When a state kills a designated terrorist on foreign soil, it simultaneously invokes the right of self-defense under Article 51 of the UN Charter, violates the sovereignty protections enshrined in Article 2(4), and either complies with or breaches international humanitarian law depending on whether the target qualifies as a combatant. The same act, the same bullet, the same body on the same sidewalk in Karachi or Sana’a or Gaza City, is legal or illegal depending entirely on which legal framework the analyst chooses to apply. This is not a gap in international law. It is the central structural feature of the targeted killing debate, and every state that conducts lethal operations abroad understands it, exploits it, and profits from the ambiguity that no government has any incentive to resolve.

The question of whether states possess the right to kill individuals on foreign territory has animated legal scholars, intelligence agencies, diplomats, and human rights advocates for decades. Israel’s Supreme Court attempted to answer it in 2006 with a landmark ruling that conditionally permitted targeted killings under strict proportionality requirements. The United States answered it through classified Office of Legal Counsel memoranda that invoked self-defense and imminent threat doctrines to authorize drone strikes across multiple continents. Russia answered it by poisoning defectors in London parks and denying involvement. India answered it by conducting a systematic campaign of eliminations across Pakistani cities while maintaining total deniability. The United Kingdom answered it by prohibiting targeted killing in domestic law while sharing intelligence with allies who conduct them. Five countries, five frameworks, five different relationships between law and lethal force. Each framework reveals something about how states construct legal cover for political violence. Together, they reveal that the legal debate on targeted killing is not a debate between clarity and confusion. It is a debate between competing forms of deliberate ambiguity, each calibrated to the domestic political needs and international positioning of the state conducting operations.

Paul Staniland of the University of Chicago has written extensively on how states construct legal justifications for political violence, arguing that the frameworks matter less for what they permit than for what they signal about the state’s relationship to international norms. A state that builds a legal framework, as Israel and the United States have done, signals that it recognizes the normative constraints on lethal force even as it carves exceptions for its own operations. A state that operates in total deniability, as India does, signals that it considers the normative constraints irrelevant to its security calculus. George Perkovich of the Carnegie Endowment has raised the question of whether India’s total-deniability approach creates a dangerous legal vacuum, arguing that the absence of any framework, domestic or international, means there is no mechanism for accountability, no standard for proportionality, and no process for determining whether a particular killing was justified or merely convenient. The five-country comparison that follows tests both positions against the operational record.

The analytical question at the core of this comparison is not whether targeted killing is right or wrong. That question admits of no universal answer because the circumstances, the targets, the methods, and the alternatives vary so dramatically across cases that moral judgment requires case-by-case analysis. The analytical question is structural: why do some states build legal frameworks for targeted killing while others operate in legal vacuums, and what consequences follow from each choice? The answer reveals a pattern that neither pure legalism nor pure realism captures adequately. Legal frameworks serve operational and political functions beyond their formal legal content. They constrain behavior at the margins, they provide institutional processes that reduce the risk of catastrophic error, and they generate the appearance of legitimacy that insulates programs from political opposition. States that operate without frameworks avoid these constraints but also forfeit the protections that frameworks provide: the institutional processes that catch mistakes before they occur, the accountability mechanisms that limit blowback when mistakes are exposed, and the legitimacy that sustains programs through changes of government and shifts in public opinion. The choice between framework and vacuum is not a choice between constraint and freedom. It is a choice between different risk profiles, and the five-country comparison reveals which risks each model generates.

The debate also implicates a fundamental tension in the post-1945 international order. The UN Charter system rests on two principles that are mutually reinforcing in theory and contradictory in practice: state sovereignty (the principle that each state controls its territory exclusively) and the prohibition on the use of force (the principle that states may not use force against each other except in self-defense or with Security Council authorization). When a state harbors non-state actors who conduct attacks against another state, and the harbor state refuses to extradite the attackers or dismantle their infrastructure, the victim state faces a choice between accepting continued attacks (respecting sovereignty) and using force on foreign soil to eliminate the threat (exercising self-defense). International law provides no clean resolution to this dilemma, and the targeted killing debate is the arena in which the tension plays out most acutely. Every state in this comparison has resolved the tension in its own way, and every resolution tells a story about power, norms, and the limits of law in governing violence between states.

The Cases

Five states conduct or have conducted targeted killings on foreign soil. Each has developed a distinctive relationship to the legal questions that such operations raise. Understanding the legal debate requires understanding each country’s framework not as a theoretical exercise but as an operational reality, a set of rules that either constrain behavior, justify behavior, or provide rhetorical cover for behavior that would proceed regardless.

Israel formalized its approach through judicial review. The Israeli Supreme Court’s December 2006 ruling in the Public Committee Against Torture v. Government of Israel case examined the legality of targeted killings conducted by the Israel Defense Forces against Palestinian militants in the occupied territories. The court did not ban targeted killings. Nor did it grant blanket authorization. It held that targeted killings could be lawful under international humanitarian law if four conditions were met: the target must be a person who directly participates in hostilities, the killing must be a last resort after arrest has been attempted or deemed impossible, the operation must satisfy proportionality by not causing excessive civilian casualties relative to the military advantage gained, and an independent investigation must follow each operation to verify compliance. The ruling created the most detailed judicial framework for targeted killing anywhere in the world, and Israel has cited it consistently in defending its operations against Hamas, Hezbollah, and Iranian nuclear scientists. The framework is real. Whether it constrains behavior in practice is a separate question that the comparison must address.

The United States built its framework through executive legal authority rather than judicial review. The Office of Legal Counsel within the Department of Justice produced a series of classified memoranda beginning in 2001 that authorized the CIA and the Department of Defense to conduct lethal operations against members of al-Qaeda and associated forces anywhere in the world. The legal theory rested on three pillars: the Authorization for Use of Military Force passed by Congress on September 18, 2001, which the executive branch interpreted as authorizing lethal force against al-Qaeda globally; the inherent right of self-defense under Article 51 of the UN Charter, which the administration interpreted as permitting preemptive action against imminent threats; and the law of armed conflict, which the administration argued applied to the “global war on terror” as an armed conflict of international scope. The most controversial element was the “imminent threat” standard. Traditional international law required imminence to mean an attack was about to occur. The OLC memos redefined imminence to include situations where a senior operational leader of al-Qaeda posed a continuing threat, even if no specific attack was known to be planned. This redefinition expanded the legal authorization for targeted killing beyond anything previously recognized in international law and provoked sustained criticism from legal scholars who argued it rendered the imminence requirement meaningless.

Russia operates without any legal framework for targeted killing. The assassinations of Alexander Litvinenko in London in 2006, the attempted poisoning of Sergei Skripal in Salisbury in 2018, and the killings of Chechen dissidents across Europe and Turkey follow no published legal doctrine, invoke no right of self-defense, and claim no authorization under international humanitarian law. Russia’s approach is pure deniability. When confronted with evidence of state involvement, Russian officials deny responsibility, propose alternative explanations, and refuse cooperation with international investigations. The Salisbury poisoning, which used the Novichok nerve agent and affected multiple British civilians in addition to the intended targets, demonstrated the operational consequences of operating without legal constraints: the absence of a proportionality requirement meant the operatives used a weapon that endangered an entire community. The Russian model proves that targeted killing without legal framework does not mean targeted killing without consequences. The diplomatic fallout from Salisbury, including the largest coordinated expulsion of Russian diplomats in history, showed that the international community applies its own accountability standards when the perpetrating state refuses to apply any.

India’s approach to the targeted killing question represents something distinct from any of the other four cases. India does not acknowledge conducting targeted killings at all. The systematic elimination of India’s most-wanted terrorists on Pakistani soil, documented in the comprehensive overview of the campaign and analyzed across dozens of individual cases, proceeds under total deniability. The Indian Ministry of External Affairs has never confirmed involvement in any killing. No Indian court has examined the legality of the operations. No parliamentary committee has reviewed the targeting criteria. No domestic statute authorizes or prohibits the activity. India occupies the most extreme position on the legal spectrum: not a permissive framework (like the United States), not a conditional framework (like Israel), not a deniability-with-evidence framework (like Russia), but a complete legal void. The operations exist in the operational record but not in the legal record. Perkovich’s concern about the dangerous legal vacuum applies most directly to India, where the absence of any framework means that every element of the targeted killing debate, from target selection criteria to proportionality assessment to post-operation review, occurs entirely within the intelligence establishment with no external oversight of any kind.

