Over-the-Horizon Operations – Part III: How May Force Be Used?

by | Dec 23, 2021

Over-the-Horizon Strikes III

Part I of this series examined when the United States may lawfully resort to force in the form of over-the-horizon (OTH) operations, a question that looms large given President Biden’s July announcement that the United States is developing the capability in light of its withdrawal from Afghanistan. Part II discussed where those operations may be conducted by examining the circumstances in which such operations may cross into the territory of Afghanistan.

In this final installment in the series, I consider how OTH operations must be conducted by surveying the law that applies to them, whether conducted from outside Afghanistan, by platforms in Afghan national airspace, or by forces inserted into the country. The determinative questions are whether the operations would occur in the context of an armed conflict, and if so, whether that conflict would be international or non-international in character. That topic provoked extensive discussion at the Afghanistan and international law workshop held by the Naval War College’s Stockton Center for International Law earlier this month.

Granular analysis of the practical consequences of the applicable legal regimes is beyond the scope of this post. However, in broad terms, status-based targeting is permitted in international or non-international armed conflict; individuals may be attacked either because of their affiliation with a particular armed group or based on conduct like directly participating in hostilities. In use of force situations that do not qualify as armed conflict, lethal force is only permissible based on the conduct of the individuals against whom it is employed. Specifically, it is allowed to defend oneself, others, and, arguably, certain property. Additionally, different detention rules apply in international and non-international armed conflicts, as well as in peacetime situations.

For better or worse, the general trend is in the direction of a merger of the international humanitarian law (IHL) rules governing the two categories of armed conflict, and international human rights law is proving increasingly influential in armed conflicts. But, here, the analysis is limited to identifying the legal regime that would govern the prospective U.S. OTH operations.

International Armed Conflict

 The applicability of IHL depends upon the existence of an armed conflict. Most IHL rules, whether treaty or customary in character, are intended for application in international armed conflict (IAC). Such conflicts occur when “hostilities” between two or more States arise. Although some minor disagreement exists in the IHL community over the requisite intensity of such hostilities, there is broad consensus that the threshold is low. For instance, the Department of Defense Law of War Manual defines IAC as “any situation in which there is hostile action between the armed forces of two parties, regardless of the duration, intensity or scope of the fighting” (section 3.4.2.).  There can be no doubt that U.S. OTH operations against Afghan forces would constitute “hostilities” for the purpose of classifying the situation as an IAC.

A more challenging question is whether operations against the Taliban may be characterized as against the state of Afghanistan given the lack of formal recognition of the Taliban as the country’s “legitimate” government. For instance, the Taliban has not assumed Afghanistan’s seat in the United Nations.

The Taliban’s status animated both earlier installments in this series. In Part I and Part II, I took the position that a group in control of a State’s territory and its instruments of governance is a de facto government and must be treated as acting for that State. This position, as was discussed, reflects the law of State responsibility’s treatment of members of a de facto government as “organs” of the State. That being so, their conduct is attributable to the State on the same basis as a State’s formal officials and its agencies; there is no need to assess whether the actions in question were the result of “instructions, or direction or control” of the State concerned, as is usually the case with attribution of a non-State actor’s conduct. The UN International Law Commission’s Articles on State Responsibility (ASR) capture these standards in Articles 4 and 8, respectively.

In my estimation, there is no reason to treat the Taliban differently with respect to classifying the conflict as international. Indeed, when the United States launched strikes against both al Qaeda and the Taliban in October 2001, the conflict was treated as international even though only three governments recognized the Taliban as Afghanistan’s government. Many of the legal debates that surfaced once hostilities began centered on aspects of IAC treaty law to which Afghanistan was party. Noteworthy in this regard were the debates over the relationship between the 1949 Geneva Convention III on prisoners of war and its companion treaty, Geneva Convention IV, which addresses civilians. Had the conflict not been international, those debates would have made little sense.

This interpretation appears to be consistent with that set forth in the DoD Law of War Manual (section 3.3.2. ):

Even if a State does not recognize an opponent as the legitimate government of a State, under certain circumstances, rules of international armed conflict may apply to a conflict between a State and a government that it does not recognize. For example, members of the regular armed forces who profess allegiance to a government or authority not recognized by the Detaining Power nonetheless would be entitled to POW status if they fall into the power of the enemy during international armed conflict.

Accordingly, IHL rules applicable in IACs would govern any OTH strikes against the Taliban. In some cases, this law will also apply to OTH operations targeting organized armed groups (OAG). As noted by the Appeals Chamber of the International Criminal Tribunal for the former Yugoslavia (ICTY) in its Tadić decision on jurisdiction, an IAC also exists when there are hostilities between a State and an OAG that is operating under the “overall control” of another State; in such circumstances, the two States concerned are both parties to the IAC (para. 145). The Chamber held that the State’s involvement must go “beyond the mere financing and equipping of such forces and involve[e] also participation in the planning and supervision of military operations. By contrast, international rules do not require that such control should extend to the issuance of specific orders or instructions relating to single military actions.” The International Criminal Court (ICC) has also adopted this standard (e.g., Lubanga, para. 541).

