May 30, 2012 | Boston University International Law Journal

A Defense of Drone Attacks in Pakistan Under Humanitarian Law

May 30, 2012 | Boston University International Law Journal

A Defense of Drone Attacks in Pakistan Under Humanitarian Law

Since the terrorist attacks of September 11, 2001, international law has had to grapple with the fundamental challenges that large-scale violence carried out by non-state actors poses to the traditional inter-state orientation of international law. Questions related to the “adequacy” and “effectiveness” of international humanitarian law, international human rights law and the law related to the use of force have been particularly pronounced. This paper focuses on the international humanitarian law implications of American drone attacks in northwest Pakistan. A highly-advanced modality of modern warfare, armed drones highlight the possibilities, problems, prospects and pitfalls of high-tech warfare. How is the battlefield to be defined and delineated geographically and temporally? Who can be targeted, and by whom? Ultimately, this paper concludes that American drone attacks in northwest Pakistan are not unlawful as such under international humanitarian law, though, like any tactical decision in the context of asymmetric warfare, they should be continuously and closely monitored according to the dictates of law with sensitivity to facts on the ground.


When the American people elected Barack H. Obama President of the United States on November 4, 2008, many observers envisaged a new focus for the United States’ foreign policy. Gone, it was thought, was the derided “cowboy diplomacy” of his predecessor in the White House, President George W. Bush. Three of President Obama’s first five executive orders made significant policy changes that were expected to positively influence foreign relations: the closing of detention facilities in Guantanamo Bay, Cuba, ordering a review of detention policies and ensuring the lawfulness of interrogations.1 Domestically, a caustic presidential campaign gave way to an increase in support for the new leader, who would soon receive what is arguably the world’s highest secular honor, the Nobel Peace Prize.2 The title of President Obama’s 2009 Nobel Lecture, “A Just and Lasting Peace,” encapsulated the optimistic atmosphere. Many thought that nearly anything was possible now that a conscientious American leader with the will to ‘speak truth to power’ had arrived to guide the international community into the second decade of the twenty-first century. 

Ironically, both President Obama’s most ardent supporters and those cynics who derided his Nobel Lecture for being unrealistic and naıve seem to have missed a number of revealing nuances in his speech in Oslo. While stressing the need to adhere to internationally agreed upon rules regulating the use of force, he asserted a “right to act unilaterally if necessary to defend my nation.”3 However, he did not expressly state whether this doctrine equated to, approximated, or diverged from each State’s right of self-defense under the law related to the use of force.4 The world was also told evil still existed, that non-violent civil disobedience would not have stopped Hitler and the Holocaust and that negotiating with Al Qaeda would ultimately prove fruitless.5 As President Obama put it, “[t]o say that force is  sometimes necessary is not a call to cynicism – it is a recognition of history, the imperfections of man and the limits of reason.” 6 Despite the proud claim Obama made on the campaign trail in Berlin during the summer of 2008 that he was a “fellow citizen of the world,”7 today Guant ´ anamo remains open, the United States has not “resigned” the 1998 Rome Statute of the International Criminal Court, and many of the Bush Administration’s policies in the fight against terrorism remain in place, albeit without the language of the “War on Terror.”8 This is the context in which this article examines the lawfulness of the use of drone attacks in northwest Pakistan under international humanitarian law. 

Drone attacks in Pakistan are one of the most important and controversial aspects of the Obama Administration’s approach to fighting terrorism.9 This article begins by exploring developments in drone technology and translates that discussion into the context of the American drone campaign in northwest Pakistan. Assessing these attacks under international humanitarian law requires determining whether an armed conflict paradigm applies and, if so, how the armed conflict at issue should be classified. This article then examines three persistent issues that have arisen in the context of the American drone campaign in northwest Pakistan: first, the question of collateral damage, with particular reference to the drone attack that killed Tehrik-e-Taliban Pakistan (TTP) leader Baitullah Mehsud in August 2009; second, the concern raised in the 2010 Report of Philip Alston, the Special Rapporteur on Extrajudicial, Summary, or Arbitrary Executions, that “[it] is not possible for the international community to verify the legality of a killing, to confirm the authenticity or otherwise of intelligence relied upon, or to ensure that unlawful targeted killings do not result in impunity;”10 and, finally, the legal implications of Central Intelligence Agency (CIA) involvement in drone attacks. Ultimately, this paper concludes that American drone attacks in northwest Pakistan are not unlawful as such under international humanitarian law, though, like any tactical decision in the context of asymmetric warfare, they should be continuously and closely monitored according to the dictates of law with sensitivity to the  facts on the ground.


One of the great ironies of social contract theory is that despite humanity’s flight from the state of nature to a more organized and ‘“civilized” existence, violence remains a constant reality, both between and within States. Whether this state of affairs was, or is, inevitable is beyond the scope of this article. However, it is clear that technology has expanded and diversified the capacity for ‘“civilized”‘ violence. No longer limited to the fist, the rock or the boulder, individuals now have access to weapons of various degrees of precision and devastation, from the bullet to the nuclear weapon. This diversity in weaponry raises practical questions regarding accessibility and control, as well as normative questions related to how and when the use of such weapons is acceptable, if ever. Developments in drone technology should be viewed within this historical evolution toward “better” weapons. 