The United Kingdom occupies the final position in this comparison: the state that prohibits targeted killing in domestic law while enabling it through intelligence cooperation. British law does not authorize lethal operations by intelligence agencies. MI6 operates under the Intelligence Services Act 1994, which permits activities abroad that would otherwise be illegal under UK law if authorized by the Foreign Secretary, but the authorization framework has never been publicly interpreted to cover assassination. The Investigatory Powers Tribunal and the Intelligence and Security Committee of Parliament provide oversight, but their inquiries into specific operations are classified. At the same time, the United Kingdom shares intelligence with the United States, Israel, and other allies who conduct targeted killings, and British intelligence has been cited in connection with drone strikes that killed British citizens abroad. The UK framework creates a form of legal compartmentalization: the killing itself occurs outside British jurisdiction and British legal authority, but the intelligence that enables the killing originates within the British system. Whether intelligence sharing that contributes to a targeted killing makes the sharing state legally complicit is one of the unresolved questions in international law, and the United Kingdom’s ambiguous position keeps it unresolved.

The Sovereignty Barrier

Article 2(4) of the United Nations Charter prohibits the use of force against the territorial integrity or political independence of any state. The prohibition is absolute in text and qualified in practice. Every targeted killing conducted on foreign soil without the host state’s consent violates Article 2(4) on its face. The sovereignty barrier is the foundational legal obstacle that every state conducting targeted killings must either overcome, circumvent, or ignore.

Israel has navigated the sovereignty barrier by operating primarily in territories it controls or in states with which it is in a de facto state of conflict. The occupied Palestinian territories, where the majority of Israeli targeted killings have occurred, present a complex sovereignty question because Israel disputes the applicability of the Fourth Geneva Convention while simultaneously exercising effective control. Operations against Hamas and Hezbollah leaders in Gaza, Lebanon, Syria, and Iran involve clear sovereignty violations, but Israel argues that these states either harbor hostile non-state actors or are themselves party to the hostilities. The Israeli Supreme Court’s 2006 ruling sidestepped the sovereignty question almost entirely, focusing instead on the humanitarian-law framework governing the conduct of hostilities rather than the jus ad bellum question of whether the hostilities were lawful in the first place. This analytical move was deliberate. By framing targeted killing as a question of how force is used rather than whether force is authorized, the court avoided the sovereignty debate that would have forced it to address Israel’s legal basis for using force on Palestinian territory.

The United States confronted the sovereignty barrier more directly because its drone operations occurred in states that were nominally at peace with Washington, including Pakistan, Yemen, and Somalia. The legal response was the “unwilling or unable” doctrine: the argument that when a host state is unwilling or unable to address a terrorist threat emanating from its territory, the threatened state retains the right to use force on that territory in self-defense. The doctrine first appeared in a 2001 letter from the United States to the United Nations Security Council following the September 11 attacks, invoking Article 51 and identifying Afghanistan as the state harboring al-Qaeda. Over the following decade, the doctrine expanded to cover operations in Pakistan (where the government privately consented to drone strikes while publicly condemning them), Yemen (where the government sometimes consented and sometimes objected depending on the political situation), and Somalia (where effective government authority was largely absent). The “unwilling or unable” test remains the most significant challenge to the sovereignty barrier in contemporary international law, and its acceptance or rejection determines the legality of most targeted killing operations. Daniel Byman of the Brookings Institution has argued that the doctrine reflects operational reality: states that harbor terrorists and refuse to act against them cannot invoke sovereignty as a shield. Critics respond that accepting the doctrine would allow any state to claim that another state is “unwilling or unable” and use that claim to justify military operations on foreign soil, effectively gutting the sovereignty protections that underpin the entire UN system.

Russia bypasses the sovereignty barrier entirely. The Litvinenko poisoning occurred on British soil. The Skripal attack occurred on British soil. The killing of Zelimkhan Khangoshvili in Berlin’s Tiergarten park in August 2019 occurred on German soil. In none of these cases did Russia seek consent, invoke self-defense, or argue that the host state was unwilling or unable to address a threat. Russia’s position is that these events did not occur as alleged, that the individuals were killed by other parties, or that the accusations are politically motivated fabrications. The sovereignty violation is total and unacknowledged. The international response has been diplomatic rather than legal: expulsions, sanctions, and condemnation, but no prosecution in international courts and no formal finding of state responsibility by any international tribunal. The Russian precedent demonstrates that the sovereignty barrier, while legally absolute, is diplomatically negotiable. A state with sufficient power, particularly a permanent member of the Security Council with veto authority, can violate sovereignty repeatedly without facing legal consequences, only political ones.

India’s relationship to the sovereignty barrier mirrors Russia’s in one respect and differs from it in another. Like Russia, India operates without consent and without legal justification. Unlike Russia, India does not leave forensic evidence of state involvement. The motorcycle-borne killings documented across Karachi, Rawalpindi, Sialkot, and other Pakistani cities do not involve nerve agents or radioactive materials traceable to state programs. They involve handguns, motorcycles, and operatives who disappear into traffic. Pakistan alleges Indian involvement and has presented its allegations at diplomatic forums, but the evidentiary chain connecting the killings to Indian intelligence agencies remains circumstantial. The sovereignty violation, if Indian involvement is established, is identical in legal character to Russia’s violation of British sovereignty in the Skripal case: the unconsented use of lethal force on another state’s territory. The practical difference is that India’s method of operation makes the violation harder to prove and therefore harder to prosecute. The shadow war’s operational methodology is designed to exploit this evidentiary gap, and the legal debate cannot ignore the reality that deniability is itself a legal strategy, one that makes the sovereignty barrier functionally irrelevant by ensuring that no state or institution can attribute the violation with sufficient certainty to trigger a formal response.

The sovereignty question acquires additional complexity when the host state provides tacit consent. Pakistan’s relationship with the US drone program illustrates this complexity. Between 2004 and 2018, the United States conducted hundreds of drone strikes in Pakistan’s Federally Administered Tribal Areas. Pakistan publicly condemned every strike as a violation of sovereignty. Privately, Pakistani military and intelligence officials provided targeting intelligence, cleared airspace, and allowed CIA operations from Shamsi airfield in Balochistan until 2011. The legal status of these operations depends on which Pakistani position is treated as authoritative: if the private consent governs, the strikes did not violate sovereignty because consent was given; if the public condemnation governs, every strike was unlawful. The dual-track approach served both governments: the United States gained operational access while Pakistan maintained plausible deniability before its domestic audience, which overwhelmingly opposed the strikes. The sovereignty question was not resolved; it was managed through diplomatic performance. India’s operations in Pakistan exist in a different register because there is no evidence of Pakistani consent, tacit or otherwise. Pakistan’s objections are genuine, and the sovereignty violation, if Indian involvement is established, is unambiguous in a way that the US-Pakistan arrangement was not.

The sovereignty barrier also raises the question of proportionality at the state level, distinct from the operational proportionality question within international humanitarian law. When a state violates another state’s sovereignty to conduct a targeted killing, the violation must be proportionate to the threat that justifies it. A single killing of a senior terrorist commander who has planned mass-casualty attacks might be defensible as a proportionate sovereignty violation. A campaign of dozens of killings across multiple cities over several years represents a sustained and systematic violation of sovereignty that raises different legal questions. The scale of India’s alleged operations, documented across Karachi, Rawalpindi, Sialkot, Lahore, and other locations, suggests not a one-time exception but a permanent redefinition of the India-Pakistan sovereignty relationship. Pakistan’s sovereignty is being treated as penetrable by design, not as a barrier to be overcome in exceptional circumstances. The legal framework does not address this distinction between episodic and systematic sovereignty violation, but the practical and political consequences are vastly different.

The United Kingdom confronts the sovereignty barrier obliquely. British intelligence does not conduct killings on foreign soil, but British intelligence products contribute to killings conducted by allies. When a GCHQ intercept provides the geolocation data that enables a US drone strike in Yemen, and that strike kills a British citizen, the sovereignty question bifurcates. The United States violated Yemeni sovereignty (assuming no consent). Did the United Kingdom share legal responsibility? The issue arose publicly in connection with the killing of Reyaad Khan, a British citizen killed by a Royal Air Force drone strike in Syria in September 2015. The strike was conducted under UK authority, making it a direct British use of force on Syrian territory, but the legal justification invoked self-defense under Article 51 rather than addressing the sovereignty question directly. The Intelligence and Security Committee’s subsequent review of the targeting process was conducted in classified session, and its conclusions remain unavailable. The UK position on sovereignty remains deliberately unclear, and the ambiguity serves British interests by allowing continued intelligence cooperation with states that conduct targeted killings without requiring the United Kingdom to take a legal position on whether that cooperation constitutes complicity.

The Self-Defense Justification

Article 51 of the UN Charter preserves the inherent right of self-defense in the event of an armed attack. The article was drafted to address attacks by states against states, not attacks by non-state actors operating from third-party territory. The gap between the text’s original scope and its contemporary application is the central legal battleground in the targeted killing debate.

