Armed Conflict Cybersecurity & Tech Foreign Relations & International Law Terrorism & Extremism

The Unlawful U.S. Killing of Ayman al-Zawahiri

Ben Saul
Wednesday, August 17, 2022, 10:01 AM

The U.S. killing of the al-Qaeda leader in Afghanistan was not justified in self-defense or under the international law of war or international human rights law. It looks more like an extrajudicial execution, or revenge murder, for past acts of terrorism.

Ayman al-Zawahiri (Волк27,; CC BY-SA 4.0,

Published by The Lawfare Institute
in Cooperation With

The recent killing of the most senior al-Qaeda leader in Afghanistan is another triumph of American exceptionalism over international law. U.S. President Joe Biden boasted that “justice” has been “delivered,” but he offered no explicit legal justification for the death of Ayman al-Zawahiri. Certain assumptions about its justification may be drawn from the U.S.’s previous expansive legal positions on counterterrorism, discussed below and in a previous Lawfare piece on the legality of the al-Zawahiri strike. However, the extensive and sympathetic western media coverage has largely omitted questions about its legality, illustrating how effectively the U.S. shapes the narrative of the war on terror. Almost a year after the U.S. withdrawal from Afghanistan, the legality of the killing has implications for the U.S.’s militarized “over the horizon” strategy for combating terrorism in Afghanistan.

The killing is most accurately described as extrajudicial execution or revenge murder designed to deter others from participating in terrorist groups. It is also another body blow to the “rules-based international order” that the U.S. demands others—but apparently not itself—respect. There are three reasons why the killing violates international law. First, it is not a lawful exercise of self-defense against any continuing armed attack by al-Qaeda, but a violation of Afghanistan’s sovereignty and the prohibitions on the use of force and intervention in a foreign state. Second, it is not authorized by the law of war, since there is no longer an armed conflict against al-Qaeda to which the law of war applies, and al-Zawahiri may not have been a military target in any case. Finally, it violates the right to life under international human rights law—an obligation that applies to a U.S. strike abroad of this kind.


First, the U.S. does not presently have a right of national self-defense against al-Qaeda under Article 51 of the U.N. Charter or customary international law. Military force in self-defense is permitted only when a country is a victim of an “armed attack.” The attacks of 9/11 were treated by most countries, and the U.N. Security Council, as an “armed attack” on the U.S. This was novel at the time, since the International Court of Justice (ICJ) has long maintained, since Nicaragua v. United States (1986), that self-defense is available only if a non-state armed group is “sent” by a government, not where a group independently attacks. The Taliban, then the de facto government of Afghanistan, harbored al-Qaeda but did not “send” it to attack the U.S.

The 9/11 attacks more readily satisfied the ICJ’s traditional requirement that violence by a non-state group (sent by a state) must be of a sufficient gravity, or scale or effects, to approach what regular state armed forces could carry out. Further, a forcible U.S. response was viewed as “necessary”—another requirement of self-defense—because the attack was a continuing one, given al-Qaeda’s declared hostile intent and capabilities. These were also evident in al-Qaeda’s earlier attacks on the USS Cole in Yemen in 2000 and the bombing of U.S. embassies in Kenya and Tanzania in 1998.

However, two decades after that attack, it is no longer plausible to argue that al-Qaeda is still mounting a continuing armed attack against the U.S. that could justify killing al-Zawahiri in self-defense. The U.S. rapidly degraded al-Qaeda after it invaded Afghanistan in late 2001, despite occasional, persistent al-Qaeda plots. At some point, probably fairly early after the U.S. toppled the Taliban, the U.S. successfully halted and repelled al-Qaeda’s attack. Indeed, most U.S. hostilities in Afghanistan since 9/11 involved the Taliban, not al-Qaeda. In July 2022, the U.N.’s Resolution 1267 sanctions monitoring team concluded that “Al-Qa[e]da is not viewed as posing an immediate international threat from its safe haven in Afghanistan because it lacks an external operational capability and does not currently wish to cause the Taliban international difficulty or embarrassment.”

The U.S. has not publicly provided evidence contradicting that assessment. Indeed, the U.S. intelligence assessment from February 2022 indicated al-Qaeda senior leaders “lack an operational presence in Afghanistan” and that threats to U.S. interests will most likely occur where al-Qaeda’s affiliates operate—such as Yemen, Somalia, and West Africa—rather than against the U.S. “homeland.” The latest U.S. assessment of August 2022, after al-Zawahiri’s killing, concludes that al-Qaeda “does not have a capability to launch attacks against the US or its interests abroad from Afghanistan.”