The first reported use of a “drone” was in 1919, when the inventor of autopilot technology and the gyroscope, Elmer Sperry, sunk a German battleship with a pilotless aircraft.11 The Vietnam War saw drones used for surveillance purposes.12 Drones have the advantage of being able to gather valuable intelligence without the inherent risk to human life that a traditional sortie by a piloted craft would pose.13 Drone gathered intelligence can be assessed in real time by multiple actors, including military lawyers, leading to a potential increase in transparency and accountability of the decision-making process, particularly if the drone at issue has an attack capability and the use of that capability is being contemplated.14 With no fear of retaliation, more “objective” decisions may be reached without the interference of fatigue and stress.15 Dozens of States now possess drone technology for surveillance purposes; however, as with military hardware and software generally, some States have more sophisticated capabilities than others.16

There are two generations of drones. The first generation is used exclusively for surveillance. The 2009 Manual on International Law Applicable to Air and Missile Warfare refers to these as “unmanned aerial vehicles,” or an “unmanned aircraft of any size which does not carry a weapon and which cannot control a weapon.”17 Depending upon the model, first generation drones can be flown remotely or can operate autonomously, can fly for days or for considerably shorter periods of time, can be built for multiple uses or for one mission only and can be as large as a traditional aircraft or much smaller.18 For example, the United States Army’s RQ-11 Raven B drone is about the size of a model plane that a young hobbyist might build in his or her basement.19 Second generation drones have an attack capability. The Manual refers to these as “unmanned combat aerial vehicles,” or an “unmanned military aircraft of any size which carries and launches a weapon, or which can use on-board technology to direct such a weapon to a target.”20 Depending upon the model, second generation drones have many of the same characteristics as first generation drones, but fewer States possess second generation drones.21 It is worth noting that the 2010 U.S. Army Unmanned Aircraft Systems Roadmap 2010-2035: Eyes of the Army foresees future drones capable of reconnaissance and surveillance of biological, chemical, nuclear and radiological weapons as well as high-yield explosives and hazards; contributing to security; operating as an attack platform in close combat and on interdiction and strike missions; facilitating improved command and control; supporting combat operations; delivering materials and supplies; and extracting personnel and damaged hardware.22

Israel and the United States have driven most of the developments in drone technology, but other key States are making rapid progress in the field.23 Beginning with Israel’s use of drones as decoys during the Yom Kippur War, its success with drones has continued to expand. During its armed conflict in Lebanon in the early 1980s, while acting as a decoy, the Israeli-developed Samson drone successfully triggered Syrian radar systems in the Bekaa Valley. This allowed Israel to destroy Syria’s considerable surface-to-air missile arsenal. Israeli drones also proved effective in monitoring Syrian-operated air bases and allowed the Israel Defense Forces (IDF) to track Syrian and Palestine Liberation Organization fighters. The sophistication of Israel’s drones has increased over time, and they have been used successfully in all major armed conflicts involving Israel over the last few decades. Drones played a particularly important role during Operation Cast Lead, Israel’s armed conflict with Hamas in the Gaza Strip between December 27, 2008 and January 18, 2009.24 Armed drone technology has supplemented the capabilities of Israeli surveillance drones and has played an important role in allowing Israel to retain its military advantage against hostile State and non-State actors in the Middle East.25

The United States’ first sustained experience with drones coincided with the first decade after the end of the Cold War. The 1990s saw the Pioneer drone fly more than three hundred missions during Operations Desert Shield and Desert Storm, searching for Scud missiles and other targets of significance for Coalition forces as they sought to dislodge Iraqi troops from Kuwait.26 During the North Atlantic Treaty Organization’s (NATO) intervention in the Balkans, President William J. Clinton used drones in both Bosnia and Kosovo.27 September 11, 2001 ushered in a new perspective for the United States on matters of international peace and security, with terrorism suddenly seen as more than just a law enforcement problem. As President Bush stated in an address to a joint session of the United States Congress shortly after September 11th, “[w]e will direct every resource at our command, every means of diplomacy, every tool of intelligence, every instrument of law enforcement, every financial influence, and every necessary weapon of war, to the disruption and to the defeat of the global terror network.”28 Drones, both first and second generation, would play an important role in this new struggle.29

The Bush Administration found great value in drone technology and used attack drones against targets in several countries, including Afghanistan, Yemen, Pakistan and Iraq.30 Under President Obama, the use of attack drones has notably accelerated. The Taliban, subdued but not defeated, has not become a non-violent political force in the new Afghanistan. Instead, aided by loyalists from strongholds in neighboring Pakistan, it continues to challenge what the United Nations Security Council has reaffirmed as the “sovereignty, independence, territorial integrity and national unity of Afghanistan” established following NATO’s 2001 invasion. 31 Relatedly, Al Qaeda’s reach, though never confined to Afghan territory alone, remains active in Afghanistan. Many of these issues involve, as President Obama has put it, the “cancer [that] is in Pakistan.”32

The border between Afghanistan and Pakistan may separate two sovereign States as a matter of law, but the Durand Line rarely functions as such in practice.33 This is the Pashtun heartland, a transnational tribalcultural geographical region with large populations of Pashtuns on either side of the border. It is because of this intermingling of culture and allegiance that the area is frequently referred to as “AfPak.” It is here where Al Qaeda’s then first- and second-in-command, Osama bin Laden and Ayman al-Zawahiri, were generally believed to be hiding until bin Laden was killed in an American raid on May 1, 2011 in Abbottabad, Pakistan. That bin Laden was found living in a large, secure compound of reinforced concrete some thirty miles from Islamabad and near the Pakistan Military Academy, with senior Pakistani military officials as neighbors, suggests the extent to which the “cancer” has spread.34 Whether due to unwillingness, inability or some combination of unwillingness and inability, 35 it seems clear that Pakistan has for some time been in breach of the substantial and exacting counterterrorism obligations that the International Court of Justice (ICJ) recognized as binding all States as a matter of customary international law in its 2005 Armed Activities on the Territory of the Congo judgment.36 As Anthony J. Blinken, National Security Advisor to United States Vice President Joseph R. Biden, reportedly put it in a conversation with Husain Haqqani, then Pakistani Ambassador to the United States, “[t]here is appeasement one day, confrontation another day, and direction a third day.”37

