Attacking Scientists and the Law of Armed Conflict

On June 13, 2025, Israel launched Operation Rising Lion, an attack on Iran’s nuclear infrastructure, ballistic missile program, and senior military leadership. Additionally, the Israel Defense Forces (IDF) conducted so-called “targeted killings” of scientists working on Iran’s nuclear development program. According to the IDF,

All of the eliminated scientists and experts, eliminated based on intelligence, were key factors in the development of Iranian nuclear weapons. Their elimination is a significant blow to the regime’s ability to acquire weapons of mass destruction.

The attack came on the day after the International Atomic Energy Agency (IAEA) Board of Governors issued a Resolution finding Iran in non-compliance with its obligations under the Treaty on the Non-Proliferation of Nuclear Weapons and the associated Comprehensive Safeguards Agreement. The resolution noted that Iran had not cooperated with the Agency regarding “undeclared nuclear material and activities at multiple undeclared locations.” That being so, the IAEA could not “verify that there has been no diversion of nuclear material required to be safeguarded under the Agreement to nuclear weapons or other nuclear explosive devices.”

In fact, a May IAEA report noted that Iran has been actively enriching uranium to a level of 60%, close to the 90% needed to produce a nuclear weapon. Its stockpile stood at 400kg, which is enough for ten nuclear warheads. In the aftermath of the Israeli attacks, there have also been media reports that Iranian scientists had successfully conducted experiments on the assembly of a nuclear weapon that would allow Iran to produce a weapon in a matter of weeks. Apparently, the scientists were divided “into several working groups to labor in secret on components of the process of weaponizing nuclear material into an actual explosive device.” The effort is said to have begun after Hamas’s October 7, 2023, attacks and ran in parallel to the ongoing enrichment program, all under the watchful eye of the IDF.

These were not the first nuclear scientists eliminated by Israel. Indeed, Israel killed Mohsen Fakhrizadeh, the so-called “father” of Iran’s nuclear weapons program, with a robotic machine gun in November 2020. Before that, it had taken out numerous other nuclear scientists, including via the use of attackers on motorcycles who affixed magnetic bombs to cars.

In an earlier post this week, I offered my views on Operation Rising Lion’s jus ad bellum framework. In this post, I am taking on the law of armed conflict (LOAC) issue of the IDF’s targeting of Iranian nuclear scientists. Of course, legal analysis is always context and fact-dependent, and I have no access to the intelligence upon which the IDF determined the scientists were lawful targets. Therefore, I will deal with the subject in the abstract, assuming for the sake of analysis that the scientists in question are significant players in the rapid development of a nuclear weapon by a party to an international armed conflict.

The Applicable Law in the Abstract

The 1977 Additional Protocol I to the 1949 Geneva Conventions (AP I) codifies the LOAC rules governing the “conduct of hostilities” during an international armed conflict. Although neither Iran nor Israel is party to that instrument (Iran is only a signatory), most of its provisions on targeting are considered broadly reflective of customary international law that binds all States (see my discussion here). International human rights law (IHRL), such as the prohibition on arbitrary deprivation of life, also applies during an armed conflict, but the LOAC rules under consideration in this post, as lex specialis, inform its application (International Court of Justice (ICJ), Nuclear Weapons Advisory Opinion, para. 25). In effect, they are the applicable law.

The foundational principle undergirding the conduct of hostilities rules is “distinction,” a customary law principle codified in Article 48 of Additional Protocol I for States parties. With regard to persons, it requires parties to a conflict to “at all times distinguish between the civilian population and combatants.” This principle has been operationalized in the customary law and Additional Protocol I prohibition on making civilians the “object of attack” (International Committee of the Red Cross (ICRC), Customary International Humanitarian Law study, rule 1; AP I, art. 51(2), U.S. Department of Defense (DoD) Law of War Manual, § 5.5.2). The prohibition does not extend to scientists who are members of the armed forces or an organized armed group (DoD Law of War Manual, § 5.7.2). This is so, irrespective of their duties; they are therefore targetable based on their status as such, around the clock, unless rendered hors de combat.

According to Article 51(3) of Additional Protocol I, civilians lose this protection from attack “for such time as they take a direct part in hostilities.” Such participation also results in the individuals concerned no longer being treated as civilians in proportionality calculations and feasible precautions in attack assessments. There is widespread agreement that the loss of protection is customary in character, thereby binding non-party States like Israel, Iran, and the United States (ICRC, Customary International Humanitarian Law study, rule 6; DoD Law of War Manual, § 5.8).

But the devil is in the details. To better understand the notion of direct participation, the ICRC convened expert meetings over five years to consider what that concept meant in practice. I was one of the participants. Unfortunately, we could not agree on several sensitive issues, so the ICRC issued its own interpretation of the rule. The resulting 2009 Interpretive Guidance on the Notion of Direct Participation in Hostilities identified three cumulative “constitutive elements” qualifying an act as direct participation, in what I believe to be an invaluable contribution to understanding direct participation.