Israel has invoked self-defense consistently since the founding of the state, but the application of Article 51 to operations against non-state actors predates the formal targeted killing program. Israel’s argument rests on the proposition that armed attacks by Hamas, Hezbollah, and other groups originating from territory outside Israeli control constitute armed attacks within the meaning of Article 51, triggering the right of self-defense regardless of whether the attacking entity is a state. The International Court of Justice partially addressed this question in its 2004 advisory opinion on the Legal Consequences of the Construction of a Wall in the Occupied Palestinian Territory, where the court stated that Article 51 addressed attacks by one state against another and did not apply to threats originating within territory under the defending state’s own control. The opinion drew sharp dissents from several judges who argued that the right of self-defense applies to attacks by non-state actors regardless of territorial origin, and Israel has consistently rejected the majority’s reasoning. The 2006 Supreme Court ruling on targeted killing proceeded on the assumption that Israel was engaged in an armed conflict with Palestinian armed groups, bypassing the Article 51 question by treating the conflict as already established rather than requiring a fresh self-defense justification for each operation.

The United States expanded the self-defense justification beyond anything Israel had claimed. The Bush administration’s initial invocation of Article 51 following September 11 received broad international support: the Security Council passed Resolution 1368 on September 12, 2001, recognizing the right of self-defense in the context of the terrorist attacks, and NATO invoked Article 5 of the North Atlantic Treaty for the first time in its history. The legal consensus fractured as the geographic scope of operations expanded. The authorization to use force in Afghanistan against al-Qaeda and the Taliban was widely accepted. The extension of that authorization to drone strikes in Pakistan, Yemen, Somalia, and Libya provoked sustained legal challenge. The core dispute concerned the temporal and geographic limits of self-defense: does the right triggered by the September 11 attacks persist indefinitely and apply globally, or does it expire when the immediate threat subsides and apply only in the theater of active hostilities? The Obama administration attempted to resolve this through the “continuing and imminent threat” standard articulated in a 2012 speech by Attorney General Eric Holder, who argued that a threat could be imminent even in the absence of evidence of a specific planned attack if the individual posed a continuing danger. Legal scholars including Nils Melzer, then the UN Special Rapporteur on extrajudicial, summary, or arbitrary executions, criticized the standard as eliminating the imminence requirement in practice while claiming to preserve it in theory.

The “unwilling or unable” doctrine functions as the operational bridge between the self-defense justification and the sovereignty barrier. If a state invokes self-defense under Article 51, it must explain why force is being directed at the territory of a state that did not conduct the armed attack. The answer, in the US and Israeli formulation, is that the host state is either unwilling to address the threat (it supports or tolerates the armed group) or unable to do so (it lacks effective control over the territory from which the threat emanates). The doctrine has not been formally adopted by any international tribunal. The International Court of Justice has not ruled on its validity. The International Law Commission has not codified it. Its status in customary international law depends on whether sufficient state practice and opinio juris exist to establish it as a binding norm. Proponents point to the US, Israeli, Turkish, French, and Australian invocations of the doctrine in counter-terrorism operations as evidence of emerging state practice. Opponents argue that the doctrine has been invoked exclusively by powerful states against weaker ones, and that accepting it would create an asymmetric legal regime in which strong states can override the sovereignty of weak states by claiming that the weak state is “unable” to address a threat.

The temporal dimension of self-defense adds another layer of complexity. Article 51 addresses the right of self-defense “if an armed attack occurs.” The present-tense construction has been interpreted by most scholars as requiring an attack to have occurred or to be genuinely imminent. The United States stretched the temporal dimension to its breaking point with the “continuing and imminent threat” standard, but other countries have faced the same challenge. Israel’s targeted killings of Hamas military leaders occur during periods of both active combat and relative calm. The Israeli position is that the armed conflict with Hamas is ongoing and continuous, making the self-defense justification available at all times regardless of whether a specific attack has recently occurred or is imminently planned. This “continuous armed conflict” interpretation eliminates the temporal limitation entirely for states engaged in open-ended conflicts with non-state actors, which is precisely the objection that scholars raise: if the conflict never ends, the right to kill never expires, and the temporal safeguard built into Article 51 becomes meaningless.

Turkey provides an instructive parallel. Turkish military forces have conducted repeated cross-border operations against Kurdistan Workers’ Party (PKK) positions in northern Iraq and northeastern Syria, invoking self-defense under Article 51 and arguing that Iraq and Syria are unwilling or unable to prevent PKK attacks originating from their territory. The Turkish invocations have received mixed international responses: the United States, a NATO ally, has generally acquiesced while criticizing specific operations; Iraq has protested violations of its sovereignty; Syria has rejected Turkish operations entirely. The Turkish case demonstrates that the “unwilling or unable” doctrine is applied selectively by the international community, accepted when invoked by allied states and questioned when invoked by adversaries. The legal standard is uniform; the political application is asymmetric.

India has never invoked Article 51 in connection with operations on Pakistani soil. India has invoked self-defense in connection with conventional military operations, including the 2016 surgical strikes across the Line of Control and Operation Sindoor, but these invocations addressed military operations conducted openly and acknowledged publicly. The shadow war operates in a different register entirely. Because India does not acknowledge the operations, it cannot invoke a legal justification for them. The absence of invocation does not mean the self-defense argument would be unavailable. India could, in principle, argue that Pakistan’s harboring of designated terrorists, its refusal to act on extradition requests, and its documented history of using non-state actors as instruments of state policy create precisely the “unwilling or unable” scenario that justifies defensive action. India chooses not to make this argument because making it would require acknowledging the operations, which would transform a deniable intelligence campaign into an acknowledged act of war. The legal justification is sacrificed to preserve operational deniability, and India’s silence on Article 51 is itself a strategic choice.

Russia’s relationship to self-defense is entirely cynical. Russia has invoked self-defense in connection with its military operations in Syria, arguing that it was acting at the invitation of the Assad government to combat terrorist threats. Russia has never invoked self-defense in connection with poisonings in London or Berlin, because doing so would require acknowledging the operations. The Russian approach demonstrates that the self-defense justification is optional for states that prefer deniability to legal legitimacy, and that the international system has no mechanism for forcing a state to provide a legal justification for an operation it denies conducting. The gap between what the legal framework requires (justification) and what the political system accepts (denial) is the space in which most targeted killings operate.

The United Kingdom invoked self-defense under Article 51 for the Reyaad Khan strike in Syria, arguing that Khan posed an imminent threat to British national security through his role in planning attacks against the United Kingdom from Islamic State territory. The invocation was notable because it was the first time the UK had used lethal force under Article 51 authority outside of a recognized armed conflict in which British forces were engaged as a belligerent. The legal basis was debated in Parliament, where the Prime Minister argued that the government’s authority to act in self-defense was inherent and did not require prior parliamentary approval. The argument established a precedent for UK use of force on foreign soil against non-state actors, but the precedent has not been tested since, and the legal community remains divided on whether the Khan case represents a one-time exception or a new baseline for British self-defense doctrine.

The Combatant Status Question

International humanitarian law, the body of rules governing the conduct of armed conflict, applies only when an armed conflict exists. If targeted killing occurs outside of armed conflict, the applicable legal framework is international human rights law, which prohibits extrajudicial killing without exception. The classification question, whether a targeted killing occurs within or outside of armed conflict, determines which legal regime applies and therefore whether the killing can ever be lawful.

Israel’s Supreme Court resolved this question for Israeli operations by holding that a state of armed conflict existed between Israel and Palestinian armed groups. The court classified the conflict as an “international armed conflict” of a specific character, one involving an organized armed group rather than a state, and applied the rules of international humanitarian law accordingly. Within that framework, the court developed the “direct participation in hostilities” test: individuals who directly participate in hostilities may be targeted during the period of their participation, but civilians who do not participate retain full legal protection. The critical interpretive question was what constitutes “direct participation.” A fighter firing a weapon clearly participates directly. A bomb-maker building an explosive device probably participates directly. A financier transferring funds to an armed group might participate directly, depending on how proximate the financial support is to the armed activity. A propagandist recruiting fighters might not participate directly at all. The Supreme Court left these line-drawing questions to be resolved case by case, which meant that Israeli military lawyers, not judges, would determine in real time whether a particular individual’s activities met the direct-participation threshold. The International Committee of the Red Cross published its own “Interpretive Guidance on the Notion of Direct Participation in Hostilities under International Humanitarian Law” in 2009, led by Nils Melzer, which proposed a more restrictive test: only individuals performing a “continuous combat function” within an organized armed group could be targeted at any time, while other civilians could only be targeted during the specific act of direct participation. Israel has not adopted the ICRC’s framework, preferring its own broader interpretation.