The 2010 Report of the United Nations Commission of Inquiry into the Facts and Circumstances of the Assassination of Former Pakistani Prime Minister Mohtarma Benazir Bhutto (Bhutto Report) illustrates the extent to which Islamist terror has come to debilitate the Pakistani State and establish northwest Pakistan as a virtually autonomous terrorist controlled territory.38 While serious questions about the culpability for Prime Minister Bhutto’s assassination remain, what is undeniable from the Bhutto Report is the ineptitude and factionalism of the Pakistani State on security matters and the potent threat Islamist militants continue to pose to it. The Bhutto Report is a tragic account of violence, largescale police raids on the media, desaparecidos, the Red Mosque confrontation between the State and Islamist militants in Islamabad, suicide bombings throughout Pakistan (directed at security forces and civilians alike) and a Pakistani military unable to stop Islamist militants from establishing “no go” areas in geopolitically strategic parts of the country. 39 It is a story of Al Qaeda and the Taliban, both the TTP and the Afghan Taliban.40 The Bhutto Report’s depictions of Pakistan in 2007, the year Prime Minister Bhutto was assassinated in Rawalpindi, describe that year as “one of the most violent years in Pakistani history, with dramatic increases both in extremist attacks carried out by radical Islamists against local targets, including suicide bombings, and in the use of force by the authorities against opposition movements.”41 A report by the non-governmental organization Campaign for Innocent Victims in Conflict echoes this situation, estimating that the number of civilian casualties in Pakistan in 2009 due to the armed conflict in the northwest of the country exceeded the number of civilian casualties in Afghanistan during the same year.42 The American drone campaign in northwest Pakistan must be examined within this context.

Although dispositive conclusions about the lawfulness of individual American drone attacks in northwest Pakistan under international humanitarian law do not hinge upon statistics, considering the total number of deaths and injuries to civilians and the degree of damage to civilian objects in relation to the total number of members of the Taliban and its Al Qaeda allies killed and injured, as well as insurgent assets destroyed,43 does help one better appreciate the impact the drone program has had on the ground. Of course, statistics should be approached with a critical mind and carefully assessed in terms of methodological rigor, relevance and applicability.44 Gathering statistics for drone attacks in northwest Pakistan is complicated by the presence of the Pakistani Armed Forces and restrictions on media access to areas where drone attacks take place. The statistical studies that do exist rarely define in a methodical manner the distinction between, on the one hand, combatants and civilians who directly participate in hostilities and, on the other hand, bona fide civilians, that is, civilians who do not directly participate in hostilities.45 These shortcomings make it difficult to ascertain legally crucial facts, such as whether any of the purported civilians killed or injured at the time were acting as human shields, either voluntarily or involuntarily, and whether any of the purported civilians were directly participating in hostilities at the time.46 Nonetheless, it is worth looking at three of the most significant statistical studies that focus on northwest Pakistan: a February 2010 report conducted by Bergen and Tiedemann under the auspices of the New America Foundation (NAF),47 a July 2010 British Broadcasting Corporation (BBC) study48 and an ongoing research project carried out by Roggio and Mayer under the auspices of the Long War Journal (LWJ).49

The NAF report covers the period between 2004 and February 24, 2010, its date of publication. Particularly striking is that the number of drone attacks that took place during the first fifty-five days of 2010 were, at eighteen, exactly twice the number of drone attacks that took place during the entire four years from 2004 to 2007.50 Overall, during the almost six years and two months covered by the study, the United States carried out 114 drone attacks, resulting in between 830 and 1,210 total deaths, with between 550 and 850 of the dead being militants.51 This means that slightly more than 30% of deaths from drone attacks in the study were civilian deaths, with the percentage dropping to slightly less than 25% if one focuses only on 2009, the most active year covered by the study.52 Ongoing reporting by the NAF continues to bring the Bergen and Tiedemann study up to the present.53 

The BBC study confirms that although drone attacks took place during the Bush Administration, they have increased significantly since President Obama took office in January 2009. Between January 2008 and January 2009 there were twenty-five drone attacks in northwest Pakistan. From late January 2009, when Obama became President, through June 2010, there were nearly ninety such attacks.54 The number of deaths during these respective periods is estimated to have been just under 200 during the former and over 700 during the latter.55 The period under review also saw a reported widening of potential targets of American drone attacks, from Al Qaeda alone to include Al Qaeda’s ally the TTP.56

Finally, the LWJ’s ongoing research project also shows a steady and consistent increase in the number of drone attacks since 2004. It gives raw figures for civilian casualties and Taliban and Al Qaeda casualties, as well as distributions for drone attacks over various tribal areas in northwest Pakistan. The research indicates drone attacks have killed 2,195 militants and 138 civilians since 2006.57 An early January 2010 report from Roggio and Mayer suggests that drone attacks have become more precise over time, which has correlated with a corresponding decrease in collateral damage.58

As these studies reveal, the number of member of the Taliban and its Al Qaeda allies who have been killed and injured in northwest Pakistan as a result of drone attacks considerably exceeds the number of deaths and injuries to civilians, though it is important to note that international humanitarian law does not require or endorse any particular ratio.59 Since the introduction of attack drones in northwest Pakistan, several of the highest-ranking militants operating in the area, including Al Qaeda’s leader in Afghanistan and Pakistan, Sheikh Fateh al Masri; Al Qaeda’s chief finance officer, Mustafa Abu Yazid; Qari Mohammad Zafar, who the United States had wanted for his alleged involvement in a 2006 attack on the United States Consulate in Karachi; senior Al Qaeda operative and member of its military shura, or council, Mustafa al Jaziri; militants wanted by the United States for their alleged involvement in the 1998 bombings of the United States Embassies in Tanzania and Kenya; and TTP leader Baitullah Mehsud, have been killed.60


