions in which the detainees and displaced persons are held, which their support facilitates and perpetuates. For instance, a U.S. government representative recently acknowledged that the facilities are “overcrowded, insecure, and face significant challenges involving the provision of medical care and sustenance.” It is also clear that these States are providing some humanitarian assistance to alleviate the conditions of those living within the detention facilities and camps (see the responses of the U.S. and UK governments at Annexes 2 and 3 of the Amnesty report). Meanwhile, the United States is leading diplomatic efforts to encourage and enable the repatriation of detained ISIS fighters and displaced foreign nationals from northeast Syria.
Yet, these measures are inadequate to ameliorate the position of the many thousands of individuals the AANES holds. While some repatriations have occurred, there is no end in sight to the detention of the majority of fighters and displaced persons who continue to face “deeply degrading” conditions of detention. The indefinite duration of the detentions, as well as the conditions in which the detainees are held, raise serious questions regarding the AANES’s compliance with international law. They additionally implicate the international responsibility of those States that assist and facilitate the SDF’s detention activities.
Potential Avenues of State Responsibility
There are two elements to an internationally wrongful act engaging a State’s responsibility: breach of an international legal obligation that is binding on the State; and attribution (Articles on State Responsibility (ASR) art. 2). Dealing first with breach, the principal bodies of international law that apply to the AANES’s detention operations are LOAC and international human rights law (IHRL). Potential IHRL violations are beyond the scope of this post, but have been addressed extensively elsewhere (see e.g. here).
The fighters and other individuals deprived of their liberty in northeast Syria were detained during the non-international armed conflict (NIAC) between ISIS on the one side and the SDF and the U.S.-led coalition on the other. Therefore, the law of NIAC applies to the detentions. This, however, includes very little detail regarding the detention-related obligations that bind the parties to the conflict. Common Article 3 to the Geneva Conventions states only that detainees must be treated “humanely, without any adverse distinction” and prohibits cruel treatment and torture. It does not define what it means by “humane treatment” or articulate whether a conflict party could violate this obligation by, for example, holding detainees in overcrowded facilities or failing to provide them with adequate water or medical care. The International Committee of the Red Cross (ICRC), meanwhile, in its most recent Commentary to common Article 3, emphasizes that the meaning of humane treatment is “context-specific and has to be considered in the concrete circumstances of each case” (para. 587).
As I highlighted in a previous post, there is considerable ambiguity regarding the precise LOAC obligations that bind the parties to a NIAC when they hold detainees. Nevertheless, given the dire conditions highlighted by Amnesty International and other organizations, as well as the allegations of systematic torture in some detention facilities, it seems likely that at least some LOAC violations have occurred. This post therefore proceeds on the basis that the AANES’s ongoing detentions of ISIS fighters and their families involve breaches of common Article 3.
Breach of an international legal obligation alone is insufficient to lead to State responsibility. That violation must also be attributable to a State. It is therefore necessary to examine whether any of the LOAC violations linked to the detentions are attributable to one or more of the supporting States. Here, it is pertinent to distinguish between violations committed by members of the AANES that may be attributable to a State, and LOAC violations committed directly by members of a State’s own military forces or other officials.
Regarding the former, reporting indicates that members of the AANES are responsible for most, if not all, of the LOAC violations complained of. It is the SDF, together with other security organizations within the AANES, that bear day-to-day responsibility for the security of the facilities and the treatment of detainees. The most likely basis on which LOAC violations, such as the inhumane treatment or torture of detainees, could be attributed to a supporting State is on the grounds that the SDF’s conduct in breach of common Article 3 was performed under the State’s instructions, direction or control (art. 8 ASR). However, as I have written previously, the threshold for proving attribution on this basis is extremely high (see Paramilitary Activities, para. 115). It is unlikely that any of the SDF’s conduct could be attributed to a State on this basis in the absence of evidence that the individuals concerned acted on the State’s instructions or under its control when committing the specific acts that violated LOAC. Proof of a broader level of State involvement in the SDF’s detention activities, such as by providing training, assisting in detention operations, or financing security measures at detention facilities, is insufficient to lead to attribution.