Note that “overall control” is a lower standard than “effective control,” which is the threshold for attributing actions of an OAG to a state under the law of State responsibility (ASR Commentary, page 48). As the ICRC has correctly noted in its recent commentaries to the first three 1949 Geneva Conventions, this distinction paradoxically could result in a State becoming a party to an IAC based on its relationship with an OAG, but that State would not be responsible for the group’s IHL violations because it does not exercise effective control over the group’s actions (see, e.g., 2020 GC III Commentary, paras. 302-306). Nevertheless, the ICJ confirmed the distinction in its 2007 Genocide Convention judgment (paras. 404-406). Thus, if the Taliban is in overall control of a non-State organized armed group on Afghan territory, U.S. OTH operations against that group would be governed by IAC law, including that involving targeting and detention.

Non-international Armed Conflict

OTH operations that are not directed at either the Taliban or a group operating under its “overall control” usually take place in the context of a non-international armed conflict (NIAC). Examples include the U.S. NIACs with al Qaeda and ISIS that have been underway for years.

NIACs involve hostilities between a State and an OAG (or between OAGs) and are governed by a smaller collection of IHL rules, some of which vary from their IAC counterparts. There are three requirements for the hostilities to comprise a NIAC. First, the armed group concerned must be sufficiently organized. Second, the violence between the State and the group must be protracted. Finally, the fighting must reach a certain threshold of intensity. (On NIACs, see Yoram Dinstein’s comprehensive treatise, Non-International Armed Conflict’s in International Law).

The organization requirement distinguishes a NIAC from unorganized violence, such as a riot by a mob. It stems from the OAG’s status as a “Party” to the conflict in Common Article 3 of the four 1949 Geneva Conventions. Absent some degree of organization, there would be no “Party” to comprise the State’s adversary.

To an extent, an OAG must constitute a cohesive armed force, although, as noted by the ICTY, there is no requirement that its organizational structure mirror that of traditional State armed forces (Boskoški, para. 197). It should operate in a coordinated, collaborative manner, with shared leadership at some point in the chain of command, although the execution of operations can be highly decentralized. Indicia of organization include “the force or group’s internal hierarchy; the command structure and rules; the extent to which military equipment, including firearms, are available; the force or group’s ability to plan military operations and put them into effect; and the extent, seriousness, and intensity of any military involvement” (Lubanga, para. 537).

This raises the question of how to characterize separate groups that operate with a common purpose. It is a determinative matter because an OAG must satisfy the protracted and intense violence requirements discussed below before its operations qualify as taking place in a NIAC. If the group is part of a larger OAG that has met the criteria, operations against the former will be governed by the IHL applicable in NIACs even if its actions, or those directed against it, are insufficiently protracted or intense. However, if its identity is distinct, IHL will not govern operations directed against it until the requisite thresholds of protractedness and intensity are reached. In my estimation, groups that operate with entirely independent chains of command or fail to coordinate and collaborate to a degree are separate for classification of conflict purposes. The mere fact that they have a common enemy or are motivated by shared values does not qualify them as components of a single OAG.

To illustrate, the U.S. Government has asserted in litigation before the D.C. District Court (Gul v. Biden) that “[a]lthough al-Qaida’s remaining core leadership has been degraded and currently poses a more limited threat to the U.S. homeland, that leadership continues to assert direction for, seek the loyalty of, and encourage cooperation and growth among regional al-Qaida affiliates, which continue to pursue recruitment and plotting, especially in unstable or vulnerable areas, and threaten local U.S. personnel, interests, and partners.” (On-file with author). If the assertion is accurate, “direction” and “cooperation” would satisfy the organization criterion for the various al-Qaeda affiliates.

By contrast, the practice of merely swearing “bay’at” (an oath of allegiance to a leader), as is often done with respect to ISIS, does not, without more, suffice to render the group part of the broader OAG. As I have maintained elsewhere,

Only in those cases in which a superior-subordinate organizational relationship emerged whereby ISIS directed and controlled the operations of the group swearing allegiance, did the latter become part of ISIS’ OAG. Should this occur, support by the group of ISIS undertakings may take the form of [e.g.] intelligence activities, such as surveilling potential targets. But absent such a relationship, and even if the group periodically cooperates and collaborates with ISIS, the groups would not be part of ISIS’ OAG.