It is undeniable that American drone attacks in northwest Pakistan have had a significant impact in terms of deaths and injuries to civilians and damage to civilian objects. In light of the stated purpose of these attacks, to facilitate the defeat of the Taliban and its Al Qaeda allies, this quantum of harm may or may not be  justifiable in terms of ’morality, ethics or policy, but these considerations are not, or are at least not wholly, considerations that determine ’the legal analysis. For example, it is possible to support the use of drone attacks, either in particular situations or entirely, as a matter of morality, ethics or policy, and still conclude that any attack is unlawful. The inverse position is also possible.61 

This article’s legal frame of reference for assessing drone attacks is rooted in international humanitarian law, and as such, it operates irrespective of the law related to the use of force.62 Put differently, whether or not drone attacks are legal under international law related to the use of force is not dispositive as to their legality under international humanitarian law. As with the juxtaposition of considerations of morality, ethics and policy, there need not be any correlation between ’the legal analysis under the law related to the use of force and’ the legal analysis under international humanitarian law. As a matter of law, these are completely separate analyses.63 

Before assessing the lawfulness of each American drone attack in northwest Pakistan under international humanitarian law, it is necessary to first draw the  distinction in law between situations of violence and situations of armed conflict and then to understand how international humanitarian law classifies situations of armed conflict. 

i. Determining Whether a Situation of Violence Amounts to a Situation of Armed Conflict

For decades following the Second World War, international humanitarian law did not provide a clear definition of armed conflict, despite the fact that international humanitarian law’s application relies upon the existence of an armed conflict. None of the Four Geneva Conventions of 1949,64 nor either of the two 1977  additional Protocols to the Geneva Conventions65 define armed conflict, and the consensus view is that the existence of an armed conflict is determined on the basis of the particular facts and circumstances.66

A number of important developments in international law in the last two decades have begun to clarify the threshold between armed conflict and situations falling short of armed conflict. Three of these are worth considering here: Prosecutor v. Tadi´c, the seminal case heard by the International Criminal Tribunal for the Former Yugoslavia (ICTY);67 the Supreme Court of Israel’s 2006 judgment in Public Committee Against Torture v. State of Israel;68 and the 2010 Final Report of the International Law Association’s Use of Force Committee on the Meaning of Armed Conflict in International Law (Final Report).69 

Tadi´c stands as perhaps the most consequential decision of any international criminal tribunal since Nuremberg. One of the reasons for this is because it gives a comprehensive definition of armed conflict. The ICTY Appeals Chamber determined an armed conflict exists “whenever there is a resort to armed force between States or protracted armed violence between governmental authorities and organized armed groups or between such groups within a State.”70 After the Appeals Chamber rejected Du˘sko Tadi´c’s jurisdictional challenge, the case returned to the Trial Chamber for a decision on the merits. The Trial Chamber affirmed the Appeals Chamber’s comprehensive definition of armed conflict in its judgment and then clarified it by noting that the threshold between armed conflict and situations falling short of armed conflict requires inquiry into the “intensity of the conflict and the organization of the parties to the conflict.”71 The threshold for armed conflict itself contains two sub-thresholds, intensity and organization, and both of these must be satisfied before an armed conflict can be said to exist as a matter of law.72

Targeted Killings also provides some clarity as to the threshold between armed conflict and situations falling short of armed conflict. In that case, the Israeli Supreme Court was tasked with assessing the lawfulness of Israel’s policy of targeted killings in the West Bank and Gaza. Between the outbreak of the Second Intifada in 2000 and 2005, Israel’s targeted killing policy resulted in the death of nearly 300 suspected terrorists, over 100 civilian deaths and hundreds of injuries.73 The most famous example of this policy was the 2004 attack on Sheikh Ahmed Yassin, founding member of Hamas, when an Israeli helicopter gunship killed him in Gaza. In discussing whether an armed conflict paradigm applied to Israel’s targeted killing policy, the court boldly stated that “there is no doubt today that an armed conflict may take place between a state and groups or organizations that are not states, inter alia because of the military abilities and weapons in the possession of such organizations and their willingness to use them.”74 The court was also clearly influenced by the State Attorney Office’s argument that Israel had suffered proportionately a much greater loss in terms of terrorism victims during the period under review than the United States had on September 11, and that these attacks were in the nature of an “‘unceasing, continuous and murderous barrage of attacks, which are directed against Israelis wherever they are, without any distinction between soldiers and civilians or between men, women and children.’”75 Thus, Targeted Killings’ main contribution to the discussion lies in its focus on the willing use of violence on a significant scale by an organized group.76

A final important development in international law in the last two decades about the threshold between armed conflict and situations falling short of armed conflict is the Final Report. It gives an overview of how international law has classified situations of violence during three time periods: 1945-80, 1980-2000, and 2000-10.77 It concludes, along the same lines as Tadi´c, that “all armed conflict involves, at a minimum, intense fighting among organized armed groups.”78 In addition to providing significant reference to State practice, the Final Report distinguishes the two key Tadi´c criteria of intensity and organization with reference to the jurisprudence of international courts and tribunals.79 Relevant factors for intensity include the number of involved fighters, the war-fighting capacity, the number of casualties, the extent of population displacement and potential Security Council involvement.80 A number of factors may be persuasive in satisfying the organization element, inter alia, a leadership hierarchy, the provision of military training and a command structure.81