1. The act must be likely to adversely affect the military operations or military capacity of a party to an armed conflict or, alternatively, to inflict death, injury, or destruction on persons or objects protected against direct attack (threshold of harm), and

2. There must be a direct causal link between the act and the harm likely to result either from that act, or from a coordinated military operation of which that act constitutes an integral part (direct causation), and

3. The act must be specifically designed to directly cause the required threshold of harm in support of a party to the conflict and to the detriment of another (belligerent nexus).

Despite broad, and justified, agreement on the elements, consensus as to how they should be interpreted in particular situations is lacking, as will be explained below (see NYUJILP, vol. 42.3). Whatever the correct interpretation, however, if a civilian’s actions satisfy all three elements, the individual is targetable “for such time” as they so participate, a concept that has itself provoked disagreement (my view coincides with DoD Law of War Manual, § 5.8.4).

In addition to Article 51(3) of Additional Protocol I, the notion of direct participation also appears in Common Article 3 to the 1949 Geneva Conventions, which provides, “Persons taking no active part in the hostilities … shall in all circumstances be treated humanely.” Humane treatment excludes, by the terms of the article, “violence to life and person.” By implication, those taking an active part in hostilities are subject to attack. Although Common Article 3 was originally included to address only “conflicts not of an international character,” international tribunals have since interpreted the provision as reflecting a “minimum yardstick” applicable in all armed conflict (ICJ, Paramilitary Activities, para. 218; ICTY Appeals Chamber, Tadić, para. 94).

Applying the Elements to Nuclear Scientists

Threshold of Harm

It is self-evident that the employment of a nuclear weapon is “likely to adversely affect the military operations or military capacity” of the enemy, thereby satisfying the “threshold of harm” element. But harm is a much broader concept than merely attacking the enemy. As noted by the ICRC, “military harm should be interpreted as encompassing not only the infliction of death, injury, or destruction on military personnel and objects, but essentially any consequence adversely affecting the military operations or military capacity of a party to the conflict” (Interpretive Guidance, p. 47).

This being so, the mere possession of a nuclear weapon would almost always satisfy the harm criterion. After all, the enemy’s possession of a nuclear weapon would exert significant influence on the opposing party’s strategic, operational, and even tactical-level military decision-making, for great care would have to be taken to avoid operations that might trigger its use. At the same time, a nuclear weapon gives the side possessing it greater operational leeway. It can afford more missteps because the weapon limits the enemy’s options for exploiting bad decisions. This dynamic was well-illustrated by Russia’s brazenness in conducting unlawful attacks on Ukrainian population centers and Ukraine’s hesitancy in conducting robust attacks into Russian territory, even though it is undeniably entitled to do so.

But what if the enemy has not yet acquired the capacity, as is the case with Iran? In this regard, the Interpretive Guidance observes,

The qualification of an act as direct participation does not require the materialization of harm reaching the threshold but merely the objective likelihood that the act will result in such harm. Therefore, the relevant threshold determination must be based on “likely” harm, that is to say, harm which may reasonably be expected to result from an act in the prevailing circumstances (p. 47).

Scientific contributions are a sine qua non factor in developing a nuclear weapon. And once fielded, the adversary’s military operations will, as explained, inevitably be adversely affected. Accordingly, there should be no question that the first constitutive element of direct participation is satisfied, so long as the individual is, in fact, involved (see Schmitt/Schauss on certainty vis-à-vis targets).

Belligerent Nexus

As understood in the context of direct participation, belligerent nexus denotes an act that is “specifically designed to [inflict harm satisfying the other elements] in support of a party to an armed conflict and to the detriment of another” (Interpretive Guidance, p. 58). Thus, the fact that a scientist is conducting research that might be useful in acquiring a nuclear weapon is insufficient, in itself, to render that person a targetable direct participant. This is so even if the research does factually contribute to its development. Instead, the research and related activities must be carried out in order to produce or employ such a weapon; only then will this constitutive element be satisfied.

Of course, if a scientist is knowingly participating in a nuclear weapons development program or otherwise conducting research to facilitate the acquisition and use of a nuclear weapon, belligerent nexus exists. But what of a scientist who is forced to participate in a nuclear program or does not realize the purpose to which the research will be put? The Interpretive Guidance suggests, correctly, in my view, that what matters is the “design of the act or operation” and not “the mindset of every participating individual” (p. 59). Thus, a scientist who is coerced into nuclear weapons research remains a direct participant in the same way that someone coerced into gathering intelligence for the enemy is directly participating.

