1. Introduction
On 7 October 2023, Hamas launched a horrendous attack on Israel. The Israel Defense Forces (IDF) have been responding with Operation ‘Iron Swords’, consisting of massive air strikes and ground offensives in the Gaza Strip. More than two years later, there is good reason to hope that the hostilities have come to a more than intermittent end.Footnote 1 Overall, the ensuing international legal debate has been focusing more on the jus in bello than on the jus contra bellum. This observation first applies to the manner in which Israel has tried to justify its course of action in international legal terms: Israel’s letters of 7 October 2023 addressed to the United Nations Secretary-General and the President of the UN Security Council (UNSC) fall short of an explicit reference to Article 51 of the UN Charter,Footnote 2 and instead only announce that ‘(t)he state of Israel will act in any way necessary to protect its citizens and sovereignty from the ongoing terrorist attacks originating from the Gaza Strip and carried out by Hamas and other terrorist organizations’.Footnote 3 In a paper issued by the Ministry of Foreign Affairs in 2023, in order to provide an overview of ‘key legal aspects of the hostilities’ from the perspective of the state of Israel, the jus contra bellum also received only cursory attention: under the rubric ‘the applicable law’, the jus contra bellum is not even mentioned, and under the rubric ‘Israel’s right and obligation to defend itself’, it is only said that ‘in relative terms, taking into account the size of Israel’s population, the scope of the attack launched on October 7 is equivalent to the 9/11 attacks many multiple times over’, that the ‘threat is both grave and immediate’, and that ‘(i)n these circumstances, Israel is both entitled and compelled to act so as to deny Hamas and other terrorist groups in Gaza the capacity to continue attacking its citizens and territory’.Footnote 4 Second, neither the UN General Assembly (UNGA), nor the UNSC, nor the International Court of Justice (ICJ) have expressed a position on the legal situation under the jus contra bellum. Third, as far as the relatively sparse scholarly coverage of the jus contra bellum aspects is concerned, it seems as if Marko Milanovic set the tone early on when he concluded his legal analysis in the following words: ‘The combined effect of all of this uncertainty is such, in my view, that the jus ad bellum is of little practical use with regard to the war in Gaza’.Footnote 5
As will be set out in this article, Milanovic is right that an international legal analysis of Israel’s use of force in the Gaza war in terms of jus contra bellum presents very significant challenges. This is so, first, because the legal status of the Gaza Strip at the material time makes the Gaza war an extraordinary constellation from the perspective of the jus contra bellum, so that a comprehensive legal analysis of the case inevitably leads along a pathway paved with an entire bundle of serious legal controversies. Second, a difficulty results from the fact that it remains impossible for an outside observer, at the time of writing, to fully and reliably appraise all the facts relevant to reach firm legal conclusions at all steps of the legal analysis. However, these difficulties must not lead to a sidelining of the prohibition of the use of force as a principal yardstick for the international legality of Israel’s use of force. Despite the challenges, it is possible to reach a number of conclusions regarding Israel’s use of force in the light of the jus contra bellum with quite some confidence. Apart from this, a close look at Israel’s use of force in the light of the jus contra bellum helps in describing, with as much precision as possible, the legal controversies concerned, and by doing so it is hopefully possible to contribute to a deeper and hence better understanding of the current state of the jus contra bellum with some of its more important intricacies.
The analysis proceeds in essentially five steps. Following a summary of the relevant factual background (Section 2), some preliminary remarks will be made regarding the open question of Palestine’s statehood under customary law and its (ir)relevance for the jus contra bellum (Section 3). A fairly brief section will then be devoted to the applicability of the prohibition of the use of force to Israel’s military operations in the Gaza Strip since 7 October 2023 (Section 4). Thereafter, it will be asked whether, at the material time, Israel has been facing a situation allowing it to exercise its right of individual self-defence as recognised in Article 51 of the UN Charter (Section 5). The final step of the analysis will consider whether Israel has been acting in pursuit of a legitimate goal of self-defence and whether its use of force was necessary to reach that goal and has remained proportionate to the latter (Section 6).