What this overview of Tadi´c, Targeted Killings and the Final Report suggests is that international law classifies a situation of violence as a situation of armed onflict where the “facts on the ground, the facts of fighting,” 82 are of a certain minimum intensity and when the fighters are organized.83 It is the focus on the two key Tadi´c criteria of intensity and organization that marks the Rubicon between “situations of internal disturbances and tensions, such as riots, isolated and sporadic acts of violence and other acts of a similar nature”84 and situations of armed conflict. Once this line has been crossed, international humanitarian law will apply.85 However, given the relative nature of the two criteria of intensity and organization, the fact that they can only be interpreted in light of particular facts and circumstances, one should be careful not to dismiss the “intrusion” of considerations of discretion and policy preference in this area of law.86 As Major General Charles J. Dunlap, Jr., United States Air Force (Retired), has put it, “[a]ll law, but especially LOAC [law of armed conflict], necessarily involves subjectivity implicit in human reasoning that may be troubling to those of a technical mind-set accustomed to the precision that their academic discipline so often grants.”87

ii. Classifying Situations of Armed Conflict

When a situation of violence amounts to a situation of armed conflict, how international humanitarian law applies depends upon how the armed conflict at issue is classified. The prevailing view since at least the adoption of the Four Geneva Conventions in 1949 until 1977 held there were two types of armed conflicts: (1) international armed conflicts between States and situations of occupation88 and (2) “armed conflict[s] not of an international character [occurring] in the territory of one of the High Contracting Parties.”89 1977 saw the “internationalization” of armed conflicts involving national liberation movements through the adoption of Additional Protocol I to the Geneva Conventions ( Additional Protocol I).90 Additional Protocol II to the Geneva Conventions (Additional Protocol II) created a set of treaty rules that would apply to armed conflicts that satisfied the Common Article 3 baseline created in 1949 for non-international armed conflicts and in addition involved fighting between the armed forces of a State in its territory and “dissident armed forces or other organized armed groups which, under responsible command, exercise such control over a part of its territory as to enable them to carry out sustained and concerted military operations and to implement this Protocol.”91 From 1977 until September 11, 2001, the consensus view was that international humanitarian law recognized three types of armed conflicts: international armed conflict, Common Article 3 noninternational armed conflict, and Additional Protocol II non-international armed conflict.92

September 11 demonstrated that a State can suffer an “armed attack”93 irrespective of whether such an attack can be attributed to another State and that the victim State can lawfully respond in self-defense on this basis.94 Since September 11, the ICJ has held that a victim State’s response in self-defense requires prior State attribution,95 but this thinking was not without pointed disagreements on the bench96 and has drawn criticism in the academic literature.97 It should also be recalled that although the ICJ has a unique role to play as the “principal judicial organ of the United Nations,”98 a divided court of fifteen judges in The Hague does not make international law.99 However, these are questions of the law related to the use of force, not questions of international humanitarian law. What is relevant from this context is that the attacks of September 11 were committed by a nefarious combination of a State actor in the form of the Taliban regime, the de facto government of Afghanistan at the time, and a non-State actor, the international terrorist organization Al Qaeda.

Faced with what Eyal Benvenisti refers to as “transnational warfare,” or “armed conflicts between state military forces and foreign non-state actors that take place beyond state borders,”100 the consensus view on classification of armed conflicts prior to September 11, though itself not without practical difficulties, has “crumbled.”101 Some explanation lies in the changing nature of warfare and the difficulties that even the most creative interpreters of legal language face when attempting to reconcile “old” law with “new” facts. The consensus view’s classifications were based on whether the armed conflict at issue was “international” or “not of an international character” and also considered Additional Protocol II addressing more “specific” types of non-international armed conflict. However, this view no longer seems to “neatly” apply in a transnational context of asymmetric warfare between various State and non-State actors that operate in shifting capacities with fluctuating capabilities.102 As the Israeli Turkel Commission put it in its 2011 report on the May 2011 Mavi Marmara flotilla incident off the coast of Gaza, “in reality, the complexities of modern warfare pose a significant challenge when classifying an armed conflict, since not all armed conflicts can be easily classified within the framework of the traditional definition.”103

The United States Supreme Court’s 2006 decision in Hamdan v. Rumsfeld reflects this view.104 In that case, the Court, although clearly aware of the transnational nature of the armed conflict that the United States was engaged in, interpreted Common Article 3’s “not of an international character” language as applying to any armed conflict that takes place other than between two States.105 This, according to the Court, was Common Article 3’s “literal meaning.”106 Justice Clarence Thomas, joined by Justice Antonin Scalia, strongly dissented from this interpretation, stating that President Bush’s interpretation of Common Article 3 to the contrary was “reasonable and should be sustained. The conflict with Al Qaeda is international in character in the sense that it is occurring in various nations around the globe.”107 It is difficult to fault either of these interpretations. As Justices Thomas and Scalia conceded, both are “plausible, and reasonable.”108 

For all of the Court’s faith in its interpretation being the interpretation that accorded with Common Article 3’s “literal meaning,” its position that Common Article 3 applies to any armed conflict that takes place other than between two States precisely because Common Article 2109 applies to armed conflicts between two States is itself incorrect; Additional Protocol I expressly brings within the scope of Common Article 2 armed conflicts between States and national liberation movements, and national liberation movements are, of course, non-State actors. The court’s reading of the Commentary to Common Article 3 also seems to be “overly generous “given that it clearly reflects the considerable reluctance by States at the time of the Geneva Conventions’ adoption in 1949 to countenance the application of international humanitarian law to armed conflicts within their territorial and maritime boundaries; that is, to internal armed conflicts.110 As the Commentary clearly states, “[s]peaking generally, it must be recognized that the conflicts referred to in Article 3 are armed conflicts, with armed forces on either side engaged in hostilities – conflicts, in short, which are in many respects similar to an international war, but take place within the confines of a single country.” 111The point is not that the court, or the dissent of Justices Thomas and Scalia, was or was not correct in its interpretation, but that the nature of “transnational warfare” makes less convincing hard and fast interpretations of what had hitherto been the consensus view regarding classification of armed conflicts.