Attribution is, however, easy to prove when considering a State’s own violations of LOAC obligations that are binding upon it. In such cases, the relevant breach is attributable to the State on the basis that the conduct at issue was performed by an organ of State, for example, a member of the State’s own military forces (art. 4 ASR). Assuming that no State organs are directly responsible for mistreating detainees, the most relevant obligation in this context is the duty to respect and ensure respect for LOAC enshrined in common Article 1 to the Geneva Conventions (CA1). The argument here would be that the supporting States’ organs have failed in their duty to respect and ensure respect for LOAC by continuing to support and facilitate the SDF’s detentions in the knowledge that these violate common Article 3. In this situation, the supporting State would bear international responsibility not for the breach of common Article 3 but rather for its own failure to respect and ensure respect for LOAC.
Before addressing CA1, it is worthwhile considering in brief whether the supporting States could bear responsibility for their own contributions towards LOAC violations committed by the SDF. In other words, does the law of State responsibility hold States to account for the aid or assistance they provide to a non-State actor if that assistance facilitates the latter’s conduct in violation of international law? The brief answer to this question is no. When drafting the ASR, the International Law Commission reached the conclusion that States only bear responsibility for aid or assistance that facilitates another State’s international legal violations, not support that enables violations on the part of a non-State actor (art. 16 ASR). However, customary international law may be evolving to encompass the assistance that States provide to non-State actors (see Bosnian Genocide, paras. 419-420). Therefore, if States provide detention-related support to the SDF in the knowledge that this will facilitate the continued detention of ISIS detainees in violation of LOAC, those States could bear international responsibility for their contributions towards the relevant breaches.
Common Article 1
When addressing States’ responsibility for their assistance to the SDF’s detention operations, Amnesty International relies principally on the supporting States’ violations of CA1 and IHRL (p. 225-234). Given the difficulties outlined above in holding States directly to account for LOAC violations committed by the SDF, this focus is unsurprising. However, it raises questions regarding the scope of CA1 and the obligations this imposes on States that choose to provide support to a non-State actor engaged in conflict.
The duty to respect and ensure respect for LOAC enshrined in CA1 appears in identical form at the beginning of each of the four Geneva Conventions but its meaning is contested. The ICRC and many legal scholars consider that the provision imposes both negative and positive obligations on States to ensure that other States and non-State actors engaged in conflict respect LOAC, even if the former have no involvement whatsoever in the hostilities. However, this broad interpretation of CA1 is not widely accepted by States.
As Professors Michael Schmitt and Sean Watts have highlighted, the ICRC’s assertion that all States are bound by an obligation not only to ensure respect for LOAC by their own organs and populations but also to bring other States and non-State actors into compliance is not supported by the norm’s drafting history or by subsequent State practice. These scholars and some States additionally contest CA1’s application to NIACs (paras. 67-73). Yet, in the specific context of States’ support to non-State actors engaged in LOAC violations, there are strong arguments in favor of the norm’s application. Given the difficulties in holding supporting States to account, outlined above, CA1 could have a considerable impact when seeking to regulate States’ assistance to non-State actors in conflict situations.
Negative Obligations
While the application of CA1 to NIACs remains a subject of debate, the International Court of Justice (ICJ) ruled in Paramilitary Activities that the parallel provisions of customary international law impose obligations on States not to encourage non-State actors engaged in a NIAC to act in violation of common Article 3 (paras. 220, 255). This duty clearly applies to States that provide support to armed groups that are party to a NIAC. And as Professor Marko Milanovic observes, “it makes no sense to prohibit encouragement, but not to prohibit more material forms of aid and assistance. If, in other words, encouragement of a serious violation of [LOAC] is unlawful, then so is the provision of weapons that facilitate such violations.” The humanitarian logic for this extended interpretation of the judgment is compelling.