2. The factual background
The armed wing of Hamas began firing rockets from the Gaza Strip into Israel in the early hours of 7 October 2023.Footnote 6 Concomitantly, armed fighters crossed the border between the Gaza Strip and Israel.Footnote 7 Over 1,200 people were killed, and thousands more were injured on that day, including, in particular, at a music festival near the kibbutz of Re’im and in the kibbutzim of Kfar Azza, Be’eri, Nachal Oz, Nir Oz, and Nirim.Footnote 8 Acts of violence, many of them of an unspeakable level of cruelty,Footnote 9 continued for almost three days before the Israeli military regained control of the situation on 9 October 2023.Footnote 10 On 1 November 2023, Ghazi Hamad, spokesperson for the Hamas-led Palestinian governing body in Gaza, announced that the attacks of 7 October 2023 were just the beginning – vowing to launch ‘a second, a third, a fourth’ attack until the country is ‘annihilated’.Footnote 11 According to the Israeli military, 12,000 rockets had been fired at Israel from the Gaza Strip territory by December 2023.Footnote 12 Since then, there have been reports of repeated rocket strikes against Israel, most of which were intercepted before impact.Footnote 13 Over the course of the attack, Hamas fighters took a total of 251 hostages and abducted them into the Gaza Strip. In support of Hamas, Hezbollah and the Houthi rebels followed suit, also launching attacks on Israel from Lebanese and Yemeni territory, respectively.Footnote 14 Israel’s military response began on 7 October 2023 and, since then, the Gaza Strip has been shelled more heavily than ever before.Footnote 15 By its own account, the Israeli military dropped 6,000 bombs on Gaza between 7 and 12 October 2023 alone.Footnote 16 The population of the Gaza Strip was denied all access to humanitarian relief between 9 and 21 October 2023.Footnote 17 During this time, the Israeli military also launched a ground offensive, during which the northern and southern parts of the Gaza Strip were separated. On 13 October, the IDF issued a widespread warning to civilians, urging 1.1 million people to move into the south within 24 hours, tentatively suggesting that the military campaign against Hamas might remain limited to the northern part of the Gaza territory.Footnote 18 However, aerial bombardment and the ground offensive were subsequently extended southward, causing mass movement in the confined area.Footnote 19 Over the escalation of the conflict, the IDF repeatedly urged the population of the Gaza Strip to move or evacuate within the territory to avoid being caught up in hostile acts.Footnote 20 Above that, the IDF have established certain ‘no-go-areas’ – including a buffer zone with the Israeli border – and fortified the separation between northern and southern Gaza.Footnote 21 While the exact number of casualties has been the subject of persistent debate,Footnote 22 the Hamas-led Ministry of Health reports a total of 67,173 fatalities, 169,780 injuries, and 436,000 housing units damaged or destroyed.Footnote 23 The hostilities have led to a severe deterioration in living conditions within the Gaza Strip.Footnote 24 This humanitarian plight has been exacerbated by continued restrictions on the facilitation of relief, leading the International Criminal Court (ICC) to issue arrest warrants against Prime Minister Benjamin Netanyahu and former Minister of Defence Yoav Gallant, which appear to be based principally on the allegation of starving civilians as a method of warfare.Footnote 25 International organisationsFootnote 26 and state representativesFootnote 27 have also warned of an intensifying humanitarian catastrophe and urged the parties to the conflict to – at least temporarily – agree to a ceasefire. In the case brought against Israel by South Africa on the Application of the Convention on the Prevention and Punishment of the Crime of Genocide in the Gaza Strip, the ICJ issued three orders to indicate provisional measures, requiring Israel to provide access to humanitarian relief for the Palestinian population and, in the third order, to ‘halt its military offensive, and any other action in the Rafah Governorate, which may inflict on the Palestinian group in Gaza conditions of life that could bring about its physical destruction in whole or in part’.Footnote 28 The UNGAFootnote 29 and the UNSCFootnote 30each have also addressed the hostilities in resolutions, calling on the parties to the conflict to agree to a ceasefire. Efforts to negotiate a ceasefire were temporarily successful, and the latter came into effect on 19 January 2025.Footnote 31 The agreement initially consisted of three phases. The first and second required Hamas to set the remaining hostages free in exchange for the release of Palestinian prisoners; the second phase also foresaw the IDF’s withdrawal from Gaza, and the third phase envisaged a plan to reconstruct the Gaza Strip.Footnote 32 However, it appears that further negotiations came to an impasse.Footnote 33 Israel