Another example of the challenges raised by the contemporary context of classification of armed conflict is the 2006 Report of the Commission of Inquiry on Lebanon Pursuant to Human Rights Council Resolution S-2/1 (Lebanon Report).112 According to the report, even though it was only the IDF and Hezbollah that engaged in hostilities with one another during the summer of 2006, with the Lebanese Armed Forces (LAF) not actively participating in hostilities, the situation was said to have amounted to a “legally cognizable international armed conflict.”113 The Lebanon Report gave three reasons for this: (1) Hezbollah formed part of the Lebanese coalition government; (2) the Lebanese national narrative of Hezbollah’s resistance to Israel’s previous occupation of south Lebanon in past decades had somehow transformed it into a force with a “national” spirit; and (3) the IDF caused damage within Lebanese territory during the armed conflict, including attacks on the LAF.114 While these reasons could be challenged, what is particularly interesting for present purposes is that the Lebanon Report’s classification of the armed conflict, although admittedly a unique case, appears to be a “mirror image” of Hamdan’s reasoning; namely, that any armed conflict that takes place other than between two States is “not of an international character.” 

Although the classification question is clearly not an area of international law free of controversy or inconsistency, how a particular armed conflict is classified is actually less important than one might think when it comes to the application of many international humanitarian law norms. There are a number of reasons for this. As the ICJ determined in its 1986 Military and Paramilitary Activities in and Against Nicaragua judgment, Common Article 3 contains international humanitarian law’s “minimum rules applicable to international and to non-international conflicts.”115 The International Committee of the Red Cross’ (ICRC) Customary International Humanitarian Law (CIHL) affirms that many of the norms embodied in Common Article 3 generally apply irrespective of how the armed conflict at issue is classified under international humanitarian law, as a matter of customary international law.116 Common Article 3 norms apply as a matter of customary international humanitarian law to all armed conflicts, not simply to those “not of an international character.” Discussing the legality of targeting decisions in its 1996 Legality of the Threat or Use of Nuclear Weapons advisory opinion, the ICJ noted that the “cardinal principles contained in the texts constituting the fabric of humanitarian law”117 are the principles of distinction, proportionality and the Martens Clause.118 The ICJ did not concern itself with how the armed conflict at issue was actually classified. These particular international humanitarian law norms are reflected in both international treaty law and customary international law and must undergird any international humanitarian law analysis.119


Having set forth the distinction between situations of violence and situations of armed conflict and the debate regarding how international humanitarian law classifies situations of armed conflict, we can turn to some of the persistent issues and challenges that arise when assessing American drone attacks in northwest Pakistan. Is the United States engaged in an armed conflict with the Taliban and its Al Qaeda allies? If so, how is this armed conflict to be classified? What geographic and temporal constraints might apply? Given this article’s focus on drone attacks in northwest Pakistan, is the United States engaged in an armed conflict there?

’Though President Obama has called for a “new beginning between the United States and Muslims around the world, one based on mutual interest and mutual respect,”120 the United States remains of the view that it is engaged in an armed conflict with the Taliban and its Al Qaeda allies, which began on September 11 and continues to the present. This is clear from President Obama’s public statements, which by their very nature inject themselves into a context of international legal meaning and significance. 121 The United States Department of State Chief Legal Advisor Harold Hongju Koh reiterated this position in unmistakable language at
the 2010 annual meeting of the American Society of International Law in Washington, DC, when he stated that, “as a matter of international law, the United States is in an armed conflict with al Qaeda, as well as the Taliban and associated forces, in response to the horrific 9/11 attacks, and may use force consistent with its inherent right to self-defense under international law.”122 Thus, from the United States’ perspective, international humanitarian law applies to its engagements with the Taliban and its Al Qaeda allies. This means, of course, that international humanitarian law would be the legal frame of reference for assessing the lawfulness of American drone attacks in northwest Pakistan.

However, if the United States is not in an armed conflict with the Taliban and its Al Qaeda allies in northwest Pakistan, then the application of an international human rights law paradigm to the facts would determine one’s legal conclusions. Under international human rights law, it would seem to be more difficult to justify each American drone attack in northwest Pakistan, as international human rights law requires that States use force as a last resort, “to protect against concrete, specific, and imminent threats of death or serious physical injury.”123 The legal criteria of proportionality, necessity, imminence and legality would have to be carefully parsed and applied to the facts of each attack under this paradigm. 124 Having said that, however, an argument could be made that contemporary advances in information technologies and global capabilities have relaxed this standard somewhat, at least since September 11. According to this line of thinking, one could argue that American drone attacks on suspected terrorists in northwest Pakistan in the autumn of 2010 may not necessarily have been unlawful under international human rights law given an asserted increase in the terrorist threat to the United States and Europe, and to American interests in Europe,125 and given the “consent” of the Pakistani State.126 “IHRL [international human rights law] in this specific respect produces much the same result as would IHL [international humanitarian law], thereby reducing the significance of determining which model controls in the first place.”127 

As discussed above, the contemporary context of warfare has rendered “outdated” the traditional insistence that armed conflicts can take place only between States or within them. Such armed conflicts do take place, as they always have, but just as States can no longer credibly “hide behind” “matters which are essentially within the domestic jurisdiction of any state”128 in an era of each State’s “responsibility to protect its populations from genocide, war crimes, ethnic cleansing and crimes against humanity,”129 so it must also be recognized that a State’s territorial and maritime boundaries are as much capable of harboring violence as they are of preventing its intrusion from outside. It is revealing that one of President Obama’s first acts in office was to select Richard Holbrooke as Special Representative for Afghanistan and Pakistan, two States in one region, or, as the late Ambassador Holbrooke would put it, “two very distinct countries with extraordinarily different histories, and yet intertwined by geography, ethnicity, and the current drama.”130