It must be cautioned that this view is not universally shared, a point made at the Stockton Center workshop. Moreover, the discussion above bears only on classification of a conflict as non-international under IHL. It has no direct bearing on qualification of a group as an “associated force” under the Authorization for the Use of Military Force (AUMF). Nevertheless, the United States took an analogous approach in 2016 when it noted, “Merely engaging in acts of terror or merely sympathizing with al-Qa’ida or the Taliban is not enough to bring a group within the scope of the 2001 AUMF. Rather, a group must also have entered al-Qa’ida or the Taliban’s fight against the United States or its coalition partners.”

The second requirement is that the hostilities be “protracted.” The ICTY’s Appeals Chamber set forth the condition in its Tadić judgment (para. 70). It also appears, for instance, in Article 8 of the ICC Statute, which sets forth war crimes over which the court has jurisdiction during a NIAC.

Unfortunately, the period that suffices as protracted is unclear. For example, in the 1997 Abella (Tablada) case, fighting in Argentina raged for just over one day. Nevertheless, the Inter-American Commission on Human Rights opined that this period satisfied the requirement; therefore, Common Article 3 of the Geneva Conventions and other rules applicable in NIACs governed the incident (para. 156). Although a bright-line test for determining when hostilities are sufficiently protracted remains elusive, that was a highly questionable conclusion.

The condition of being protracted has implications for OTH operations into Afghanistan against OAGs that are not affiliated with the Taliban. By it, “isolated and sporadic” (a term appearing, inter alia, in Article 8) attacks against the United States by a group would not trigger a NIAC no matter how intense they might be, nor would an isolated and sporadic U.S. response.

Thus, for instance, a severe terrorist attack that is unlikely to be repeated would not be subject to the IHL that applies in NIACs. This is so even if the attack occurs after a period that qualified as a NIAC. Accordingly, if an OAG such as al-Qaeda goes into a long period of inactivity, a later attack mounted by a reconstituted group would have to separately satisfy the protracted criterion assessment. Of course, if the group used the period for purposes related to the earlier NIAC, as in recruiting, seeking further financing and arms, planning, intelligence gathering, and the like, the original NIAC would continue.

This dynamic is related to the third requirement, intensity, which was set forth in the ICTY Trial Chamber’s 1997 Tadić judgment to distinguish a NIAC from “banditry, unorganized and short-lived insurrections, or terrorist activities” (para. 562). NIACs must be both protracted and intense, a point made in 2014 by an ICC trial chamber in Katanga (para. 1217).

There is no bright-line test for the requisite intensity of a NIAC. Article 1(2) of the 1977 Additional Protocol II is often cited as bearing on the matter, as it excludes “internal disturbances and tensions, such as riots” from the ambit of the NIACs that instrument addresses. And drawing on both ICTY and ICC case law, Yoram Dinstein has observed,

The indicia include (inter alia) the numbers of casualties; the diffusion of violence over territory; the deployment of military units against the insurgents; the types of weapons used; the siege of towns; and the closure of roads. The type of weapons used is often an acid test: when tanks, artillery or attack helicopters are in operation, the intensity bar is crossed (p. 45).

At the Stockton Center workshop, there was an interesting discussion about when a NIAC is over such that IHL no longer governs any residual hostilities that may occur. A suggestion was made that once the intensity threshold is reached, the fact that the level of violence subsequently drops and remains at a lower level does not mean the NIAC is over. This is an erroneous assertion.

It may be the case that the level of violence waxes and wanes, as it does in most armed conflicts, whether international or non-international. If the violence dips below the requisite level of intensity only temporarily, the NIAC continues unabated. This is particularly important in the case of transnational terrorism qualifying as a NIAC because breaks are common between acts of violence that, when aggregated, reached the intensity threshold. But once the dip in the level of violence between an OAG and the state is no longer temporary, the NIAC has ended even. This is so even though isolated and sporadic or insufficiently intense hostilities continue to occur. The same is true if the group breaks apart and can no longer satisfy the organization criterion.

The question remains whether IHL applies to the U.S. operations in situations where the qualifying violence occurred outside Afghanistan; this was the case in 2001 with al Qaeda and its attacks into the United States.

International humanitarian law’s geographical scope of applicability during a NIAC has been contentious for some time. It derives from the fact that Common Article 3, which is generally seen as defining “non-international armed conflict,” refers to “an armed conflict not of an international character occurring in the territory of one of the High Contracting Parties” (emphasis added).

There are three views. By the first, the NIAC is limited to the borders of the State that is a Party to that conflict. However, this narrow view is fading away, in no small part because NIACs often “spill over” into neighboring States. Therefore, the ICRC has maintained that IHL continues to apply to activities with a nexus to the conflict in those areas. This is a reasonable approach, although it begs the question of the precise legal justification for application in the spill-over area, but not beyond.