However, the ICRC is also of the view that belligerent nexus is not satisfied if the person involved is “totally unaware of the role they are playing in the conduct of hostilities” (p. 59). It is difficult to imagine how a scientist contributing to nuclear weapons design and production would not appreciate the role they are playing. Still, I tend to agree with the ICRC. As I wrote in a 2005 paper for one of the expert meetings, “It is not necessary that the individual foresaw the eventual result of the operation, but only that he or she knew his or her participation was indispensable to a discrete hostile act or series of related acts” (p. 532).

Direct Causation

It is with respect to the constitutive element requiring a direct causal link between the activity in question and the adverse effect on enemy military operations or capacity that most interpretive play exists. In this regard, there is broad agreement that acts contributing to the “general war effort” are too attenuated to qualify; that relationship is instead one of “indirect” causation. The Interpretive Guidance defines the general war effort as “activities objectively contributing to the military defeat of the adversary (e.g., design, production and shipment of weapons and military equipment, construction or repair of roads, ports, airports, bridges, railways and other infrastructure outside the context of concrete military operations)” (p. 51).

Among the experts taking part in the ICRC’s direct participation project, there was consensus on this point. For instance, we discussed workers in a munitions factory at some length, agreeing that they were not direct participants and, therefore, could not be attacked (e.g., while on their way to work). Additionally, they counted as civilians during proportionality calculations and feasible precautions in attack assessments (Interpretive Guidance, fn. 123).

As we drilled deeper, however, interpretive fault lines appeared. One was the ICRC’s contention that “direct causation should be understood as meaning that the harm in question must be brought about in one causal step.” As examples of activities that did not meet this standard, it proffered “scientific research and design, as well as production and transport of weapons and equipment unless carried out as an integral part of a specific military operation designed to directly cause the required threshold of harm” (Interpretive Guidance, p. 53).

This evoked concern among a significant number of the experts, especially in the context of applying the interpretation to the assembly of improvised explosive devices (IED). The ICRC argued that the activity, “unlike the planting and detonation of that device, [does] not cause that harm directly” (p. 54). Yet, by the time of the meetings, IEDs, which were often produced in the very areas where they would soon be used, were causing more coalition casualties in Afghanistan and Iraq than any other weapon.

Unsurprisingly, therefore, the experts with military experience firmly rejected the premise that IED assemblers could not be attacked while away from the location (e.g., a garage or basement) at which IEDs were being assembled or that they counted as civilians in the requisite proportionality calculation and feasible precautions assessment. For us, the production and assembly of an IED is integral to its imminent use, unlike the production of munitions far from the battlefield that might never find their way there. We believed, for instance, that temporal and geographic proximity were useful considerations in assessing those assembling IEDs, although the ICRC did not (Interpretive Guidance, p. 55). Context mattered to us.

Nuclear weapons development arose analogously while addressing the subject, albeit not with the fervency surrounding IED assembly. Indeed, the Interpretive Guidance mentioned that discussion in a footnote (n. 121) that, in the past few days, has loomed large in debates over the targeting of Iranian scientists.

Although, during the expert meetings, civilian scientists and weapons experts were generally regarded as protected against direct attack, some doubts were expressed as to whether this assessment could be upheld in extreme situations, namely where the expertise of a particular civilian was of very exceptional and potentially decisive value for the outcome of an armed conflict, such as the case of nuclear weapons experts during the Second World War.

This accurately characterizes the discussions, during which the experts took diverse approaches to the issue (see Summary Report, Fourth Meeting, p. 48-50). My recollection is that the suggestion that all weapons are not equal and that context matters found a fair degree of traction; I certainly found it compelling, although, admittedly, not everyone did.

The other American participant in the direct participation project was my friend, the late Hays Parks, who had written on the subject in his classic 1990 piece, Air Law and the Law of War. Taking a famously comprehensive look at State practice, Parks distinguished civilians supporting the “war effort” from those involved in the “military effort” (p. 133). He argued that

[C]ivilians working directly towards the military effort—and this would include those scientists employed on the various research projects such as radar, rockets, etc., during World War II—by the nature of their duties are far more combatants than civilians. Their civilian status should not be permitted to jeopardize the safety from attack of the innocent civilian, whose numbers make up the vast majority of any nation’s population (p. 132).

For Parks, the “work of some ‘civilians’ has become so critical to military success that those individuals are civilians in name and garb only.” As he explained in a well-known 1989 legal memorandum written while serving as Chief of the Army’s International Law Branch, “One rule of thumb with regard to the likelihood that an individual may be subject to lawful attack is his (or her) immunity from military service if continued service in his (or her) civilian position is of greater value to a nation’s war effort than that person’s service in the military. A prime example would be civilian scientists occupying key positions in a weapons program regarded as vital to a nation’s national security or war aims.”