To the extent a sufficient “nexus” exists between the armed conflict that ensued between the United States and the Taliban and its Al Qaeda allies in the wake of September 11, and the acts of a person who participates in hostilities, whether by his or her very nature, as a combatant, or directly, as a civilian,131 then it can be said that the “fight,” as a matter of law, “follows the fighter.”132 This was the case during the Second World War, when the United States Army Air Force shot down and killed Japanese Admiral Isoroku Yamamoto, the planner of the attack on Pearl Harbor, when he was flying to Bougainville Island on a planned inspection of Japanese soldiers, and it remains the rule today.133

The effect of this is that international humanitarian law applies both to the immediate area of hostilities, that is, within Afghanistan, and “further afield,” the only requirement being, to use the ICTY’s language in its 2002 Prosecutor v. Kunarac judgment, one of “substantial[ ] relat[ion].”134 In this sense, it would not be entirely correct to assert, as O’Connell does, that, “[a]rmed conflict has a territorial aspect. It has territorial limits. It exists where (but only where) fighting by organized armed groups is intense and lasts for a significant period.”135 Kunarac states “there is no necessary correlation between the area where the actual fighting is taking place, and the geographical reach of the laws of war.”136 Of course, this understanding of armed conflict must be read in conformity with the two key Tadi´c criteria of intensity and organization.137

An alternative view, distinct from the “extension” argument just discussed, is that the United States is engaged in an armed conflict in northwest Pakistan in particular given that it has intervened there in a way that has overlapped with the internal armed conflict in which Pakistan itself is engaged. The ICTY Appeals Chamber acknowledged this “overlap” understanding of armed conflict in its 1999 Tadi´c judgment when it stated:

It is indisputable that an armed conflict is international if it takes place between two or more States. In addition, in case of an internal armed conflict breaking out on the territory of a State, it may become international (or, depending upon the circumstances, be international in character alongside an internal armed conflict) if (i) another State intervenes in that conflict through its troops, or alternatively if (ii) some of the participants in the internal armed conflict act on behalf of that other State.138

It is also worth mentioning in this context that Pakistani Minister of State for Foreign Affairs Hina Rabbani Khar’s condemnation of American drone attacks in northwest Pakistan on March 17, 2011 was phrased in the language of armed conflict: “[s]uch strikes constitute a matter of serious concern and raise issues regarding respect for human rights and humanitarian law. Irresponsible and unlawful conduct cannot be justified on any grounds.”139 Admittedly, the Minister of State’s statement was not particularly clear in recognizing that the United States was engaged in an armed conflict with the Taliban and its Al Qaeda allies on the Pakistani side of the Durand Line, but it is a legally significant admission that contributes to the case for the existence of an armed conflict in northwest Pakistan. Proceeding from the position that the situation of violence that the United States is engaged in with the Taliban and its Al Qaeda allies in northwest Pakistan amounts to an armed conflict, international humanitarian law would be the appropriate legal frame of reference for assessing the lawfulness of individual drone attacks in northwest Pakistan.

Given that the United States has conducted almost 300 drone attacks in northwest Pakistan in recent years and that international humanitarian law would require an exacting and individualized assessment for each of these attacks, space constraints preclude broad and sweeping generalizations about the compliance of each of these attacks under this branch of law. International humanitarian law is extraordinarily fact intensive, and the meaning attached to many of its key principles, in particular the principle of proportionality, is often contested and prone to political manipulation. 140 As the ICJ noted in its 1980 Interpretation of the Agreement of 25 March 1951 Between the WHO and Egypt advisory opinion, “a rule of international law, whether customary or conventional, does not operate in a vacuum; it operates in relation to facts and in the context of a wider framework of legal rules of which it forms only a part.”141

The rest of this section examines three persistent issues identified at the beginning of the paper: (1) collateral damage; (2) accountability to the international community; and (3) the legal implications of CIA involvement in the drone attacks. These issues reveal some of the main challenges from an international humanitarian law perspective. First, although international humanitarian law recognizes that it is “[unable] to eliminate the scourge of war . . . [and instead] endeavours to master it and mitigate its effects,”142 the principle of proportionality does not forbid collateral damage when such damage is outweighed by a particular attack’s “concrete and direct military advantage anticipated.”143 Indeed, it would not be putting it too strongly to say that State practice
supports, even endorses, the inevitability of collateral damage, the “incidental loss of civilian life, injury to civilians, damage to civilian objects, or a combination thereof.”144 Of course, international humanitarian law does prohibit disproportionate attacks,145 and violations of this prohibition can entail State responsibility,146 individual criminal responsibility147 or both.148 However, it is clear that collateral damage as such is not necessarily unlawful under international humanitarian law and that proportionality is a calculus of intangibles that balances military and civilian concerns. Proportionality must be assessed within the context of particular facts and circumstances, and conclusions of law cannot be drawn in abstracto.

With this understanding of the international humanitarian law principle of proportionality, one can begin to assess the legal implications of the drone attack that killed Baitullah Mehsud. When Hellfire missiles from an American Predator drone killed Mehsud in South Waziristan, it was reported that he perished along with his wife, his mother- and father-inlaw, seven bodyguards and a TTP lieutenant.149 It was alleged that Mehsud played a key role in supporting the insurgency in Afghanistan and in numerous acts of terrorism inside Pakistan itself, including the Marriott Hotel bombing in Islamabad in September 2008 and the assassination of Prime Minister Bhutto.150 Given Mehsud’s leadership position in the TTP and the crucial role he played in the ongoing armed conflict in AfPak, opposing both the United States and Pakistan, the drone attack seems to have been proportionate despite the reasonable foreseeability of collateral damage. It would be disingenuous not to recognize this as being as much an ethical judgment as a legal judgment, but such is the nature of legal conclusions in this area of law.151