The United States classified the conflict with al-Qaeda as a “non-international armed conflict” of global scope, a classification that had no precedent in international humanitarian law. Traditional non-international armed conflicts occur within a single state’s territory, between government forces and organized armed groups. The US position that the conflict with al-Qaeda constituted a non-international armed conflict spanning multiple continents required accepting that armed conflict could exist in a geographically unbounded form, with hostilities occurring wherever members of the armed group were located. The classification had enormous practical consequences: by declaring the existence of an armed conflict, the United States could apply the law of armed conflict rather than human rights law, which meant that members of al-Qaeda could be targeted as combatants rather than arrested as criminal suspects. The OLC memos applied a “functional combatant” test: individuals who held operational leadership positions within al-Qaeda or who were actively engaged in planning attacks qualified as targetable combatants. The test was applied internally by executive-branch lawyers with no judicial review and no adversarial process. The targeting of American citizen Anwar al-Awlaki in a September 2011 drone strike in Yemen tested the limits of this framework. The OLC produced a separate memorandum addressing the question of whether the government could kill an American citizen without due process, concluding that the citizen’s constitutional rights yielded to the government’s interest in national security when the citizen posed an imminent threat and capture was infeasible. The memorandum was released in redacted form in 2014 following litigation by the American Civil Liberties Union.

Russia does not classify its targeted killings as occurring within any armed conflict. Litvinenko was a former intelligence officer living as a British citizen. Skripal was a retired double agent living under diplomatic protection. Khangoshvili was a Georgian citizen of Chechen origin living in Germany as an asylum seeker. None of these individuals was engaged in armed conflict. The killings, if attributed to the Russian state, fall entirely within the framework of international human rights law, which prohibits extrajudicial killing absolutely. The applicable legal regime is unambiguous: these killings, if proven to be state-sponsored, are unlawful under every applicable framework. Russia’s response is not legal argument but factual denial. The gap between law and practice is not bridged by legal interpretation; it is obscured by denial, disinformation, and the structural inability of international institutions to compel a permanent Security Council member to submit to jurisdiction. The Russian case is instructive for the broader debate because it reveals the floor of the legal system: when a state with sufficient power refuses to engage with the legal framework at all, the framework has no enforcement mechanism. The Security Council, which is the only UN body with binding enforcement authority, cannot act against a permanent member because the member can exercise its veto. The International Court of Justice requires consent to jurisdiction that Russia has not given for most matters. The European Court of Human Rights has issued rulings, but Russia formally withdrew from the Council of Europe in March 2022, removing itself from the court’s jurisdiction entirely. The Russian experience demonstrates that international law on targeted killing operates as a system of voluntary compliance, binding only on states that choose to be bound, and the states most likely to conduct targeted killings are precisely the states least likely to accept binding legal constraints.

India’s targeted killings in Pakistan present a classification challenge that neither the Israeli nor the American framework fully addresses. Is the conflict between India and Pakistan an armed conflict? The two states have fought four conventional wars (1947, 1965, 1971, 1999 Kargil). They exchanged fire during the 2019 Balakot crisis and the 2025 Operation Sindoor. But the targeted killings occur outside the periods of active hostilities, during what would conventionally be classified as peacetime. If the killings occur in peacetime, international human rights law applies, and extrajudicial killing is prohibited. If the killings occur within an ongoing armed conflict, international humanitarian law applies, and the targets may be lawful if they meet the combatant-status criteria. India’s position, which is silence, avoids the classification question entirely. Pakistan’s position is that India is conducting extrajudicial killings on Pakistani soil in peacetime, violating both Pakistani sovereignty and international human rights law. The legal debate cannot resolve this question without first resolving the factual question of Indian involvement, which remains officially denied. The comparison between India’s approach and Israel’s framework reveals the fundamental difference: Israel built a legal architecture to justify operations it acknowledges. India operates in a legal void because it acknowledges nothing.

The United Kingdom’s position on combatant status aligns broadly with the mainstream of international humanitarian law: individuals may only be targeted during armed conflict, the targeting must satisfy proportionality and necessity, and collateral damage to civilians must be minimized. The Intelligence Services Act 1994 does not authorize lethal operations, and the UK has no domestic equivalent of the US Authorization for Use of Military Force. When the UK conducted the Reyaad Khan strike, it argued that Khan was a combatant in the Syrian armed conflict who posed an imminent threat to British security, combining the armed-conflict and self-defense frameworks into a hybrid justification. The precedent is narrow and has not been expanded, but it demonstrates that even a state committed to restrictive legal interpretations will adapt its framework when operational imperatives demand it.

The combatant-status question exposes a deeper structural problem in the law of armed conflict. International humanitarian law was designed for conflicts between organized military forces wearing uniforms, carrying arms openly, and operating under a chain of command. Non-state armed groups rarely meet all of these criteria, yet they conduct operations that cause mass casualties and threaten national security. The legal architecture has not evolved at the same pace as the threat landscape. The result is that states apply Cold War-era legal categories to twenty-first-century security challenges, stretching definitions and creating interpretive innovations that the drafters of the Geneva Conventions could not have anticipated. The “direct participation in hostilities” concept, the “continuous combat function” test, and the “functional combatant” analysis are all attempts to fit modern realities into a legal framework designed for a different era of warfare. Each attempt creates new ambiguities that states exploit according to their operational needs.

The question of who determines combatant status is as important as the substantive standard. In Israel, the determination is made by military lawyers embedded in the operational chain of command, with post-hoc judicial review available through the Supreme Court. In the United States, the determination is made by intelligence analysts and executive-branch lawyers, with no judicial review. In India, the determination, to the extent it occurs at all, is made entirely within the intelligence apparatus with no oversight mechanism of any kind. The entity that determines combatant status has enormous power: it decides who lives and who dies, and the reliability of its determinations depends on the quality of its intelligence, the independence of its judgment, and the accountability structures that govern its work. When the determining entity is the same institution that benefits operationally from classifying individuals as targetable, the risk of confirmation bias is significant. The American experience with “signature strikes,” drone attacks targeting individuals based on behavioral patterns rather than positive identification, illustrates the danger: some individuals targeted on the basis of behavioral signatures turned out to be civilians, and the post-strike review process was unable to correct the error before the killing occurred.

The Accountability Mechanism

Legal frameworks exist not only to authorize or prohibit but to provide accountability: a mechanism for reviewing whether a particular operation complied with the applicable rules. The five countries diverge most sharply on accountability, and the divergence reveals the most important truth about the targeted killing debate: the legal framework matters less than the accountability mechanism that enforces it.

Israel has the most robust accountability mechanism of any state conducting targeted killings, and it still falls far short of genuine oversight. The Supreme Court’s 2006 ruling required independent investigation after each operation, but the court did not specify what institution would conduct the investigation, what standards it would apply, or what consequences would follow a finding of illegality. In practice, post-operation reviews are conducted by the Military Advocate General’s office, which is part of the military chain of command rather than an independent civilian body. The reviews examine whether the operation complied with the targeting criteria established by the legal department of the IDF, but the criteria themselves are classified and not subject to public scrutiny. Human rights organizations including B’Tselem and the Public Committee Against Torture have documented cases in which targeted killings resulted in significant civilian casualties, including the July 2002 strike on Salah Shehadeh’s apartment in Gaza City that killed fourteen civilians including nine children. The investigation that followed concluded that the operation was conducted with the best available intelligence and that the civilian casualties, while regrettable, did not render the operation unlawful. The verdict illustrated the limitation of an accountability mechanism that operates within the institution it is supposed to oversee.

The United States has layered its accountability mechanisms but has never created genuine external oversight. The targeting process for drone strikes involves multiple institutional actors: the CIA’s Counterterrorism Center identifies potential targets, the National Security Council’s Principals Committee reviews nominations, the Office of Legal Counsel provides legal authorization, and (in some periods) the President personally approves each strike. The process is elaborate and involves significant interagency deliberation, but every participant is part of the executive branch. No court reviews targeting decisions before they are executed. No independent body audits the accuracy of post-strike casualty assessments. The Senate and House Intelligence Committees receive classified briefings on the drone program but have no authority to approve or block individual operations. Philip Alston, who served as the UN Special Rapporteur on extrajudicial, summary, or arbitrary executions from 2004 to 2010, argued that the US accountability framework creates the appearance of oversight without the substance, because no external institution has the authority, the information, or the independence to challenge targeting decisions made within the classified intelligence system. The comparison between the US drone program and India’s shadow war reveals that the United States, despite its elaborate internal processes, has not produced a single criminal prosecution of a government official for an unlawful targeted killing, suggesting that the accountability mechanism is designed to validate decisions rather than question them.

Russia has no accountability mechanism whatsoever. Russian courts do not review intelligence operations. The Russian parliament does not exercise oversight over the security services in any meaningful sense. The FSB, GRU, and SVR operate under presidential authority with no external check. When European courts have found Russian involvement in targeted killings, Russia has ignored the verdicts. The European Court of Human Rights ruled in its 2014 judgment in Georgia v. Russia that Russia was responsible for the killing of Zelimkhan Khangoshvili. Russia rejected the court’s jurisdiction and refused to comply with its orders. The International Criminal Court has no jurisdiction over Russian intelligence operations because Russia has not ratified the Rome Statute. The Salisbury investigation produced detailed evidence of GRU involvement, including identification of the operatives by name, but the British criminal justice system has no mechanism to compel Russian cooperation, and the extradition request issued by the Crown Prosecution Service has been predictably refused. The Russian case demonstrates that accountability for targeted killing requires either domestic institutions willing to investigate their own government or international institutions with enforcement power. Russia has neither.