And that is the crucial question. What about operations beyond the spill-over region, a fundamental issue for OTH operations into Afghanistan against an OAG that may have launched attacks against the United States far from that country? In its Law of War Manual (para., the DoD has taken the position that

Non-international armed conflicts … are classified as such simply based on the status of the parties to the conflict, and sometimes occur in more than one State. The mere fact that an armed conflict occurs in more than one State and thus may be characterized as international “in scope” does not render it “international in character.”

I have examined the issue elsewhere and concluded this is the best interpretation of the law. However, be aware that although IHL might govern an OTH operation conducted against an OAG with which the United States is involved in a NIAC, there must still be a legal basis for mounting the operation into the country concerned, here Afghanistan. That matter was the subject of Part II in this series.

OTH Operations Outside Armed Conflict

The way an OTH operation is conducted outside an armed conflict is subject primarily to the domestic law of any State enjoying prescriptive (and subsequently enforcement and judicial ) jurisdiction over the operation and to international human rights law. The reader is cautioned that the applicability of both bodies of law is exceptionally complex, and a comprehensive discussion of the matter is well beyond the scope of this post.

Importantly, prescriptive jurisdiction over a U.S. OTH operation would not be limited to the United States. To take but one example, States generally enjoy jurisdiction over actions completed on their territory pursuant to the “objective territoriality” principle; thus, Afghanistan’s domestic law would apply. Yet, the application of domestic law is subject to any territorial constraints found in that law and to personal or functional immunities that might apply under international law.

The applicability of international human rights law obligations, especially the right to life and the detention rights residing, inter alia, in the International Covenant on Civil and Political Rights, is likewise complicated. Most important is the question of whether the United States would even owe a human rights law obligation to individuals against which OTH action is taken. The question of the extraterritoriality of human rights obligations looms especially large in this regard (see Marko Milanovic).

For instance, the United States has taken the position that the Covenant on Civil and Political Rights does not apply beyond the borders of the United States (see Fourth Periodic Report to the Human Rights Committee, para. 505). Additionally, the prevailing view among States and international human rights law tribunals is that States only bear human rights law obligations when they exercise control over the territory concerned or the individual whose rights are affected. This view led the European Court of Human Rights in Bankovic to famously reject jurisdiction in a situation involving a lethal airstrike into the Federal Republic of Yugoslavia during the 1999 IAC between NATO members and that state. The court held that the attacking States did not exercise effective control over the deceased (para. 75). More recently, the Grand Chamber of the European Court of Human Rights declined to apply the European Convention’s human rights law regime to a phase of armed conflict between Russia and Georgia in which the belligerents were actively vying for control over territory.

The scholarly community increasingly challenges these approaches are. Over time, an interpretation by which a State owes human rights obligations to individuals when it can control the exercise or enjoyment of human rights, as was the case in Bankovic vis-à-vis the right to life, is likely to emerge. The trend seems to be in that direction (see, e.g., herehere, and here). However, it probably has not reached that point yet.

Whether human rights law applies to such OTH operations would have significant bearing on how they are carried out. In particular, human rights law does not permit status-based targeting. Instead, a law enforcement paradigm by which the lethal action must be in relatively immediate defense of individuals at risk of death or grievous bodily injury applies. Moreover, international human rights law imposes significant limitations on detention. Any deviation from these human rights rules would be subject to a strict regime of limitation by which that deviation must, inter alia, be for a legitimate aim, be provided for in law, occur only after less intrusive means have been exhausted, and be proportionate to the state’s interest in the circumstances. This is a high threshold to have to cross.


 In Part I of this series, I examined when a State may engage in OTH operations, with particular focus on President Biden’s announcement that the United States was developing such a capacity in light of our withdrawal from Afghanistan. Three were identified: authorization or mandate by the UN Security Council, self-defense, and request from another State.

Part II addressed the follow-on question of where OTH operations may be conducted.  With regard to Afghanistan, I concluded that U.S. operations into the country would be permissible with Security Council authorization or mandate; with the consent of the Taliban; in self-defense against an “armed attack” by the Taliban or non-State groups that operate on the Taliban’s behalf; or when independent non-State groups conduct an armed attack against the United States if the Taliban is “unwilling or unable” to put an end to their activities in Afghanistan.

Finally, in this third installment, I have explained that the law governing how OTH operations must be conducted depends on whether they occur during an IAC, a NIAC, or during peacetime. International humanitarian law would apply in the first two cases, although the applicable rules in that body of law would be determined by whether the OTH operation occurred in an IAC or NIAC. Outside of armed conflict, the law of States exercising jurisdiction would provide the legal framework, alongside international human rights law. In that regard, the applicability of human rights law extraterritorially is a hotly contested issue.


Michael N. Schmitt is the G. Norman Lieber Distinguished Scholar at the United States Military Academy. He is also Professor of Public International Law at the University of Reading, Strauss Center Distinguished Scholar and Visiting Professor of Law at the University of Texas, and Professor Emeritus at the United States Naval War College.