O’Connell suggests that the drone attack on Mehsud likely violated international humanitarian law because the TTP leader had, at the time of the fatal attack, reportedly been receiving an intravenous transfusion and, as such, was an enemy person hors de combat.152 An enemy person cannot be targeted and does not qualify as a military objective when he or she is bona fide hors de combat.153 According to the Commentary to Common Article 3, the essence of this “categorical imperative”154 is that the sick and wounded must be “respected and protected.”155 CIHL defines an enemy person hors de combat as one who is “no longer participating in hostilities, by choice or circumstance.”156 Article 41(2)(c) of Additional Protocol I gives an exclusive list of three subtypes of enemy persons hors de combat, and the relevant subtype for present purposes is the third subtype, that of an enemy person who “has been rendered unconscious or is otherwise incapacitated by wounds or sickness, and therefore is incapable of defending himself.”157 It is important to stress, however, that none of these subtypes apply if the enemy person at issue does not “abstain[ ] from any hostile act.”158 As the Commentary to Article 41 of Additional Protocol I stresses, “[a]ny hostile act gives the adversary the right to take countermeasures until the perpetrator of the hostile act is recognized, or in the circumstances, should be recognized, to be ‘hors de combat’ once again.”159 

O’Connell’s criticism of the drone attack on Mehsud on enemy person hors de combat grounds is unconvincing. As the TTP leader at the time of the attack, Mehsud was a highly-skilled and disciplined militant who survived previous threats and drone attacks and was wanted by both the United States and Pakistan.160 It is reasonable to assume that he would not have left himself exposed and defenseless in a region well known for American drone strikes, which South Waziristan surely was and remains. Even if one assumes that Mehsud’s relatives were civilians bona fide, his over half a dozen bodyguards and the TTP lieutenant were not. Was Mehsud communicating with his bodyguards or with the lieutenant about military matters at the time of the attack? It seems reasonable to conclude that he was given the circumstances. If he was, then, according to the Commentary to Article 41 of Additional Protocol I, this would have been considered a hostile act that would have divested him of whatever protected status he might otherwise have been able to claim as an enemy person hors de combat.161 Even if one were to conclude that Mehsud had been an enemy person bona fide hors de combat at the time of the attack and, as such, could not have been targeted, it is important to recognize that Mehsud’s bodyguards and the TTP lieutenant were military objectives and, as such, could have been targeted. Given the hostile terrain in AfPak and the real challenges posed by the embedding of civilians bona fide by militants in violation of international humanitarian law,162 it could be argued that an attack directed at Mehsud’s bodyguards and the TTP lieutenant alone would not have been disproportionate in an alternative proportionality calculus despite reasonably foreseeable risk of collateral damage to Mehsud and his relatives.163 A final point is that while O’Connell is certainly correct to note that doubt should be resolved in favor of recognizing a person’s protected status,164 a “reasonable man”165 could have had no doubt in Mehsud’s case. Indeed, given Mehsud’s background and the context of the insurgency in AfPak, it would have been unreasonable to have doubted the legality of the attack.

A second persistent issue that has arisen is criticism that the United States has been insufficiently forthcoming and transparent about its drone attacks. In its “Conclusions and Recommendations” section, Philip Alston’s Report of the Special Rapporteur on Extrajudicial, Summary or Arbitrary Execution (Alston Report) stresses the need for States that engage in targeted killings to publicly reveal the rules of law they believe justify each killing, the bases upon which they use deadly force rather than attempt capture, the nature of procedural safeguards that ensure compliance with international law and procedures for post facto assessment and remedial measures.166 This concern evokes United States Supreme Court Justice Louis D. Brandeis’ famous quip that “[s]unlight is said to be the best of disinfectants; electric light the most efficient policeman.”167

These interrelated concerns of transparency may or may not be understood as a matter of moral, ethical or policy preference, but they are not required from an international humanitarian law perspective.168 Certainly, American drone attacks must comply with the “cardinal principles contained in the texts constituting the fabric of humanitarian law,” and the United States must perform its treaty obligations in good faith169 and interpret these obligations “in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose.”170 Apart from these broad legal  obligations, the United States, as with all other States that engage in methods and means of warfare, is not required by international humanitarian law to reveal its tactical “playbook.” The United States need not reveal its understanding of civilians who directly participate in hostilities, though it must, of course, interpret this concept in good faith. Indeed, the Alston Report itself admits that the weight of State practice supports this understanding: “[because] there is no commonly accepted definition of DPH [direct participation in hostilities], it has been left open to States’ own interpretation – which States have preferred not to make public – to determine what constitutes DPH.”171 The point is not that “much about the [Obama] Administration’s position remains unclear”172 anymore than it is to bemoan vague treaty language or vague provisions in United Nations Security Council Resolutions. The United States’ position is strategically vague but legally defensible, and international humanitarian law does not require more than this.

A final persistent issue that has arisen in the context of the American drone campaign in northwest Pakistan is the legal implications of CIA involvement. If, as this article maintains, the situation of violence that the United States is engaged in with the Taliban and its Al Qaeda allies in northwest Pakistan amounts to an armed conflict, then one must grapple with the so-called “combatant’s privilege,” that is, the “right to participate directly in hostilities.”173 As the Inter-American Commission on Human Rights put it in its 2002 Report on Terrorism and Human Rights, this is, “in essence[,] a license to kill or wound enemy combatants and destroy other enemy military objectives.”174 Only certain organizations and persons have this right by their very nature; all other organizations and persons do not. As a civilian intelligence agency, the CIA falls into the latter category. It is not part of the American armed forces, and it has not been incorporated into it.175


Read in Boston University International Law Journal


Afghanistan Al Qaeda Pakistan