resumed major military operations on 18 March 2025 after having reinstated its policy of blocking humanitarian relief from entering the Gaza Strip on 2 March 2025.Footnote 34 The air strikes conducted by Israel, which reportedly ended the ceasefire, have been described as pre-emptive, based on Hamas' ‘readiness to execute terror attacks, build up force and re-arm’.Footnote 35 Upon the renewed hostilities, Israel’s blockade of all relief supplies lasted for a total of around two months. Slow deliveries were resumed through a private mechanism at first, bypassing experienced aid distributors such as the UN.Footnote 36 Several reports of persons being killed while attempting to collect food at the designated aid points emerged shortly thereafter.Footnote 37 Overall, Gaza became at risk of famine, with widespread malnutrition, disintegrating medical services, and severe shortages of clean water.Footnote 38 Under mounting international pressure in response to these conditions,Footnote 39 Israel signalled willingness to negotiate another temporary truce tied to the release of the remaining hostages, even as it prepared for further large-scale operations.Footnote 40 In this vein, hospitals and aid groups were told to prepare for evacuations ahead of an Israeli capture of Gaza City, reportedly aimed at dismantling the remaining militarised Hamas infrastructure.Footnote 41 Hamas was still believed to hold at least 20 hostages presumed alive in its power.Footnote 42 An Israeli airstrike on 9 September apparently targeted Hamas officials engaged in backchannel negotiations in Qatar, fuelling international grievances with Israel’s conduct in the war.Footnote 43 The next day, several countries – including some historically aligned with Israel – shifted their positions in the UNGA, backing a resolution in support of recognising Palestinian statehood.Footnote 44 Since then, efforts led by the US to reinstate a ceasefire gained particular traction.Footnote 45 In October 2025, the hostilities came to a close. In November 2025, at the time of completing work on this article, the situation on the ground remains fragile.
3. The uncertainty regarding Palestine’s statehood
On 7 October 2023, the Gaza Strip formed part of the territory destined for the realisation of the right to self-determination of the Palestinian people.Footnote 46 While Hamas established control over the Gaza Strip in 2007,Footnote 47 no separate state of Gaza under the government of Hamas was created under international law because the Palestinian population in the Gaza Strip did not decide to break away from the larger Palestinian self-determination territory.Footnote 48 It is, however, a matter of profound controversy whether on 7 October 2023 a Palestinian state, including the Gaza Strip, had already been created in the realisation of the right to self-determination of the Palestinian people.Footnote 49 On the one hand, in 2012, the UNGA granted Palestine the status of ‘non-member observer state’Footnote 50 and, in 2024, it declared, with 143 states in favour, ‘that the State of Palestine is qualified for membership in the United Nations in accordance with Article 4 of the Charter of the United Nations and should therefore be admitted to membership to the United Nations’.Footnote 51 Also, in 2020, the Prosecutor of the ICC considered that Palestine qualified as a state under international law in view of its internationally recognised right to self-determination, the detrimental impact of Israel’s ongoing breaches of international law on the effectivity of the Palestinian Authority’s operative government in the entire territory subject to the right of self-determination, and the high number of bilateral recognitions of Palestinian statehood.Footnote 52 On the other hand, however, on 7 October 2023, the effectivity of the Palestinian Authority’s government over said territory continued to suffer from significant limitations, including its complete inability essentially to govern the Gaza Strip. A Pre-Trial Chamber of the ICC explicitly refrained from considering Palestine a state under general international law in 2021, while no less finding that Palestine is a state party to the ICC Statute.Footnote 53 In 2024, the ICJ, while reaffirming the ‘right (of the Palestinian people) to an independent and sovereign State, living side by side in peace with the State of Israel within secure and recognized borders for both States’,Footnote 54 abstained from determining that a state of Palestine has been created in the realisation of that right.Footnote 55 In the light of this prevailing uncertainty, the existence, on 7 October 2023, of a Palestinian state, including the Gaza Strip, will neither be assumed nor excluded in this study. Proceeding in that agnostic way does not decisively complicate the legal analysis; on the contrary, as will be shown, the legal situation under the jus contra bellum does not depend decisively on the question of Palestinian statehood at the material time.
