Editor’s note: The Reiss Center on Law and Security and Just Security are producing a series, What You Need to Know: Unpacking the Law in Russia’s War Against Ukraine. In brief question-and-answer interviews with leading experts, we probe some of the most urgent and unsettled legal questions in a tragic conflict that threatens to reshape the international legal and political landscape for years to come.

What are the most important ways that international humanitarian law (IHL) guards against food insecurity?

Before answering this question, it is worth noting the context. Russia’s invasion of Ukraine is a war of aggression in violation of article 2(4) of the United Nations (U.N.) Charter and customary international law. Aggressive war is also an international crime (Rome Statute of the International Criminal Court (Rome Statute), article 8bis). These are continuing violations, which is to say that the perpetuation of the illegal war is part of the aggregate jus ad bellum violation, even when its specific components do not violate IHL. Along these lines, the Human Rights Committee has reasoned (correctly in my view) that all killings in an aggressive war violate the right to life. Logically, the same principle can be extended to components of the illegal war that impair other rights, including those relating to food security.

Having said that, the focus of our exchange is IHL, which also has several rules that are critical to limiting the effect of armed conflict on food security. Those rules apply equally to both parties to the conflict. Moreover, unlike the crime of aggression, which attaches only to a narrow category of persons in leadership roles, IHL applies to individual participants throughout the command chain. Serious violations can qualify as war crimes.

Several IHL rules are tailored to preserving food security in armed conflict. Additional Protocol I (1977) – applicable here as Ukraine and Russia are among the treaty’s 174 parties – prohibits the starvation of civilians as a method of warfare, provides enhanced legal protection to “objects indispensable to the survival of the civilian population” (including food and food infrastructure), and regulates humanitarian access to populations in need (including with reference to consent to humanitarian action and the facilitation of access when consent is granted). On the applicability of these rules to Russian actions earlier in the war, see here and here. Geneva Convention IV (1949) – also applicable to the conflict in Ukraine and ratified by 196 states – requires occupying powers, such as Russia in several Ukrainian regions, to ensure that the occupied population is supplied with food and other essentials, including by bringing resources in and, where supplies remain inadequate, granting humanitarian access. Geneva Conventions III and IV require detaining authorities to ensure food rations of sufficient quantity, quality, and variety to keep detainees in good health and to prevent weight loss or nutritional deficiencies. With the requisite intent, starvation of civilians as a method of warfare qualifies as a war crime ( Rome Statute, article 8(2)(b)(xxv)).

Compliance with the core rules of distinction, proportionality, and precautions are also important to limiting the impact of war on food security. For example, even when food is not targeted, belligerents are required to take all feasible precautions to limit attacks’ incidental damage to civilian food and food systems as well as incidental civilian injury or death. The attack may not go ahead if those expected incidental impacts would be excessive in relation to the concrete and direct military advantage anticipated. Damage to food and food infrastructure should weigh heavily in the latter analysis, given its indispensability to human survival. Although not their primary focus, the heightened protections accorded to dams, dykes, and nuclear plants, as well as the rules protecting the environment in armed conflict also help to protect water and food systems.

Since terminating the Black Sea Grain Initiative (BSGI), Russia is reported to have engaged in multiple attacks on food and food export infrastructure in Ukraine’s Black Sea ports and on alternative food shipping routes. What does IHL have to say about those attacks?

​​The attacks on food and food-specific infrastructure very likely violate the prohibition on targeting civilian objects and almost certainly violate the framework of heightened protection accorded to objects indispensable to civilian survival. The legality of attacks on general export infrastructure is harder to ascertain without more information, but there are at least questions as to its compatibility with the requirements of distinction, precautions, and proportionality.  

Pursuant to the foundational IHL rule of distinction, belligerents may only target objects that, by their nature, location, purpose, or use, make an effective contribution to military action, such that their destruction, capture, or neutralization would return a definite military advantage. Any food or export infrastructure that does not satisfy those criteria is protected as a civilian object. Targeting such an object would be unlawful. Done with the requisite level of intent, it would be a war crime (Rome Statute, article 8(2)(b)(ii)). If there is doubt as to its civilian status, it should be presumed protected.

Ordinarily, an object that has a civilian use, but also contributes militarily pursuant to the standard just mentioned – a “dual-use object” – would qualify as a military objective. As such, it can be targeted, as long as the rules mentioned above regarding proportionality and precautions are satisfied. However, as I will explain, food is subject to heightened protection.

The most likely argument in defense of Russia’s attacks on grain stores and food-specific infrastructure is that Ukrainian food exports provide an effective military contribution through sustaining the war effort economically. This argument is not persuasive.

