What international legal rules determine when the U.S. government can resort to force against a non-state armed group such as the Khorasan Group, if that organization plans an imminent strike on the United States? And when can a state lawfully use force to stop a homegrown armed group from launching an attack? In this context, does international humanitarian law (IHL) limit state action before the first attack against a non-state actor (NSA)? As recent posts on Just Security reveal, this question troubles us very much in this era of non-international armed conflicts (NIACs). It is especially vexing since it points to the specter of a legal void. According to the orthodox view, NIACs exist (and IHL applies) only after a high level of intensity of violence is reached. What is more, at least in the context of transnational attacks, the application of international human rights law (IHRL) is many times uncertain. Moreover, in such cases, the law on inter-state resort to force is not entirely helpful since its adaptation to the relations between an attacking state and an NSA – essentially a group of individuals – is imperfect.  What, then, if a state decides to conduct a surprise or preemptive attack against a NSA? What would be the pre-attack norms or set of rules that govern the relations between the parties?

In a thoughtful post, Adil Haque suggests that IHL would constrain parties as long as an organized armed group has the “capacity to sustain military operations.” IHL thus essentially governs pre-attack relations as long as such capacity exists.  Michael J. Adams and Ryan Goodman, too, challenge the pre-attack intensity requirement, which is commonly traced to the ICTY’s Tadic ruling. As they point out, the key question is not whether an NIAC exists, but rather, what is the applicable law before violence reaches the intensity level that Tadic envisions. They imply that IHL might “bind” the first attack, but leave us to ponder the relationship between IHL and other bodies of law that might regulate this initial resort to force. To Haque, and some extent also to Adams and Goodman, IHL’s “early” applicability might be desirable since IHL has constraining aspects, and as such could be preferable to the alternatives.

The root of the problem we face as international lawyers is found in traditional international law’s failure to address adequately resort to force in all non-international settings.

Historically, this stems from international law’s reluctance to deal with violence within state borders. Owing to anachronistic perceptions of sovereignty, the international law on jus ad bellum is still perceived to be silent on intra-state resort to force. This failure does not stop at the state level: in certain circumstances, it also affects transnational resorts to force against NSAs. Since we don’t know when states are legally justified to resort to force internally, we find difficulties in judging the actions of third parties that assist them. Without saying something about internal jus ad bellum, we are very limited, for example, in judging the legality of the Russian intervention in favor of the Assad regime in Syria. We can only assess the actions of the regime and Russia in bello, but we don’t have a satisfying legal vocabulary to judge the intervention itself and the resort to force in the first place.

Lawyers don’t like vacuums, and therefore, as I argue in a recent article, much of the discussion “whether a NIAC exists” (or, in its more precise version “whether IHL applies”) is really about the legality of the use of force itself. In this discourse, IHL actually serves as a proxy for jus ad bellum questions. For instance, much of the post 9-11 debate on US targeting policies asks whether there is an armed conflict between the US and certain NSAs. Critics say no, proponents say yes. Both implicitly acknowledge that the question is crucial since IHL is (generally perceived as) more permissive (at least concerning targeting) than the legal status quo. In short, if there is an armed conflict, people can be targeted in a manner that is otherwise impermissible. The lawfulness of an attack hinges on the question whether an armed conflict exists.

Of course, this argument is circular. Since the existence of an armed conflict is a question of fact, saying that an attack is permissible if one exists is like saying “an attack is permissible if other attacks take place.”  A state can provoke an armed conflict, and then justify its actions by referring to the existence of one. In theoretical terms, this allows moving from the general regime of human rights-based law enforcement to the IHL-based state of exception, without any normative regulation.

This predicament is indeed rooted in international law’s approach toward internal resorts to force—within the four corners of a single state—but quickly transposes onto transnational questions, through the problem of consent-based interventions. Earlier this week, Rebecca Crootof and Sarah Weiner rightly noted that “states involved in … consent-based interventions have an obligation to independently determine if human rights law or international humanitarian law is the governing legal regime—and act accordingly.” This is hard to dispute. However, the upshot is that presumably, if there is an internal armed conflict, and IHL applies, a third-state can be invited by the government to intervene and join the hostilities. Absent any additional conception of internal jus ad bellum, this consent can serve as a standalone justification for third party involvement, even if the inviting government instigated the internal hostilities itself and for unjustifiable reasons.

As detailed in my article, I suggest adding another layer to the analysis. Whether “an armed conflict exists,” however we define it, will always be relevant to the applicability of the key aspects of IHL. Yet, we also have to ask whether the first resort to hostilities was lawful, in order to avoid the circularity. This requires identifying the pre-armed conflict norm that regulates the switch between the normal and the exceptional, between law enforcement and hostilities.

Contrary to the orthodox approach, I argue that such a legal standard exists: IHRL’s right to life can be viewed as a prohibition on the first resort to hostilities between a state and NSAs. It is widely accepted that under IHRL lethal force can be justified only in self-defense or defense of others. Internal resort to hostilities would thus be justified only in face of a threat to self or others of such magnitude that would necessitate military force. The legality or illegality of third party assistance would piggyback on the lawfulness of the initial resort to force of the consenting party.

In black-letter terms, the “arbitrary” deprivation of life, prohibited in Article 6(1) of the ICCPR, does not have to be limited to IHL violations taking place during active hostilities (as implied by the ICJ in the Nuclear Weapons advisory opinion), but can also refer to the decision to resort to force itself. The same analysis would apply to Article 2 of the European Convention of Human Rights: it can be understood as prohibiting resort to hostilities unless in “defense of any person from unlawful violence” or if undertaken lawfully (meaning, in self-defense) against an insurrection or organized use of armed force. If these articles govern a single decision to resort to lethal force by a police officer or a police unit, there is no reason to exclude from their purview much broader decisions on force, e.g. by military means. Of course, necessity and proportionality considerations are embedded in this analysis: if a state decides to resort to hostilities in absence of a threat to life or limb that is comparable in its scale and effects to the threats that emanate from armed hostilities, then the ensuing killings could be considered arbitrary deprivations of life.

In sum, this framework could apply directly to internal confrontations between states and NSAs and, by extension, to foreign interventions with governmental consent. This is because the consenting government, obviously, cannot circumvent its own IHRL obligations – in this case, its internal jus ad bellum obligations – simply by utilizing a third party. The intervening party will also be in the wrong in such a case, by assisting in the commission of an internationally wrongful act. The application of this framework to non-consensual transnational attacks against NSAs, however, is less straightforward. Namely, it would require a broad understanding of extra-territorial application of IHRL, as regulating the pre-attack phase even without the attacker’s presence on the ground (an interpretation that I support, but will not develop here). Moreover, this proposal does not replace the international law on jus ad bellum that would also apply in such cases, at least in the relations between the attacker and the territorial state.

Undoubtedly, there are many practical and theoretical questions that this proposal raises. I address some of them in my article. Yet, there is much to be gained by realizing that the application of IHL cannot in itself legalize armed violence in NIACs. Without an external reference point—one that mirrors inter-state jus ad bellum—we quickly find ourselves immersed in IHL technicalities while losing sight of the bigger question: whether resort to force was at all legally justified.