India occupies the same accountability vacuum as Russia but for different reasons. Russia has accountability-resistant institutions that actively protect the security services from scrutiny. India has no institutions engaged in the targeted killing question at all, because the official position is that no targeted killings are occurring. The Indian judiciary has not examined the question because no case has been brought. The Comptroller and Auditor General does not audit intelligence operations. The Cabinet Committee on Security, which oversees RAW and the Intelligence Bureau, operates in total secrecy with no public reporting requirement. The Parliamentary Standing Committee on Home Affairs receives periodic briefings from intelligence officials but has no authority to review individual operations, and its proceedings on intelligence matters are classified. The accountability gap is total: there is no institution, domestic or international, that reviews whether a particular killing on Pakistani soil was authorized, proportionate, necessary, or lawful. The absence of accountability is not an oversight; it is a structural feature of India’s total-deniability approach. Accountability requires acknowledgment, and acknowledgment is precisely what the shadow war’s operational design is built to prevent.

The United Kingdom has formal accountability mechanisms that function more effectively than those of any other state in this comparison, primarily because the UK conducts fewer targeted killings and exercises greater restraint. The Investigatory Powers Tribunal reviews complaints about intelligence agency conduct. The Intelligence and Security Committee of Parliament oversees MI6, MI5, and GCHQ. The Investigatory Powers Commissioner’s Office audits the use of surveillance powers. These mechanisms have produced substantive investigations, including the Investigatory Powers Tribunal’s examination of GCHQ’s mass surveillance programs, which found certain practices unlawful. The limitation is that the oversight mechanisms apply to British intelligence activities, not to the activities of allied states that use British intelligence. If GCHQ provides geolocation intelligence that enables an American drone strike, the British oversight system reviews the sharing decision, not the strike itself. The legal question of whether sharing intelligence that contributes to an unlawful killing creates legal liability for the sharing state remains unresolved, and the UK oversight system is not designed to answer it.

Where the Comparison Breaks Down

The five-country comparison illuminates the structural choices that states make when they decide to kill on foreign soil, but it also obscures differences that matter. The comparison works when it holds all five cases in the same analytical frame: sovereignty, self-defense, combatant status, accountability. It breaks down when the cases diverge on dimensions that the legal framework does not capture.

The first divergence is target selection. Israel targets members of armed groups with which it is in active armed conflict. The United States targets members of designated terrorist organizations. Russia targets defectors, dissidents, and individuals who have personally betrayed the state. India targets designated terrorists who have been charged in specific attacks against Indian citizens. The legal framework treats all targeted killings as equivalent acts of lethal force, but the political and moral logic of each category is different. Killing a Hamas military commander who is actively directing rocket attacks against Israeli civilians occupies a different position on the moral spectrum than killing a former intelligence officer who has provided information to a foreign government. The legal framework does not capture this distinction because international humanitarian law and human rights law address the method and context of killing, not the moral character of the target. The case of Jamal Khashoggi, a journalist killed by Saudi operatives in the Istanbul consulate, demonstrated that public opinion draws sharp moral distinctions between categories of targets even when the legal analysis is identical. The Khashoggi case produced universal condemnation not because the method was different from other targeted killings but because the target was a journalist, not a terrorist.

The second divergence is collateral damage. The US drone program has killed hundreds of civilians according to independent monitoring organizations including the Bureau of Investigative Journalism and the New America Foundation. Israeli targeted killings have killed civilians in nearly every major operation, including the Shehadeh strike, the 2014 Gaza operations, and the ongoing conflict. Russia’s Skripal operation endangered multiple British civilians through Novichok contamination, and a British citizen, Dawn Sturgess, died from exposure to the discarded nerve agent container. India’s shadow war has produced minimal documented collateral damage: the motorcycle-borne killings target individuals in isolation, and the method, close-range shooting by operatives who then flee, is inherently more discriminate than drone strikes or nerve agents. The legal debate treats proportionality as a universal standard, but the operational reality is that different methods produce dramatically different collateral-damage profiles. India’s method, whatever its legal status, is operationally closer to the precision that proportionality requires than the US drone program that claims legal authority for its strikes.

The third divergence is democratic authorization. Israel and the United States are democracies with at least nominally independent judiciaries, free press, and electoral accountability. Russia is an authoritarian state where the security services operate beyond civilian control. India is a democracy, but the targeted killing program operates entirely outside the democratic framework: no parliamentary vote, no judicial review, no public debate. The United Kingdom is a democracy with strong oversight institutions, but those institutions do not extend to the activities of allied states. The comparison reveals that democratic governance does not automatically produce legal accountability for targeted killings. Israel’s Supreme Court reviewed the legality of targeted killings only because human rights organizations filed a petition; absent the petition, no review would have occurred. The US Congress authorized force against al-Qaeda but has not voted on the geographic or temporal expansion of that authorization. India’s parliament has never debated the targeted killing program because the program officially does not exist. Democratic systems provide the structural possibility of accountability, but they do not guarantee that the possibility will be realized.

The fourth divergence is international response. Israel faces sustained international criticism for targeted killings, including regular UN Human Rights Council resolutions, International Criminal Court investigations, and diplomatic consequences. The United States faces legal challenges from civil liberties organizations and criticism from international human rights bodies, but the diplomatic consequences have been modest. Russia faces diplomatic sanctions and intelligence-relationship disruptions following each exposed operation. India faces the least international consequence of any state conducting targeted killings, partly because of the deniability that prevents attribution, partly because the targets are universally designated terrorists, and partly because the international community has reached an unstated consensus that Pakistan’s harboring of terrorists responsible for mass-casualty attacks creates a degree of tolerance for the operations that address the threat. The international response pattern reveals that the legal framework is not the primary determinant of consequences. The political relationship between the perpetrating state and the international community, the target’s identity, the method’s collateral impact, and the strength of attribution evidence all matter more than the formal legal analysis.

The fifth divergence, often overlooked in legal scholarship, is the question of institutional learning and doctrinal evolution. Israel’s targeted killing program has evolved over six decades, from the relatively crude operations of the 1960s (the Eichmann capture required an entire team operating in Argentina for weeks) to the precision strikes of the contemporary era (the killing of Iranian nuclear scientist Mohsen Fakhrizadeh in November 2020 reportedly involved a remotely operated weapon system). Each evolution reflected lessons learned from operational failures, legal challenges, and political consequences. The Lillehammer affair of 1973, in which Mossad operatives killed Ahmed Bouchikhi, an innocent Moroccan waiter they mistook for a Black September operative in Norway, produced immediate legal consequences (the operatives were arrested and prosecuted in Norwegian courts) and long-term doctrinal changes (improved target verification procedures). The United States similarly evolved its drone program in response to criticism, tightening targeting criteria under the Obama administration’s 2013 Presidential Policy Guidance and then loosening them under the Trump administration’s 2017 replacement. India’s program, by contrast, shows no evidence of doctrinal evolution driven by legal or accountability pressures, because no such pressures exist. The total-deniability model insulates the program from the feedback mechanisms that have shaped Israeli and American practice, which means that operational mistakes, if they occur, will not produce the institutional learning that legal frameworks and accountability mechanisms generate.

The sixth divergence concerns the relationship between targeted killing and broader strategic objectives. Israel’s program is embedded within a comprehensive security strategy that includes conventional military operations, intelligence collection, diplomatic engagement, and civilian defense measures. Targeted killing is one tool among many, and its effectiveness is assessed in relation to the broader strategic picture. The United States similarly embedded its drone program within a counter-terrorism strategy that included special operations, intelligence partnerships, financial disruption, and diplomacy. India’s shadow war operates alongside conventional military capability (as demonstrated by the 2016 surgical strikes and Operation Sindoor) but exists in a separate institutional compartment. The absence of any public framework for the shadow war means that its relationship to India’s broader security strategy is unclear. Whether the targeted killings serve the strategic objective of deterring future attacks, the operational objective of degrading terror networks, the political objective of demonstrating resolve, or some combination of all three is a question that India’s total-deniability approach prevents from being publicly debated. Strategic clarity requires institutional transparency, and India’s approach provides neither.

What the Comparison Teaches

The five-country comparison yields a single overarching conclusion: international law on targeted killing is not settled. It is contested. The contest is not between states that obey the law and states that violate it. It is between states that exploit different forms of ambiguity to sustain operations that the legal framework can neither clearly authorize nor effectively prohibit.