The United States asserts that “war-sustaining” objects qualify as military objectives (Department of Defense Law of War Manual §§, For the most comprehensive defense of that view, see Goodman. However, this assertion has not gained traction with most states and is rejected by the majority of experts, for whom a tighter connection to military action is required (e.g. Dinstein, pp.126-127). For that reason, in the absence of a more proximate military contribution, a strong argument can be made that the strikes are straightforwardly illegal as attacks on civilian objects. 

Even if one were to assume the war-sustaining theory of military objectives in general, that theory could not support attacks on food or food-specific infrastructure, such as grain silos. First, as objects indispensable to civilian survival, food and food-specific infrastructure cannot be targeted for their sustenance value unless they provide sustenance exclusively to combatants (Protocol I, article 54(2), 54(3)(a)). To the extent that Russia has targeted food and food infrastructure to put pressure on global (civilian) food supplies and thereby elicit an alleviation of certain economic sanctions, this targeting of indispensable objects for their sustenance value would be prohibited. Second, food and food infrastructure can only be targeted for reasons other than sustenance value if they provide “direct support to military action,” and even then, only if the targeting does not cause civilian starvation (Protocol I, article 54(3)(b)). Whatever one’s view on the war-sustaining theory of military objectives, general support to the economy is clearly not direct support to military action. There is also reason to believe that the elimination of Ukraine’s capacity to supply the world with grain will lead to starvation in areas of the world with greatest need. However, given the relative complexity of that causal link (discussed below), the lack of direct support to military action provides a more straightforward basis for the strikes’ illegality. Any Russian invocation of alleged Ukrainian IHL violations to defend its attacks as lawful reprisals would be irrelevant here, as food and food infrastructure may not be the target of reprisals (Protocol I, article 54(4)).

In contrast, the general export infrastructure of Ukraine’s Black Sea ports is not likely to qualify as indispensable to civilian survival. Therefore, it would be analyzed under the ordinary rule of distinction. Although maritime ports are often characterized as military objectives (Dinstein, p.142), this must be assessed on a case-by-case basis. Certainly, port infrastructure that is being (or is going to be) used for channeling military supplies or launching military operations qualifies as a military objective by use or purpose, without any need to resort to the war-sustaining theory. However, it is doubtful that this describes the infrastructure targeted since the termination of the BSGI. Depending on their strategic criticality, it is also possible for certain ports or port elements to qualify as military objectives by location. However, even assuming some of Ukraine’s port infrastructure might qualify as military objectives on one of these grounds, it would be unlawful to attack without satisfying the requirements of proportionality and precautions. Those, of course, are highly fact-specific evaluations, some of the details of which I’ll revisit below.

In terminating the BSGI, Russia has indicated that it will now obstruct the passage of food from Ukraine to the world and has stated that “all ships en route to Ukrainian ports in the Black Sea will be considered as potential carriers of military cargo.” Is its new posture compatible with IHL? Given the critical role of Ukraine as a supplier of global nutrition, is Russia “weaponizing food,” as some have argued?

It is important to emphasize at the outset that Russia’s warning to vessels sailing in the waters of the Black Sea to Ukrainian ports cannot, under any circumstances, create a “free-fire zone.” Nor, for that matter, can Ukraine’s tit-for-tat response threatening ships en route to Russian ports. Beyond that, an IHL analysis of Russia’s operations in the Black Sea depends on whether it is engaged in a legal blockade under the law of naval warfare. Parenthetically, to revisit a point made at the outset, a blockade lacking jus ad bellum basis is an enumerated act of aggression in both the Rome Statute definition of the crime of aggression (article 8bis(2)(c)) and the UN General Assembly Definition of Aggression (article 3(c)). This should not be lost in the IHL analysis that follows.

To qualify as a blockade, Russia’s operation must have been declared with specificity as to the starting time, geographical limits, and period for neutral vessel exit (San Remo Manual, paras 93-94; Newport Manual, § 7.4.2). The San Remo Manual commentary notes that this would ordinarily include a Notice to Mariners and communication to the International Maritime Organization (IMO) (pp. 172, 177). The London Declaration concerning the Laws of Naval Warfare (1909) specifies direct notification to neutral states and local authorities (article 11). It is at least questionable that the Russian Ministry of Defense message on Telegram specifying that “all vessels sailing in the waters of the Black Sea to Ukrainian ports will be regarded as potential carriers of military cargo” after 00.00 Moscow time on Jul. 20, 2023 was addressed or specified in the ways necessary to satisfy these criteria. No declaration of blockade appears to have been circulated via the IMO. Additionally, to qualify as a legal blockade under the law of naval warfare, the encirclement must be enforced effectively and impartially on ships of all nations (San Remo Manual, paras 95, 100; Newport Manual, §§ 7.4.3, 7.4.4). If it does not satisfy those requirements, Russia’s operation would not gain the broader law-of-blockade permissions regarding the capture or targeting of neutral merchant vessels carrying Ukrainian exports.