4. The applicability of the prohibition of the use of force to Israel’s use of force and a brief remark on Hamas’ attack on Israel and the jus contra bellum
This prohibition, which the ICJ has accurately called a ‘cornerstone’ of the UN Charter,Footnote 56 is laid down in Article 2(4) of the Charter and also forms part of customary international law.Footnote 57 The ICJ therefore tends to refer to it as the ‘principle of non-use of force in international relations’.Footnote 58 In the present case, the question arises whether this principle is applicable to Israel’s use of force. This would certainly be so if, on 7 October 2023, the Gaza Strip qualified as part of the territory of a state of Palestine. Yet, as we have just seen, this remains an open question. It is therefore worthwhile to interrogate whether Israel’s use of force in the Gaza Strip also fell within the scope of the prohibition of the use of force, assuming this territory, at the material time, was not subject to the sovereignty of a foreign state. Some have voiced the view that the current jus contra bellum is strictly interstate in nature and, therefore, that the scope of the prohibition extends only to the use of force on the territory of another state or one that otherwise affected external manifestations of such a state.Footnote 59 However, such an interpretation is neither required nor persuasive in view of the fact that Article 2(4) of the UN Charter covers the use of force by states in their international relations, not only if it is carried out ‘against the territorial integrity or political independence of any state’ but also if it occurs ‘in any other manner inconsistent with the Purposes of the United Nations’. It is eminently sensible to assume that the latter formulation includes the use of force by a state on a self-determination territory outside its own borders;Footnote 60 for such an interpretation ensures that the territory – on which the right of a people to create a state in the exercise of its international right to self-determination is intended to be realised – is protected against the forcible intrusion of a foreign state in the same manner as the territory of a state and the population of such a state. It is impossible to point to any legitimate interest of a state not to refrain in the same manner from the use of force on the territory of a foreign self-determination territory, as the same state is bound to do with respect to the territory of a foreign state. This interpretation of the prohibition of the use of force is increasingly accepted in international legal scholarship.Footnote 61 It was also implicitly endorsed by the ICJ in its 2024 advisory opinion in Legal Consequences Arising from the Policies and Practices of Israel in the Occupied Palestinian Territory, Including East Jerusalem, in which the Court applied the prohibition of the use of force to the question of the legality of Israel’s continued military presence in the Occupied Palestinian Territory without recognising Palestinian statehood.Footnote 62 The ICJ approach is in line with the position taken by a large number of states in the course of the proceedings.Footnote 63 The better arguments thus support the view that the jus contra bellum has been applicable to Israel’s use of force in the Gaza Strip since 7 October 2023, even on the assumption that the latter territory was not, at the material time, part of a Palestinian state.Footnote 64
It bears noting that Hamas, being a non-state organisation, is not bound by the jus contra bellum. Following the transnational armed attacks by the non-state organisation Al-Qaeda on 11 September 2001, a debate arose as to whether the scope of application of the prohibition of the use of force under customary international law should be extended to cover violent action of non-state transnational armed groups.Footnote 65 Yet, such a development of customary international law has not taken place.Footnote 66 It is thus impossible to make a persuasive case that Hamas was bound by the prohibition of the use of force under international law on 7 October 2023. Assuming the existence of a state of Palestine on that date, this state would, of course, have been bound by the customary prohibition of the use of force. But as Hamas’ attack on Israel would not have been attributable to such a state and as one would have had to consider the same state unable to prevent that attack,Footnote 67 there would not have been a violation of the prohibition of the use of force by a state of Palestine through the Hamas attack or in connection with it. Rather than constituting an unlawful use of force under international law, this attack on Israel entails conduct that violated the law of armed conflict and may qualify as war crimes, crimes against humanity, and genocide.Footnote 68
5. On the conditions necessary to trigger Israel’s right of individual self-defence
5.1. The lack of clarity in Israel’s position and its reference to an ‘ongoing armed conflict’Footnote 69
Israel has yet to articulate a comprehensive and fully consistent exposition of its own view regarding the legality of its use of force in the Gaza Strip under the jus contra bellum.Footnote 70 In its letter of 7 October 2023, Israel did not refer explicitly to Article 51 of the UN Charter; nor did it specifically mention its right of individual self-defence. Instead, as mentioned at the beginning of this article, it was declared that ‘(t)he state of Israel will act in any way necessary to protect its citizens and sovereignty from the ongoing terrorist attacks originating from the Gaza Strip and carried out by Hamas and other terrorist organizations’.Footnote 71 A little later, Israel stated that the Hamas attack took place in the context of an ongoing armed conflict and that, for this reason, ‘legally speaking, these attacks do not necessitate an analysis of the conditions under which Israel may resort to the use of