The ICRC characterizes the duty not to assist another actor’s LOAC violations as falling within the “ensure respect” element of CA1 (paras. 158-163). However, it arguably forms part of States’ obligation to respect LOAC through the conduct of their organs and agents. When State officials assist a non-State actor, they gain an enhanced ability to influence its conduct. The United States and certain other members of the anti-ISIS coalition have a huge degree of influence over the SDF’s detention-related activities because of the non-State actor’s dependence on those States for military and other support. The ICRC has recognized the significance of such support relationships, and the ICJ similarly acknowledged their importance in the Bosnian Genocide case (para. 434).
These relationships of influence are key to enhancing non-State actors’ compliance with norms of international law. If a State develops or sustains a relationship with a non-State actor engaged in conflict, thus furthering its own interests in the hostilities, the State’s obligation to respect and ensure respect for LOAC should clearly extend not only to its own armed forces, its population, and to other individuals or groups acting on its behalf, but also to any non-State actors it supports and leverages to achieve its own conflict-related objectives. If it is internationally wrongful for States to encourage LOAC violations on the part of a non-State actor, it must be similarly unlawful for them to provide material or other assistance to a non-State actor that facilitates such violations.
While this interpretation of CA1 may not have been contemplated by the parties to the Geneva Conventions at the time these were agreed, negative obligations of this nature follow naturally from the ICJ’s judgment in Paramilitary Activities and are necessary to ensure the norm’s effective operation. Moreover, in cases where such a support relationship exists, giving rise to an ability on the part of the State to influence the non-State actor’s behavior, it is arguable that CA1 or its customary equivalent imposes positive as well as negative obligations on the supporting State.
Positive Obligations
The scope of States’ positive obligations under CA1 is particularly contentious. The ICRC and some scholars take an expansive view of such obligations while certain States and other commentators deny that CA1 imposes any positive duties on States to ensure external actors’ compliance with LOAC. Drawing upon the ICJ’s judgment in Bosnian Genocide, however, there may be a sensible middle ground between these opposing positions. While the ICRC’s assertion that States are subject to a duty to ensure the LOAC compliance of all conflict parties is overly broad and somewhat unrealistic (see para. 153), the same is not true if the duty is restricted to situations where a relationship of influence exists.
To follow the ICJ’s judgment in Bosnian Genocide, the State’s “capacity to influence effectively” is key to the imposition of such positive obligations (para. 430). The ICJ highlighted a range of factors that might be considered when determining whether a State can influence another actor, including “the geographical distance of the State concerned from the scene of the events, and … the strength of the political links, as well as links of all other kinds, between the authorities of that State and the main actors in the events.”
Taking such factors into account, the United States and other members of the anti-ISIS coalition clearly have the capacity to influence the SDF’s detention-related conduct. When a State assists a non-State actor, this typically generates a degree of dependence by the non-State actor on the State that the State can leverage to exert its influence over the non-State actor’s conduct. In the case of the SDF, that level of dependence is particularly high. Continued U.S. support is “essential for SDF survival” in view of the threats the group faces not only from ISIS but also from Turkey and the Syrian regime.
When a State chooses to assist a non-State actor engaged in conflict, therefore, in addition to its negative duty not to knowingly facilitate LOAC violations, the State should be subject to a positive duty to ensure that the non-State actor does not use that assistance in a manner that violates international law. The extent of that duty might differ, however, according to the strength of the State’s relationship with the non-State actor concerned. As the ICJ acknowledged, a State’s capacity to influence “varies greatly from one State to another” (para. 430). Thus, when a State has a particularly strong relationship with a non-State actor, the steps it is expected to take might be greater than if its capacity to influence the non-State actor’s conduct is relatively weak. In the context of the SDF, this would mean that the obligations imposed on the United States, as the lead partner in the coalition with the strongest relationship with the SDF, might be greater than the duties imposed on the other States involved.