Assuming the blockade criteria are not met, Russia would have the right to visit and search neutral merchant vessels, at least when it has a reasonable suspicion that there are grounds for the vessels’ capture, such as that they are carrying arms or contraband (although the need for such suspicion is disputed) (San Remo Manual, para 118; Newport Manual, § 9.9). However, in the absence of a blockade, it could not subject those vessels to capture simply by virtue of their carrying Ukrainian food exports (San Remo Manual, paras 147, 150(c); Newport Manual, § Moreover, assuming such vessels are merely carrying those exports (and are not under Ukrainian control or direction, or otherwise supporting the war effort, such as through providing intelligence), they would not become targets, except by actively resisting lawful Russian efforts to visit and search (San Remo Manual, para. 67; Newport Manual, § 8.6.5). Finally, although Russia may warn ships away from specified areas when militarily necessary and consistent with rights of neutral navigation, vessels that travel into such areas are not transformed into lawful military objectives and do not lose their protection from attack (Newport Manual, § 

In contrast to neutral ships, Ukrainian merchant vessels could be captured and condemned under the law of prize, granting Russia full property rights (San Remo Manual, para 135; Newport Manual, §§ 9.1, 9.4) – a legal reality that Andrew Clapham has persuasively criticized as outdated, particularly (although not exclusively) as applied to an aggressor. Even assuming the general applicability of the law of prize, there is a question as to its application to vessels carrying food. The San Remo Manual provides that the exemption of food from this regime applies only pursuant to the prior consent of the belligerents (San Remo Manual, para 136(c)(ii)). However, a good argument could be made that the prohibition on starvation of civilians as a method of warfare places an additional limit on Russia’s authority under the law of prize (see below on blockades). In any event, Ukrainian merchant vessels’ liability to condemnation as prize would not make them lawful targets of attack, unless they were to engage in other actions, such as actively resisting visit, search, or capture (San Remo Manual, para 60(e); Newport Manual, § 8.6.3).

If Russia’s operation were to qualify as a blockade in legal terms, the key change would be to the situation of neutral merchant vessels. On a widely held view, an attempt to breach the blockade would ordinarily render a neutral merchant vessel liable to capture and condemnation or diversion (San Remo Manual, para 146(f); Newport Manual, § 7.4.7, 9.10; but see Clapham pp.1255-1258). Moreover, were such a vessel to attempt to resist visit or capture following a warning, it could qualify as a lawful target under the law of naval warfare (San Remo Manual, para 67(a); Newport Manual, §§ 7.4.7, 8.6.5, 9.11; again, but see Clapham pp.1255-1258).

However, even under the law of blockade, certain humanitarian protections apply. There is broad consensus, even in relatively conservative restatements, that a blockade cannot be issued with the sole or primary purpose of starving civilians or depriving them of objects indispensable to their survival, such as food (San Remo Manual, para. 102; Newport Manual, § 7.4.5). Such obstruction would violate the prohibition of starvation of civilians as a method of warfare – the legal category most proximate to the notion of “weaponizing food,” which is not itself a legal term of art (San Remo Manual commentary, p. 179). Here, Russia appears to be blocking food to coerce third states into alleviating sanctions, which could mean that the blockade’s primary purpose is indeed to deprive civilians of the quintessential object indispensable to survival. However, if the primary purpose of the blockade is instead to strangle Ukraine’s economy – one of the traditional purposes of blockade – with the denial of food a secondary or collateral consequence, there is more scope for debate.

I have argued elsewhere that the ban on starvation of civilians as a method of warfare precludes the deprivation of food by blockade pursuant to the same rules that govern the protection of food from attack (see above). Among other things, this would mean that the purposive denial of food to a civilian population would be prohibited whether or not that is the primary or sole purpose of the blockade. However, even in adopting the latter (narrower) prohibition, the influential San Remo Manual precludes establishing a blockade that would inflict disproportionate civilian damage and requires the blockading party to grant passage to food when the “civilian population of the blockaded territory is inadequately provided” with it (pursuant to certain technical arrangements) (paras 102-103). The more recent Newport Manual takes an even narrower view than the San Remo Manual on this point, denying the applicability of proportionality or any duty to allow the passage of food to civilians (§ 7.4.5). In my view, the latter position is incompatible with the ban on starvation of civilians as a method of warfare.