Admittedly, determining the endpoint of an armed attack by a terrorist group is more difficult than in the case of a state aggressor, where the end of hostilities is usually demarcated by decisive victory on the battlefield, an unconditional surrender, or a negotiated peace agreement. In contrast, terrorist campaigns are by nature more sporadic, and groups that have been heavily suppressed can resurge. Where a group has already committed an armed attack, a presumption arguably exists in favor of characterizing its continuing violence as furthering that original attack.

Nonetheless, the basic rule holds that self-defense must end when there is no longer a continuing armed attack: that is, when al-Qaeda is no longer committing armed violence against the U.S. of a gravity loosely equivalent to that of regular state armed forces. Al-Qaeda attacks on the U.S.—actual, attempted, or planned—have been very rare since 9/11, and the last one was well over a decade ago in 2009: a thwarted bombing of the New York subway, itself more crime than a wartime attack. There have been more attacks by al-Qaeda and its affiliates against other countries, but these cannot be considered an armed attack on the U.S.

The practice of a few states since 9/11 may controversially suggest that the gravity threshold of an armed attack by a terrorist group has itself been lowered. For example, the U.K. declared a right of self-defense against ISIS in Syria following a thwarted plot to attack a Victory Day parade, and possibly kill the Queen, in London in 2015. France likewise claimed self-defense against ISIS in Syria after coordinated attacks by gunmen in Paris in 2015, which killed 130 people.

According to the ICJ jurisprudence mentioned earlier, these would have been regarded as not as of sufficient gravity to constitute a military armed attack, and would thus instead have been typically treated as terrorist crimes to be met by policing. Even in 2005, for example, the U.K. did not treat the large-scale al-Qaeda attack on London as an armed attack. The more recent examples by a handful of countries self-evidently do not reflect wider state practice and global opinion, and therefore cannot alter the long existing—albeit somewhat ambiguous—rules requiring a minimum gravity threshold.

Relatedly, there is increasing support for the idea that a series of smaller acts of terrorism could “add up” over a period of time to crystallize, at some tipping point, as meeting the gravity of an armed attack, even if each individual act is not of itself sufficiently grave. There is considerable merit in this so-called “accumulation of events” or “pin-prick” theory, and the ICJ has alluded to it in a number of cases. However, the pattern of acts must still be more than a string of isolated terrorist crimes and plausibly reach the high, military “armed attack” threshold. For the reasons given earlier, al-Qaeda’s current activities do not meet that threshold.

Self-defense should, of course, be interpreted dynamically in the light of evolving threats. But any lowering of the gravity threshold, or its attenuation via the accumulation of events theory (if applied too liberally), risks obliterating any distinction between an armed attack and lesser terrorist crimes. It prematurely escalates counterterrorism responses to kill first, rather than a last resort.

The right of self-defense has a strictly protective rationale. Where there is no longer an armed attack, there is no longer any legal justification for self-defense. Self-defense does not persist until every last leader responsible for a long-past attack is eliminated. It does not authorize punishment, vengeance, or retribution. It does not permit force to deter the possibility of future attacks by a degraded, remnant group that committed a concluded armed attack in the past. If it did, self-defense would be available in perpetuity against every vanquished state adversary who might rise again one day.

There is still the possibility that al-Qaeda may have been separately planning a new, “imminent” armed attack on the U.S., which could give the U.S. a right of self-defense. However, the U.S. has made no such claim, for instance by reporting the strike as self-defense to the Security Council as required by the U.N. Charter, and presented no evidence of any specific threat of this magnitude. Even if al-Qaeda was planning terrorist acts, unless these were of the gravity of an armed attack, they would simply be crimes, not acts of war, and would have to be met by a law enforcement not a military response. Additionally, before 9/11, global opinion was against anticipatory self-defense against imminent attacks, and this does not appear to have changed since 9/11, despite a little more support for it.

For argument’s sake, if al-Qaeda were still committing an armed attack, or planning an imminent attack, any defensive action by the U.S. would also need to be legally “necessary” to repel it. Remarkably, U.S. officials have stated that al-Zawahiri, near the time of his death, was not involved in planning al-Qaeda operations—but instead provided “strategic direction” to and “urged attacks” by al-Qaeda affiliates. Depending on further specifics as to his role, his killing may not have been necessary to stop any attacks if he had no role in ordering, planning, or controlling them.