Israel exploits the ambiguity of the direct-participation standard, defining it broadly enough to cover a wide range of targets while citing judicial endorsement as evidence of legality. The United States exploits the ambiguity of the imminence standard, redefining it to cover continuing threats that would not meet any traditional definition of imminent. Russia exploits the ambiguity of attribution, conducting operations that leave traces but refusing to acknowledge them, forcing the international system to respond with diplomatic measures rather than legal ones. India exploits the ambiguity of attribution even more effectively, conducting operations that leave no state-traceable evidence and forcing the legal system to confront a situation in which the killing occurs but the perpetrator officially does not exist. The United Kingdom exploits the ambiguity of complicity, sharing intelligence that enables killings conducted by others while maintaining that the sharing itself does not constitute participation in the act.

The targeted killing legal framework comparison across these five countries reveals that each has calibrated its approach to its domestic political needs and international positioning. Israel needs judicial endorsement because its Supreme Court is influential and because Israel cares about its legal reputation in the Western legal community. The United States needs executive legal authority because the American constitutional system requires legal authorization for military operations and because the ACLU and other organizations will litigate unauthorized uses of force. Russia needs nothing because the Russian political system does not generate domestic legal challenges to intelligence operations. India needs nothing because the operations officially do not exist. The United Kingdom needs ambiguity because British public opinion and parliamentary oversight create political risk if the government is seen as complicit in unlawful killings.

Paul Staniland’s framework for understanding state-constructed legal justifications applies precisely to this pattern. Each state has constructed a framework that serves its domestic political needs, not a framework that reflects a principled analysis of international law. The legal frameworks are tools of political management, not expressions of legal conviction. This does not mean the frameworks are irrelevant. The Israeli framework creates real constraints: the proportionality requirement, however imperfectly enforced, produces operational modifications that reduce civilian casualties. The American framework creates real processes: the interagency targeting review, however captured by institutional interests, produces deliberation that excludes some potential targets. The British framework creates real limitations: the domestic prohibition on assassination constrains MI6 operations even if it does not constrain intelligence sharing. Legal frameworks shape behavior even when they do not determine outcomes.

The comparison also teaches that the most important variable in the targeted killing debate is not the legal framework but the accountability mechanism. Israel has a framework and a partial accountability mechanism: its operations are more constrained than those of any other state in this comparison. The United States has a framework and an internal accountability mechanism: its operations are more constrained than Russia’s but less constrained than Israel’s. Russia has neither framework nor accountability: its operations are the least constrained and produce the most collateral damage. India has neither framework nor accountability: its operations are operationally precise but legally ungoverned. The UK has a framework and external oversight: its operations are the most constrained of all five states, which is why it conducts the fewest targeted killings. The pattern is clear: accountability constrains behavior more effectively than legal framework alone.

George Perkovich’s concern about the dangerous legal vacuum is validated by the comparison. India’s total-deniability approach produces operationally effective results in the short term, but it creates long-term risks that the other four approaches mitigate. Without a legal framework, there is no standard for proportionality; a future operation that kills civilians will not be measured against any established rule. Without an accountability mechanism, there is no process for correcting errors; a future operation that targets the wrong person will not produce consequences for the decision-makers responsible. Without democratic authorization, there is no political mechanism for debating whether the program serves national interests; the program’s continuation depends entirely on the judgment of intelligence officials who have every institutional incentive to continue operations that justify their budgets and enhance their status. The extended legal analysis explores these long-term risks in greater depth, but the B-series overview can state the conclusion directly: India’s approach works until it doesn’t, and the absence of any framework means that there is no mechanism for determining when it has stopped working.

The final lesson of the comparison is the most uncomfortable one. The legal ambiguity surrounding targeted killing is not an accident. It is a feature of the international system that serves the interests of the states that conduct targeted killings and the states that tolerate them. Resolving the ambiguity in either direction, by clearly permitting targeted killings under specified conditions or by clearly prohibiting them, would create political costs that no major power is willing to accept. Clear permission would normalize assassination as an instrument of state policy, undermining the sovereignty norms that protect weak states from strong ones. Clear prohibition would require enforcement against states that conduct operations with broad international sympathy, including Israel’s operations against Hamas and India’s operations against terrorists responsible for mass-casualty attacks. The ambiguity persists because it allows the international community to respond to each case on its political merits rather than its legal merits, condemning operations that violate norms the community cares about (Khashoggi, Skripal) while tolerating operations that address threats the community recognizes (Hamas commanders, designated terrorists). The legal debate on targeted killing will not be resolved because resolution would require the international community to make choices it prefers to avoid.

The historical trajectory of state practice reinforces this analysis. In the 1970s, when Mossad’s Wrath of God operatives hunted the architects of the Munich Olympics massacre across European capitals, the operations were considered extraordinary and generated significant diplomatic controversy. In the 2000s, when the United States launched its drone program across multiple continents, targeted killing became routinized, conducted thousands of times with decreasing levels of public scrutiny and diplomatic consequence. In the 2020s, India’s shadow war represents the latest expansion of state practice: targeted killing conducted without legal framework, without acknowledgment, and without formal international opposition. Each decade has normalized what the previous decade considered exceptional. The direction of travel is clear: more states will acquire the capability and the willingness to conduct targeted killings, and the legal debate will continue to trail behind the operational reality, producing scholarship that describes what has already happened rather than rules that constrain what happens next.

The normative implications extend beyond the five countries examined here. If India’s total-deniability model succeeds, meaning that the operations achieve their strategic objectives without triggering legal accountability or sustained diplomatic consequences, other states will adopt similar approaches. The precedent is particularly relevant for middle powers with intelligence capabilities and unresolved security threats: countries like Turkey, Saudi Arabia, Egypt, and Iran already conduct extraterritorial operations, and India’s model offers a template for doing so without the legal infrastructure that Israel and the United States constructed. The proliferation of targeted killing programs without legal frameworks would represent a fundamental shift in the international security order, from a system in which lethal force on foreign soil is exceptional and legally constrained to a system in which it is routine and legally ungoverned. Whether that shift serves global security or undermines it depends on whether the legal debate can produce accountability mechanisms that keep pace with operational capability, and nothing in the five-country comparison suggests that it can.

The international response to India’s targeted killings reflects this pattern with precision. The United States, which conducted thousands of drone strikes during the war on terror, is in no position to condemn India for conducting dozens of targeted killings against designated terrorists. Canada, which accused India of involvement in the killing of Hardeep Singh Nijjar, faces the complication that Nijjar was designated as a terrorist by India and that Canada’s own intelligence agencies had previously monitored his activities. The United Kingdom, which shares intelligence with allies who conduct targeted killings, cannot credibly object to India’s operations without implicating its own intelligence cooperation. The international legal framework is clear in theory: targeted killings on foreign soil without consent violate sovereignty and, absent armed conflict, violate the prohibition on extrajudicial killing. The international political response ignores the legal framework because enforcing it would require states to confront their own conduct and their own complicity.

Countries do not resolve the legal ambiguity surrounding targeted killing. They exploit it. The exploitation is not lawlessness; it is a form of legal entrepreneurship in which each state constructs the minimum viable framework needed to sustain operations that serve its security interests while maintaining sufficient ambiguity to avoid accountability. The five-country comparison demonstrates that the targeted killing debate is not a conversation between law and lawlessness. It is a conversation between competing forms of managed ambiguity, each designed to serve the state that constructed it. The law follows power, and power has no interest in clarity.

Frequently Asked Questions

Q: Is it legal to kill terrorists on foreign soil under international law?

International law does not provide a single answer. The legality depends on which legal framework applies. Under international humanitarian law, individuals who directly participate in hostilities during an armed conflict may be targeted, subject to proportionality and necessity. Under international human rights law, which applies outside of armed conflict, extrajudicial killing is prohibited without exception. The critical question is whether an armed conflict exists between the state conducting the killing and the armed group to which the target belongs. Israel argues that its conflict with Hamas constitutes an armed conflict, making targeted killings lawful under IHL. The United States argued that its conflict with al-Qaeda constituted a global armed conflict, extending targeting authority worldwide. If no armed conflict exists, as in the case of Russia’s killings of defectors in Europe, the killing violates human rights law regardless of the target’s identity. India’s situation is ambiguous: India and Pakistan have an established history of armed conflict, but the targeted killings occur during periods that would conventionally be classified as peacetime, placing them in a legal gray zone that neither framework clearly governs.

Q: What does Article 51 of the UN Charter say about self-defense and targeted killings?

Article 51 preserves the “inherent right of individual or collective self-defense if an armed attack occurs against a Member of the United Nations.” The article was drafted to address attacks by states against states. Its application to attacks by non-state actors operating from third-party territory is contested. The United States and Israel interpret Article 51 broadly, arguing that an armed attack by a non-state actor triggers the right of self-defense regardless of whether the host state sponsored the attack. Most international law scholars argue that Article 51 requires the armed attack to be attributable to the host state or that the host state must have been given an opportunity to address the threat before force is used on its territory. The International Court of Justice addressed this question obliquely in its 2004 Wall advisory opinion, suggesting that Article 51 addresses inter-state attacks, but the opinion drew strong dissents and has not settled the debate.