Whichever of these approaches prevails, a complicating factor in applying the humanitarian protections of blockade law to the current case is that the paradigm on which legal analysis has focused is that of cutting off an encircled population from external supply. This is apparent most obviously in the San Remo requirement to allow the passage of food when the blockaded population is inadequately supplied. In Russia’s blockade of Ukraine’s Black Sea ports, the prominent danger is the other way around. The blockaded region is the food supplier. In addition to raising a question as to the applicability of the legal protection, this complicates analyzing the blockade’s impact on affected (but not blockaded) populations. Ukraine’s importance as a grain supplier is such that Russia’s actions will surely exacerbate severe food crises in several regions of the world. However, this is not a straightforward impact when viewed through the narrow lens of IHL.


On that point, does it matter in this case that the civilians who would be affected by the lost grain (as distinct from those who would be affected by the lost revenue) are primarily the nationals and residents of states that are not party to the conflict? Does the answer to this question differ depending on whether the question relates to attacks on food or to the obstruction of efforts to bring food to market?

The operations under analysis are clearly occurring within the geographic area of the ongoing armed conflict between Russia and Ukraine. In my view, the nationality and location of those who will endure exacerbated food insecurity due to Russia’s actions since the termination of the BSGI does not affect the legal analysis. However, the related issues of causation and foreseeability might have an impact, particularly insofar as the analysis turns either on proportionality or on whether actions were taken despite the expectation that civilians would be left in starvation conditions.

Under Geneva Convention IV, many of the provisions protecting civilians are contingent on their being nationals of the adversary. Even those of its provisions not subject to that limitation apply only to the “populations of the countries in conflict.” In contrast, Additional Protocol I – the source of the key rules under consideration here – is more capacious in its protection, defining civilians to include anyone who is not a combatant, without regard to nationality or territory. The key limitations are instead that IHL applies only to actions associated with an armed conflict (Sassòli, pp. 200-203) and the scope of its protection is often contingent on foreseeability and clarity of causation (Gillard, paras. 37-60). Together, these are important premises in the current context.

Regarding the illegality of attacking food for its sustenance value, the belligerent nexus is clear and the violation attaches to the reason for the attack and the nature of the objects, not the identity or distance of those dependent upon them. The same can be said of attacks on civilian objects on land (port infrastructure) or at sea (merchant vessels), as well as attacks on food not providing “direct” non-sustenance support to military action. Again, the relative remoteness of the civilians dependent on those objects is legally irrelevant. One or another of these categories likely characterizes most of the attacks on food and food systems following the BSGI’s termination.

However, for attacks that do not violate those requirements, the complexity of the causal chain between the attack and the harm to civilians suffering food insecurity makes the latter’s impact on IHL illegality less clear-cut. Per the analysis above, such a determination would rely either on an expectation that the attack would cause starvation or that it would cause disproportionate civilian death, injury, or destruction of objects. In my view, these assessments must incorporate the harm associated with the unavailability of that portion of food that would have been allocated directly to those in need (including, most obviously, the 725 thousand metric tons of grain channeled through the World Food Programme to Afghanistan, Djibouti, Ethiopia, Kenya, Somalia, Sudan, and Yemen during the functioning of the BSGI). However, much of the food security impact of the attacks is likely to occur through the intervening factors of global market forces and price inflation (combined with the policies and practices of other key food suppliers) – a causal route that may be too complex for the ultimate impacts to play a role either in IHL proportionality analysis as commonly understood, or in an analysis of whether the attacks are likely to leave civilians so inadequately supplied as to cause starvation (itself a high threshold).

A distinct complication may be thought to arise in any analysis of whether the obstruction of grain through encirclement would implicate the ban on starvation of civilians as a method of warfare. In my view, any denial of food that is purposeful or includes food destined for civilians who may otherwise be expected to face starvation or displacement implicates the crux of the ban. However, it might be suggested that denying food to external populations to strong-arm non-belligerent states into alleviating economic sanctions is too tangential to the armed conflict to be classified as a “method of warfare.” Analyzing the current situation, I take a different perspective. The use of blockade or other tools of naval warfare is itself shaped by and dependent upon the fact of armed conflict. Here, if the goal is to elicit sanctions relief, those methods are being used purposefully to deny food to civilian populations with a view to facilitating progress in the war by reducing one of its key costs (the attached sanctions). This ought to be understood straightforwardly to qualify as using food deprivation as a method of warfare. The location and identity of the affected civilians should not obscure that reality.