Finally, much discussion about self-defense against terrorism has been focused on whether, and if so how, post-9/11 practice has altered the traditional requirement that an attack by an armed group be attributable to a state. The U.S. view is well known that self-defense exists where a state is “unable or unwilling” to suppress a non-state attacker based in its territory. Since al-Qaeda is no longer committing violence equivalent to an armed attack, for present purposes the issue is moot.

Again, however, the U.S. position on self-defense remains an outlier minority view and has failed to persuade the international community to change the law. In terms of other state practice, few states have explicitly supported it; a few others may have implicitly endorsed it; some support self-defenses against terrorism without explaining under what circumstances; and many have rejected it (including the 120 states of the Non-Aligned Movement). Further, it cannot be presumed that states that remain silent have somehow acquiesced to a radical change in a peremptory international rule. The ICJ has also maintained its traditional test in cases since 9/11. Most states likely fear that the escalatory and sovereignty costs, short and long term, of widening self-defense far outweigh the perceived immediate benefits.

In the absence of credible legal justifications for the use of force, the U.S. strike violates the sovereignty of the state of Afghanistan (irrespective of whether the Taliban is recognized as its government), the international prohibition on intervention in Afghanistan’s internal affairs, and the prohibition on the use of military force against another country.

For completeness, it goes without saying that the Taliban’s violation of its (non-treaty) commitment under the 2020 Doha Agreement, which framed the U.S. withdrawal from Afghanistan, not to harbor terrorists does not give the U.S. any right to kill al-Zawahiri. There is no right to enforce treaties or remedy their breach by resorting to nondefensive military force under international law.

The Law of War Does Not Apply

Second, the U.S. had no separate authority under the law of war—a separate branch of international law—to target al-Zawahiri. This is the law that applied when the fighting started between the U.S. and al-Qaeda in Afghanistan. In late 2001, a non-international armed conflict (NIAC) began because the fighting was sufficiently intense and al-Qaeda was a sufficiently organized armed group, meeting the  threshold of Common Article Three of the four Geneva Conventions of 1949 (as authoritatively interpreted by the International Criminal Tribunal for the former Yugoslavia in the Tadic (1995) case).

The traditional test for the existence of a NIAC focuses on the territory of the state in which the sufficiently intense fighting takes place—here, Afghanistan. Applying Tadic, a NIAC ends when the criterion for existence—intense armed violence between the parties—ceases. The U.S. withdrawal from Afghanistan in August 2021 ended the hostilities with al-Qaeda that constituted the NIAC there. It is thus irrelevant that there is no peace agreement with al-Qaeda, or no victory by either side, or that al-Qaeda maintains a hostile intent against the U.S. where it is not practically engaged in combat with the U.S. It is even arguable that the NIAC against al-Qaeda ended before the U.S. withdrawal, given that U.S. operations there were largely focused on the Taliban, not al-Qaeda.

The U.S. has asserted that its conflict against al-Qaeda is a de-territorialized global one, where the law of war follows the actions of the parties, so to speak. Al-Qaeda fighters outside Afghanistan can thus be targeted under the law of war wherever they are, as in Pakistan, Somalia or Yemen, or now in post-withdrawal Afghanistan, from the U.S.’s perspective. There is some logic to this, much as the law of war undoubtedly applies wherever there is fighting between states in an international armed conflict. Otherwise, insurgents could find safe havens abroad to carry on a fight with impunity. This approach aggregates all al-Qaeda threats globally to constitute the intensity of a single NIAC.

However, again, this U.S. position is a radical, minority one. Unsurprisingly, there is scant support from other states to recognize a “global war on terror,” in which U.S. military strikes (or those of Russia or China or others) could infringe on their sovereignty at-will. The dominant view is that NIACs are assessed on the territory of the state in which the fighting is sufficiently intense and the non-state party is sufficiently organized to reach those legal thresholds. In large part this is because, by definition, NIACs are not conflicts between states, but within them.

The International Committee of the Red Cross (ICRC) modestly acknowledges that a NIAC in one territory could “spill over” into a geographically adjacent territory and still be part of that NIAC, governed by the law of war. For example, the U.S. could lawfully strike al-Qaeda targets sheltering across the border in Pakistan, under the ICRC’s view.