Q: What is the “unwilling or unable” doctrine in targeted killing law?

The “unwilling or unable” doctrine holds that when a state is either unwilling or unable to address a terrorist threat emanating from its territory, the threatened state retains the right to use force on that territory in self-defense. The doctrine bridges the gap between the self-defense justification and the sovereignty barrier: it explains why force is directed at the territory of a state that did not conduct the armed attack. The United States has been the most vocal proponent of the doctrine, invoking it to justify drone operations in Pakistan, Yemen, and Somalia. Israel has applied similar reasoning to operations in Lebanon and Syria. Critics argue that the doctrine allows powerful states to override the sovereignty of weaker states by making unilateral determinations about the host state’s willingness or ability. The doctrine has not been formally adopted by any international tribunal or codified in any treaty, and its status in customary international law remains contested.

Q: How did Israel’s Supreme Court rule on targeted killings in 2006?

The Israeli Supreme Court’s December 2006 ruling in Public Committee Against Torture v. Government of Israel held that targeted killings of individuals who directly participate in hostilities are lawful under international humanitarian law if four conditions are met. First, the target must be a person taking a direct part in hostilities. Second, the killing must be a last resort after capture has been attempted or deemed infeasible. Third, the operation must satisfy proportionality, meaning that expected civilian harm must not be excessive relative to the anticipated military advantage. Fourth, an independent investigation must follow each operation. The ruling created a conditional framework rather than granting blanket authorization, and it remains the only judicial decision by any national supreme court that directly addresses the legality of targeted killing. The ruling’s practical impact has been debated: human rights organizations argue that the investigation requirement is rarely enforced, while Israeli officials argue that the framework shapes operational planning and reduces civilian casualties.

Q: What were the US Office of Legal Counsel drone memos?

The Office of Legal Counsel within the US Department of Justice produced a series of classified legal memoranda beginning in 2001 that provided the legal authorization for the CIA and Department of Defense to conduct lethal operations against members of al-Qaeda and associated forces. The memos rested on three legal pillars: the September 2001 Authorization for Use of Military Force, the inherent right of self-defense under Article 51 of the UN Charter, and the law of armed conflict applied to a global non-international armed conflict with al-Qaeda. The most controversial aspect was the redefinition of “imminent threat” to include situations where a senior operational leader posed a continuing danger even if no specific attack was known to be planned. A separate memo addressed the targeting of American citizens, concluding that constitutional due-process protections yielded to national-security interests when the citizen posed an imminent threat and capture was infeasible. Portions of the memos were released publicly in 2014 following ACLU litigation.

Q: Why does India not acknowledge conducting targeted killings in Pakistan?

India maintains total deniability regarding the targeted killings documented across Pakistani cities. The denial serves multiple strategic purposes. Acknowledgment would transform a covert intelligence operation into an open act of war, potentially triggering military escalation between two nuclear-armed states. Acknowledgment would invite legal challenge in Indian courts, where the constitutional right to life has been interpreted broadly by the Supreme Court. Acknowledgment would require India to articulate targeting criteria, proportionality standards, and an accountability mechanism, creating legal obligations that the intelligence establishment has every incentive to avoid. Acknowledgment would also provide Pakistan with formal diplomatic ammunition to pursue the matter at the UN Security Council, the International Court of Justice, or other international forums. The total-deniability approach sacrifices legal legitimacy to preserve operational flexibility, strategic ambiguity, and escalation control.

Q: How does Russia justify targeted killings like the Skripal poisoning?

Russia does not justify targeted killings. Russia denies conducting them. The official Russian position on the Skripal poisoning in Salisbury was that Russia was not involved, that the Novichok nerve agent could have been produced by multiple countries, and that the British investigation was politically motivated. Russia maintained this denial even after British and Dutch intelligence identified the GRU operatives responsible by name and published their photographs. The denial strategy is functional: by refusing to acknowledge the operations, Russia avoids the obligation to provide a legal justification, which in turn prevents the legal debate from engaging with Russian conduct in a structured way. The approach carries diplomatic costs, as the Salisbury poisoning triggered the largest coordinated expulsion of Russian diplomats in history, but Russia has calculated that the operational benefit of maintaining a credible assassination capability outweighs the diplomatic cost of periodic exposure.

Q: Does the UN Charter permit preemptive self-defense against terrorist threats?

The UN Charter text addresses self-defense in the event of an armed attack, using language that most scholars interpret as requiring an attack to have occurred or to be imminent before the right of self-defense is triggered. Preemptive self-defense, the use of force to prevent a future attack that has not yet occurred, is not explicitly authorized by the Charter. The United States has argued that the right of self-defense includes the right to act preemptively when a credible threat exists, particularly in the context of terrorism where attacks are difficult to detect in advance. The Bush administration’s 2002 National Security Strategy formally adopted preemptive self-defense as a policy, and the Obama administration’s “continuing and imminent threat” standard extended this logic by redefining imminence to include ongoing threats that would not meet any traditional definition. The legal community remains divided: a minority of scholars accepts preemptive self-defense as consistent with the Charter’s purpose, while the majority holds that the Charter requires an actual or genuinely imminent attack.

Q: What accountability mechanisms exist for targeted killings?

Accountability mechanisms vary dramatically across states. Israel has the most developed system: the Supreme Court has established legal criteria, and the Military Advocate General’s office conducts post-operation reviews, though critics argue these reviews lack independence. The United States has an elaborate interagency targeting process involving the CIA, National Security Council, and Office of Legal Counsel, but all participants are executive-branch entities with no independent oversight. Congressional intelligence committees receive classified briefings but cannot approve or block individual operations. Russia has no accountability mechanism; Russian courts do not review intelligence operations, and international court rulings are ignored. India has no accountability mechanism because the operations officially do not exist. The United Kingdom has formal oversight bodies including the Investigatory Powers Tribunal and the Intelligence and Security Committee of Parliament, but their jurisdiction extends to British intelligence activities, not to allied operations enabled by British intelligence sharing.

Q: Can a targeted killing satisfy the proportionality requirement under international law?

Proportionality in international humanitarian law requires that the expected collateral damage to civilians and civilian property not be excessive in relation to the concrete and direct military advantage anticipated. A targeted killing can satisfy this requirement if the operation is designed to minimize civilian harm, the intelligence supporting the operation is reliable, and the anticipated harm to civilians is proportionate to the military significance of the target. Israel’s Supreme Court incorporated proportionality as one of four conditions for lawful targeted killing. In practice, proportionality assessments are conducted internally by military lawyers using classified intelligence, making independent verification difficult. The US drone program has faced sustained criticism on proportionality grounds because of documented civilian casualties. India’s motorcycle-borne operations produce minimal collateral damage, which paradoxically means they come closer to satisfying the proportionality requirement than the legally authorized drone strikes of the United States, though India’s operations lack the legal authorization that would make the proportionality analysis relevant.

Q: How does the UK balance prohibiting targeted killing domestically while sharing intelligence with allies who conduct them?

The United Kingdom prohibits its intelligence agencies from conducting assassinations under domestic law, but it shares intelligence with allied states including the United States and Israel that conduct targeted killings using, in some cases, intelligence provided by British agencies. The legal question of whether intelligence sharing that contributes to a targeted killing creates legal liability for the sharing state remains unresolved in international law. The UK government has argued that intelligence sharing decisions are reviewed through established oversight mechanisms and that sharing intelligence is legally distinct from conducting a kinetic operation. Human rights organizations have challenged this distinction, arguing that providing targeting intelligence makes the sharing state a participant in the operation. The Investigatory Powers Tribunal has examined specific cases but its findings are classified. The UK maintains deliberate legal ambiguity on the complicity question, which allows continued intelligence cooperation while avoiding a formal legal position that would either endorse or condemn the allied operations enabled by British intelligence.

Q: What is the difference between a targeted killing and an assassination?

The distinction is legal rather than operational. In international humanitarian law, a “targeted killing” refers to the deliberate use of lethal force against a specific individual when the state considers the killing lawful under applicable legal frameworks, typically during armed conflict against a combatant or a civilian directly participating in hostilities. “Assassination” implies an unlawful killing, often politically motivated, outside the framework of armed conflict. The United States officially prohibits assassination under Executive Order 12333, signed by President Reagan in 1981, but the executive branch has argued that targeted killings of enemy combatants during armed conflict do not constitute assassination and therefore do not violate the executive order. Israel has adopted similar reasoning. The distinction is functionally circular: if the legal framework authorizes the killing, it is a targeted killing; if it does not, it is an assassination. The operational act, killing a specific named individual through premeditated action, is identical in both cases.

Q: Has any state faced lasting legal consequences for conducting targeted killings?