Whereas everyone agrees that those involved in the war effort are protected from attack under both customary international law and Additional Protocol I, Parks believed civilians performing work critical to military success enjoyed that protection only under the Protocol during operations by parties to the treaty. This is because he took the position that Article 51(3) of Additional Protocol I is not reflective of customary law in its entirety. This remains the U.S. position today (DoD Law of War Manual, § 5.8.1.2)

In 2004, I proposed a similar test that focused on “the criticality of the act to the direct application of violence against the enemy.” By it, “an individual performing an indispensable function in making possible the application of force against the enemy is directly participating. In other words, the appropriate test is whether that individual is an integral facet of the uninterrupted process of defeating the enemy” (see here, p. 505, 525; see also Shamir-Borer, p. 967-71; Mc Donald, p. 11117-18). Whether that process involves employing a nuclear weapon or possessing one to frustrate the enemy’s operational flexibility (or enhance one’s own), the significance of the activity to military operations is the central factor in the assessment of the causality element.

Such an assessment lies along a rather vague continuum and is influenced by a wide variety of factors. As I observed in 2005,

[D]irect participation determinations are necessarily contextual, typically requiring a case-by-case analysis. But case-by-case determinations need not be eschewed. On the contrary, sometimes they more precisely balance military requirements and humanitarian ends than mechanical applications of set formulae (p. 534).

It is an approach consistent with that of the International Criminal Tribunal for the former Yugoslavia, which in the 1990 Tadić judgment held that it is “unnecessary to define exactly the line dividing those taking an active part in hostilities and those who are not so involved. It is sufficient to examine the relevant facts of each victim and to ascertain whether, in each individual’s circumstances, that person was actively involved in hostilities at the relevant time” (para. 616).

In the absence of a bright line test for assessing the causal element of direct participation, doubt will often surface. I tackled that issue as well in 2005. Two decades later, I stand by what I said then, especially given that the question is the development of nuclear weapons.

Many activities lie between the extremes. In such cases, the methodology that best approximates the underlying intent of the direct participation notion is to interpret the term liberally, in other words, in favor of finding direct participation. … An interpretation of direct participation that allows civilians to retain their immunity even though inextricably involved in the conduct of ongoing hostilities will engender disrespect for the law on the part of combatants endangered by their activities. Moreover, the liberal approach provides an incentive for civilians to remain as distant from the conflict as possible because they can thereby avoid being directly targeted … . While broadly interpreting the activities that subject civilians to attack might seem counterintuitive from a humanitarian perspective, it actually enhances the protection of the civilian population as a whole by encouraging distance from hostile operations (p. 534-35).

Concluding Thoughts

When events during an armed conflict capture the international community’s attention, there is often a precipitous rush to legal judgment. In some cases, that is appropriate, as in condemning direct attacks on civilians. Yet, the fact that LOAC is a body of law seeking to balance military and humanitarian considerations means that there is a baked-in margin of appreciation designed to accommodate the context in which the rules will apply. And there are few LOAC rules, the interpretation of which has generated more disagreement among States and scholars than that addressing direct participation in hostilities. It is a thorny subject, even in the abstract.

Accordingly, determining whether a particular nuclear scientist is participating in the hostilities is a complex endeavor that must consider an array of factors. How advanced is the nuclear weapons development program? How intense is the current development activity? To what extent will the possession of a nuclear weapon be a game changer at the tactical and operational level of war? Is the individual part of an established nuclear weapons development program or someone conducting general research that the program finds useful? What is the individual doing, and how critical is that activity to the program’s overall development? Are the individuals concerned aware of the nature of their research? Is delivery of the weapon viable? And so on. The constitutive elements proposed by the ICRC, considered cumulatively, are the appropriate fulcrum on which the assessment of these and many other factors must turn.

So, my very unsatisfactory answer to the question of whether the IDF attacks targeted directly participating Iranian scientists and, as such, were lawful is that I do not have enough facts. Based only upon open source material available today, I lean in the direction of the attacks complying with LOAC. Yet, I am certain a fair number of scholars and practitioners I respect will come down on the other side of the issue. The law of armed conflict is often an uncomfortable environment for those seeking clarity or certainty.

***

Michael N. Schmitt is the G. Norman Lieber Distinguished Scholar at the United States Military Academy at West Point. He is also Professor of Public International Law at the University of Reading and Professor Emeritus and Charles H. Stockton Distinguished Scholar-in-Residence at the United States Naval War College.

The views expressed are those of the authors, and do not necessarily reflect the official position of the United States Military Academy, Department of the Army, or Department of Defense.

Articles of War is a forum for professionals to share opinions and cultivate ideas. Articles of War does not screen articles to fit a particular editorial agenda, nor endorse or advocate material that is published.

 

 

 

 

 

Photo credit: Avash Media