Ultimately, however, the ICRC also maintains that violence farther afield must be assessed in the usual way: by asking whether the violence between, say, al-Qaeda and the U.S. in some other country is sufficiently intense, and the non-state party is sufficiently organized. On this approach, sporadic U.S. strikes on al-Qaeda operatives in other countries, where al-Qaeda usually cannot strike back at the U.S., would typically fail to meet the intensity threshold to constitute a conflict in that country.

However, where U.S. strikes are with the consent of and in support of a local government, itself already engaged in a NIAC in its own country, the U.S. could become a party to that conflict—even if its own hostilities against the group would not of themselves be sufficiently intense to establish a separate NIAC. An example is U.S. participation in Somalia’s NIAC with al-Shabab, which was previously allied with al-Qaeda. But that would still not give the U.S. any right to target al-Qaeda members outside the theater of hostilities in Somalia.

There is another convincing reason to doubt the legal existence of a global NIAC against al-Qaeda. Al-Qaeda is simply not a singular organized armed group fighting the U.S. worldwide. The various al-Qaeda “franchises” in different countries may have ideological and other links to the al-Qaeda leadership in Afghanistan and Pakistan, but in practice, they operate as independent groups. As one commentator succinctly put it, “In the real world … al-Zawahiri stopped being relevant years ago,” losing any authority at all over a group like Jabhat al-Nusra (now Hay’at Tahrir al-Sham) in Syria.

The application of the law of war thus depends on whether, for example, al-Qaeda in the Maghreb is an “organized armed group” fighting the U.S. in that region. The question is not whether their nominal allegiance to al-Zawahiri makes al-Qaeda everywhere a party to some imaginary, unified global conflict, or artificially attenuated global “co-belligerents” with al-Qaeda against the U.S.

Targeting al-Zawahiri Under the Law of War

In sum, since the law of war no longer applies, the U.S. has no basis in it to target members of al-Qaeda in Afghanistan. Even if the law of war still applied, however, al-Zawahiri may not have been targetable, although this depends on one’s approach to the law on targeting in NIACs, as well as more precise facts about the role of al-Zawahiri in al-Qaeda operations.

The U.S. view, shared by states such as Australia, is that any member of an organized armed group is targetable, irrespective of whether they perform a combat function. Thus, even if al-Zawahiri was no longer commanding al-Qaeda operations, as the senior political and spiritual leader of the group, he was targetable. This approach is comparable to “targeting” in international conflicts, where state military personnel can be targeted as “combatants” even if they are unarmed cooks.

In contrast, the ICRC takes a narrower view. There is no “combatant” status, or de facto combatant status, in NIAC that automatically allows members of non-state armed groups to be targeted. Rather, a person may be targeted for such time as they take “a direct part in hostilities,” whether as a civilian sporadically engaging in hostilities or as a member of an armed group.

The ICRC deems the latter to be targetable only where the person performs a “continuous combat function,” meaning they are integrated on a lasting basis into the armed group and have a combat-related role in it (whether as a fighter, as a commander, or providing direct support to fighters, such as transporting munitions). This allows for such persons to be targeted at any time as long as they remain part of the group, even if they are not engaged in combat at the time.

This approach is narrower than the U.S. view because it does not permit the targeting of group members who are not engaged in hostilities—such as spiritual, ideological, or political leaders; financiers; propagandists; recruiters; arms dealers; bomb makers; and so on. Lest the ICRC view seem too restrictive: The international law of war likewise does not allow an adversary to kill, say, members of Congress, U.S. taxpayers who finance the U.S. military, or civilian workers in U.S. weapons factories.

The U.S. view risks collapsing the distinction between unarmed civilians, who present no direct military threat warranting a forcible response, and fighters who are militarily dangerous and whose killing is necessary. Terrorist groups are not analogous to state militaries, whose personnel are separate from civilians, because terror groups are hybrids of civilian and military personnel. Thus, whether targeting al-Zawahiri would be lawful depends on more information about his exact role. As mentioned, he was reportedly not involved in planning al-Qaeda operations, but in “urging attacks” and providing “strategic direction.” If he was not involved in planning attacks, which normally includes ordering them, then his role in “urging attacks” does not necessarily mean that he was a targetable military commander, as opposed to someone playing an indirect role by inciting violence, which al-Qaeda members may or may not have acted upon. Likewise, if “strategic direction” means political and spiritual guidance, not military strategy, or propaganda or recruitment activities, al-Zawahiri would not have been taking a direct part in hostilities and thus would not be targetable.