No state that conducts targeted killings as a matter of policy has faced lasting legal consequences through international legal mechanisms. The International Criminal Court has investigated Israeli operations in the Palestinian territories, but no senior Israeli official has been prosecuted for authorizing targeted killings. The European Court of Human Rights has issued judgments against Russia for killings attributed to Russian intelligence, but Russia has refused to comply with the court’s orders. The International Court of Justice has not adjudicated any case directly concerning state-sponsored targeted killing. The consequences that states face are diplomatic and political rather than legal: sanctions, diplomatic expulsions, intelligence-relationship disruptions, and reputational damage. The absence of legal consequences reflects the structural limitation of international law enforcement: the institutions empowered to adjudicate violations lack the enforcement mechanisms to compel compliance, particularly against permanent Security Council members or major regional powers.

Q: Why do different targeted killings produce different international responses?

The international response to targeted killings is determined by political factors more than legal analysis. Four variables shape the response. First, target identity: killing a designated terrorist produces less condemnation than killing a journalist (Khashoggi) or a defector (Litvinenko). Second, method: operations that endanger civilians through nerve agents or large-scale ordnance produce stronger responses than precise operations with minimal collateral damage. Third, attribution strength: operations with clear forensic evidence of state involvement (Salisbury) produce stronger responses than operations where state involvement is alleged but not proven (India’s shadow war). Fourth, the perpetrator’s geopolitical position: the international community responds more vigorously to Russian operations in Europe than to Indian operations in Pakistan because the European states affected by Russian operations are powerful and united, while Pakistan’s complaints about Indian operations receive less diplomatic traction. The legal analysis is largely uniform across cases; the political calculus is highly variable.

Q: Does international law need to be clarified on targeted killing?

Legal scholars are divided. One position, associated with scholars like Mary Ellen O’Connell, holds that the existing law is clear: targeted killing outside of armed conflict violates human rights law, and within armed conflict it is permitted only under strict conditions that most operations fail to meet. The problem is enforcement, not clarity. The opposing position, associated with scholars like Kenneth Anderson, holds that the existing law fails to account for the reality of transnational terrorism and needs updating to provide clear rules for states facing non-state threats that originate from foreign territory. A third position, advanced implicitly by the operational behavior of the states in this comparison, holds that ambiguity serves a stabilizing function by allowing necessary operations while maintaining normative pressure against indiscriminate violence. Clarifying the law in either direction would create political costs: clear permission would normalize assassination, while clear prohibition would require enforcement against states with broad international sympathy.

Q: How does India’s approach compare to the Israeli approach to targeted killing?

Israel and India both conduct targeted killings of designated terrorists, but their legal and institutional approaches are diametrically opposed. Israel acknowledges targeted killings, subjects them to judicial review, publishes the legal criteria governing them, and conducts post-operation investigations (however imperfect). India denies conducting targeted killings, has no judicial review mechanism, has published no legal criteria, and conducts no known post-operation investigations. Israel accepts the political and diplomatic cost of acknowledgment in exchange for legal legitimacy and domestic accountability. India avoids the political cost of acknowledgment by operating in total deniability, sacrificing legal legitimacy and accountability in exchange for operational flexibility and escalation control. The Israeli model is legally stronger and operationally more constrained. India’s model is legally vulnerable but operationally more flexible. Whether India’s approach is sustainable depends on whether the deniability can be maintained as the scale of operations increases and the international scrutiny intensifies.

Q: What role does the concept of state sovereignty play in the targeted killing debate?

State sovereignty, as codified in Article 2(4) of the UN Charter, is the foundational legal barrier to targeted killings on foreign soil. The principle holds that every state has exclusive authority over its territory and that no other state may use force within that territory without consent. Every unconsented targeted killing violates this principle. The practical significance of the sovereignty barrier depends on enforcement: a state that can enforce its sovereignty against foreign operations (as the UK did through diplomatic action after Salisbury) creates consequences for the violation. A state that cannot effectively enforce sovereignty over its own territory (as Pakistan cannot fully control the FATA region or the security situation in Karachi) creates operational space for foreign operations. The sovereignty debate in targeted killing law is ultimately a debate about power: states with the military and intelligence capability to operate on foreign soil do so, and the sovereignty principle constrains them only to the extent that the host state or the international community can impose costs.

Q: What is the legal status of the “continuous combat function” test?

The “continuous combat function” test was developed by the International Committee of the Red Cross in its 2009 “Interpretive Guidance on the Notion of Direct Participation in Hostilities under International Humanitarian Law,” led by Nils Melzer. The test holds that members of organized armed groups who perform a continuous combat function, meaning their role within the group involves ongoing participation in hostilities rather than occasional or incidental involvement, may be targeted at any time during the armed conflict, not only during specific acts of participation. Civilians who participate directly in hostilities on a sporadic or one-time basis may only be targeted during the specific act of participation. The ICRC’s guidance is influential but not legally binding; it represents one interpretation of existing IHL rather than a new rule. Israel has not adopted the continuous combat function test, preferring its own broader interpretation of direct participation. The United States applies a similar concept through its “functional combatant” analysis but has not formally adopted the ICRC framework. The test’s legal status remains advisory, not authoritative.

Q: Why is the Khashoggi case considered different from other targeted killings?

The Khashoggi killing in Istanbul in October 2018 produced universal international condemnation because it violated two norms simultaneously. First, the target was a journalist, not a combatant or a designated terrorist. Jamal Khashoggi was a Washington Post columnist and Saudi dissident who entered the Saudi consulate in Istanbul to obtain marriage documents. Killing a journalist, even a dissident journalist critical of his government, crosses a normative line that killing a designated terrorist does not. Second, the method violated diplomatic norms: the killing occurred inside a consulate, a space protected by international diplomatic conventions, weaponizing the very diplomatic protections that are supposed to ensure individual safety. The combination of a protected target in a protected space made the Khashoggi case qualitatively different from operations against combatants or terrorists in non-diplomatic settings. India’s shadow war avoids both violations: its targets are designated terrorists (accepted targets in the international consensus), and its operations occur in public spaces (no diplomatic abuse). The contrast explains why Khashoggi produced universal condemnation while India’s operations have produced selective silence.

Q: Can targeted killing ever be considered a form of law enforcement rather than military action?

Technically, international human rights law permits the use of lethal force in law-enforcement operations when strictly necessary to protect life and when no less lethal alternative is available. This standard, drawn from the UN Basic Principles on the Use of Force and Firearms by Law Enforcement Officials, is far more restrictive than the IHL standard governing armed conflict. For a targeted killing to qualify as law enforcement, the operation would need to satisfy several conditions: the state must have attempted arrest, arrest must have been genuinely infeasible, the lethal force must have been the minimum necessary, and the threat must have been imminent to human life. Very few targeted killings would survive this test. The Israeli Supreme Court’s “last resort” condition approximates the law-enforcement standard by requiring that arrest be attempted before lethal force is used, but Israel applies this condition within the IHL framework of armed conflict, not the human-rights framework of law enforcement. The distinction matters because if a targeted killing is classified as law enforcement rather than military action, the proportionality and necessity requirements become significantly more demanding.

Q: What precedent does the US drone program set for other countries considering targeted killings?

The US drone program established several precedents that have influenced the global targeted killing debate. It demonstrated that a state can conduct sustained lethal operations across multiple countries using remotely piloted aircraft with limited domestic political opposition. It established the “continuing and imminent threat” standard that expanded the legal authorization for targeted killing beyond traditional imminence requirements. It created an interagency targeting process that other countries have studied and, in some cases, replicated. It demonstrated that civilian casualties, even significant ones, do not necessarily produce political consequences sufficient to halt a program. And it proved that the international legal community’s objections, however sustained and well-reasoned, have limited practical impact on a program supported by the executive branch of a superpower. For countries considering targeted killing programs, the US precedent suggests that legal authorization, operational infrastructure, and domestic political support matter more than international legal approval. India’s shadow war operates on a different operational model (human operatives rather than drones), but the strategic logic, targeting designated threats on foreign soil when the host state is unwilling or unable to act, is directly influenced by the precedent the US program established.

Q: How might the legal framework for targeted killing evolve in the coming decade?

Three developments could reshape the legal landscape. First, the International Criminal Court’s investigation of operations in Palestine may produce the first international prosecution related to targeted killing, which would create binding precedent on proportionality, necessity, and combatant-status determination. Second, the development of autonomous weapons systems and artificial intelligence for targeting decisions will force the legal community to address questions of human control, accountability for algorithmic errors, and whether existing IHL standards can apply to machine-made targeting decisions. Third, the increasing frequency of state-attributed targeted killings, from Israel’s operations to India’s shadow war to ongoing Russian assassinations, may generate sufficient state practice to crystallize new customary international law norms, either permitting targeted killing under specified conditions or prohibiting it with greater specificity. The direction of evolution is uncertain, but the status quo of deliberate ambiguity is increasingly unstable as more states acquire the capability and the willingness to conduct lethal operations on foreign soil.