Further factors of relevance to the State’s positive obligations are the severity of the LOAC violation and the means available to the State. These elements, as well as the State’s ability to influence the non-State actor, impact the steps that it might be feasible for the State to take to ensure the non-State actor’s respect for LOAC. Regarding the SDF, the United States and other coalition partners have a very high level of influence over the group’s detention activities. Moreover, the LOAC violations associated with the ongoing detentions are potentially serious, of long duration, and impact the lives of thousands of detained individuals, including many children. As such, CA1 or its customary equivalent should impose positive obligations on those States to take wide-ranging steps to ensure the SDF’s LOAC compliance.
Concluding Thoughts
The United States and other members of the anti-ISIS coalition have provided significant detention-related assistance to the SDF over the past five or more years. This has enabled the group to hold thousands of former ISIS fighters and other individuals, including children, in long-term detention. Despite some repatriations, the prospects of release for the majority of detainees remain slim. And as the detentions continue, for year after year, the conditions in which those detainees are held continue to deteriorate.
It is easy to blame the detention authority, the AANES, for this state of affairs. As the United Kingdom said in its response to the Amnesty report, “ultimately, responsibility for detention and camp facilities and the wellbeing, detention, transfer or prosecution of detainees is a matter for authorities under whose jurisdiction individuals are detained” (see Annex 3). But the AANES is a non-State actor that faces huge pressures and a lack of resources. The AANES can only continue to hold such large numbers of detainees with the assistance of States.
States have provided significant support to the SDF’s detention activities, and their assistance has undoubtedly had some impact in improving the conditions for those detained. But it does not go far enough. States, to date, seem to have effectively turned a blind eye to the potential LOAC violations their assistance facilitates in the knowledge that the associated legal risks are low. Attribution of the SDF’s violations to a supporting State is highly unlikely; the evidence is insufficient to prove that members of the SDF were acting under a State’s instructions, direction or control at the time they acted in breach of common Article 3. Other avenues of State responsibility are also far from straightforward. It remains uncertain whether the provisions of the ASR that prohibit States from facilitating international law violations by other States also apply to States’ assistance to non-State actors. And States, scholars, and the ICRC take widely differing positions regarding the impact of CA1.
When thinking about CA1 in the context of States’ support to non-State actors, there is a sensible interpretation of the norm that falls in the middle ground between these opposing positions. The duty to respect and ensure respect for LOAC is particularly pertinent when States build relationships of influence with non-State actors engaged in conflict, whereby they provide those groups with assistance and develop the capacity to guide their behavior. In such situations, there are compelling humanitarian reasons to interpret the duty not to encourage LOAC violations so as to also prohibit States from facilitating breaches of LOAC on the part of the non-State actors they support. This interpretation is logical. Moreover, it is necessary to ensure the effective operation of the norm.
The imposition of positive duties by CA1 or its customary equivalent is more contentious. However, in the context of support relationships, it makes sense for such positive duties to apply. When States choose to achieve their national security goals through the intermediary of a non-State actor, they should be subject to an obligation to take feasible steps to ensure that the recipient of their support does not use the assistance provided in a manner that violates LOAC. Such an obligation should apply to Iran when it supplies arms to Hamas and the Houthis, just as it applies to the States that support the SDF.
States’ adoption of this proposed interpretation of CA1 could ultimately lead to its acceptance as the agreed construction of the norm. This would enhance States’ ability to call out other States like Iran that make extensive use of proxies, which frequently act with complete disregard for LOAC. Of course, States’ acceptance of these duties would also impact their own assistance to non-State actors like the SDF. But if the States that support the SDF took such obligations more seriously, there could potentially be an end in sight to the indefinite detention of ISIS fighters and their families in Syria.
***
Jenny Maddocks is an Assistant Professor in the Department of Law at the United States Military Academy, West Point and the Managing Editor for Articles of War.
Photo credit: Y. Boechat (VOA)