It should be emphasized that even if al-Zawahiri’s killing were lawful under the law of war, this would not render it a lawful exercise of self-defense under the law on the use of force, as discussed earlier. The U.S. would still violate its obligation not to use non-defensive military force.

International Human Rights Law

It is arguable that the U.S. strike also substantively violated al-Zawahiri’s human right to life under the International Covenant on Civil and Political Rights (ICCPR). In peacetime, the ICCPR permits the use of force only in self-defense or defense of others against an imminent lethal threat. Al-Zawahiri, standing on a balcony in Kabul, posed no such threat. Since the killing was not justified in national self-defense or under the law of war, those areas of law do not apply as lex specialis to modify this peacetime, law enforcement-based right-to-life standard.

The issue depends on the threshold issue of whether the U.S.’s international human rights obligations applied in Afghanistan. The U.S. has long maintained that its human rights obligations do not apply outside its territory, a view decisively rejected by the ICJ, the U.N. human rights bodies that authoritatively interpret the International Covenant on Civil and Political Rights, and regional bodies such as the Inter-American human rights system and the European Court of Human Rights.

The more controversial question is whether an isolated military strike such as this one legally constitutes an exercise of jurisdiction by the U.S., so as to trigger its human rights obligations. It is well settled that human rights apply extraterritorially where a state effectively controls foreign territory (as during occupation, or under leases like at Guantanámo Bay), or has physical custody of a person.

“Hot-combat” situations, including hostilities to contest territory, have been viewed as falling outside a state’s jurisdiction, as in the European Court of Human Rights cases of Bankovic (2001) and Georgia v. Russia (No. 2) (2021). However, in Carter v. Russia (2021), the European Court accepted that Russian agents poisoning the defector Aleksandr Litvinenko in the U.K. “amounted to the exercise of physical power and control over his life in a situation of proximate targeting.” The similarly targeted nature of the al-Zawahiri strike, being outside of any hot combat in Afghanistan, would likewise arguably ground U.S. jurisdiction.


The U.S. finds itself in an awkward club of states that murder or assassinate abroad, including Russia, Saudi Arabia, and Israel. That the killing may be lawful under the U.S.’s absurdly wide 2001 Authorization for Use of Military Force, which purports authorizing perpetual global war against al-Qaeda irrespective of international law constraints, does not excuse U.S. violations of international law. In the same way, Russia cannot write its own rules for Ukraine, or China for the South China Sea, or Israel for occupied foreign territory, or Morocco for annexed Western Saharan territory. U.S. law does not rule the world. 

Terrorists evading justice in safe havens like Afghanistan is a fiendish policy challenge, despite the raft of nonmilitary tools available to contain them, including intelligence and law enforcement cooperation, sanctions, collective action through the Security Council, diplomacy, and influencing their state sponsors through incentives, among others. Unilateral, extralegal solutions may appeal to powerful states that can pursue them with relative impunity, particularly where other states—often not victims of terrorism to the same extent—cannot be convinced to change the law.

Ultimately, however, summarily executing criminal suspects without trial, however unsavory they may be, is disturbingly lawless and unjustifiable. Vigilante justice discredits the U.S. and erodes its moral authority as a self-declared champion of a “rules-based international order.” It signals to other states that the rule book is only for the weak, and spawns lawlessness by others.

It is also counterproductive, evident in the deep radicalizing effects of other U.S. war on terror practices such as torture, enforced disappearances, drone strikes causing excessive civilian casualties, unfair military commissions, and indefinite detention at Guantánamo Bay. It tends to bring tactical victories without strategic success. The relative failure of two decades of counterterrorism in Afghanistan warrants more abundant caution about whether military responses to terrorism are necessarily the most effective long-term means of countering the threat.

Ben Saul is Challis Chair of International Law at The University of Sydney, an Associate Fellow of Chatham House in London and the International Centre for Counter-terrorism in The Hague, and a UN counter-terrorism adviser. He recently co-drafted the United Nations Model Law on Victims of Terrorism. He has taught at Harvard and Oxford.

Subscribe to Lawfare