armed force (the jus ad bellum)’.Footnote 72 During the proceedings before the ICJ in the Case concerning Application of the Convention on the Prevention and Punishment of the Crime of Genocide in the Gaza Strip, Israel, however, invoked its right of self-defence in the following terms: ‘(I)t is in response to the slaughter of 7 October … and to the ongoing attacks against it from Gaza, that Israel has the inherent right to take all legitimate measures to defend its citizens and secure the release of the hostages’.Footnote 73 It is thus not entirely clear whether Israel has, in fact, invoked its right of self-defence, as recognised in Article 51 of the UN Charter, in response to the Hamas attack of 7 October 2023, or whether Israel believes that its right of self-defence was triggered and exercised at some earlier point in time by a prior attack by Hamas, and that this had rendered it unnecessary for Israel to justify its use of force since 7 October 2023 under the jus contra bellum because the armed conflict, having resulted from Israel’s prior exercise of self-defence, was ongoing on that date.Footnote 74
If taken literally, Israel’s statement – that ‘an analysis of the conditions under which Israel may resort to the use of armed force (the jus ad bellum)’ was unnecessary on 7 October 2023 because Hamas’ attack of that day formed part of an ongoing armed conflict – directly contradicts the separation between the jus contra bellum and the jus in bello: the existence of an (ongoing) armed conflict cannot provide an independent justification for a use of force under the jus contra bellum. However, Israel’s statement in question might also be read as a somewhat imprecise and abbreviated endorsement of the following legal proposition, put forward by the late Yoram Dinstein, which Eliav Lieblich has aptly termed a ‘static approach’:Footnote 75 if the right of self-defence is triggered by an armed attack of a ‘critical’ intensity, the exercise of the right of self-defence by the victim state will not be subject to the continuous application of the principles of necessity and proportionality, but it will provide justification for the use of force under the jus contra bellum ‘until final victory’.Footnote 76 Under this approach, the analysis of the legality of Israel’s use of force under the jus contra bellum would require one to go back in time in order to ascertain whether Hamas, at one point prior to 7 October 2023, had launched an armed attack on Israel that not only triggered Israel’s right of self-defence, but was of such a ‘critical intensity’ that Israel was henceforth entitled to use force against Hamas without the need for recurring jus contra bellum analysis, ‘until final victory’ over Hamas. Such an approach, however, is legally untenable: it would deprive the principles of necessity and proportionality of any meaningfulFootnote 77 limiting legal effect after the initiation of the exercise of self-defence.Footnote 78
Another way to make sense of Israel’s statement is to read it as a misleading use of jus in bello terminology in order to make the jus contra bellum point that the Hamas attack of 7 October 2023 formed part of one single continuing armed attack within the meaning of Article 51 of the UN Charter and customary international law, thus justifying, since its inception, the exercise of Israel’s right of self-defence. It cannot be doubted that a series of sufficiently proximate consecutive armed attacks emanating from the same source may constitute one single, continuing armed attack, for the purposes of the exercise of the right of self-defence.Footnote 79 Yet, Israel has not specified at all the factual basis of such an assertion, limiting the possibilities to assess fully the legal proposition that an alleged armed attack by Hamas began at some time before 7 October 2023, and continued since then.Footnote 80 The legal analysis that follows will thus proceed on the basis of Israel’s claim before the ICJ that its use of force as from 7 October 2023 ‘is in response to the slaughter of 7 October … and to the ongoing attacks against it from Gaza, that Israel has the inherent right to take all legitimate measures to defend its citizens and secure the release of the hostages’.Footnote 81
5.2. Israel’s alleged military occupation of the Gaza Strip as a possible ground for the inapplicability ab initio of Israel’s right of self-defence
South Africa has denied the possibility for Israel to exercise its right of self-defence against the Hamas attack of 7 October 2023 for the following reason: ‘What … Israel is doing in Gaza, it is doing in territory under its own control. Its actions are enforcing its occupation. The law on self-defence under Article 51 of the United Nations Charter has no application’.Footnote 82 The underlying legal position, which other states have endorsed,Footnote 83 is thus that the right of self-defence of a state is not triggered if an armed attack occurs against a state from within territory under military occupation of the state concerned. For this legal proposition, South Africa, correctly, draws authority from a statement made by the ICJ in Legal Consequences of the Construction of a Wall in the Occupied Palestinian Territory, which reads as follows:Footnote 84
The Court … notes that Israel exercises control in the Occupied Palestinian Territory and that, as Israel itself states, the threat which it regards as justifying the construction of the wall originates within, and not outside, that territory. The situation is thus different from that contemplated by Security Council resolutions 1368 (2001) and 1373 (2001), and therefore Israel could not in any event invoke those resolutions in support of its claim to be exercising a right of self-defence.
This statement, however, has remained controversial. In her separate opinion, Judge Higgins observed that:Footnote 85
[she failed] to understand the Court’s view that an occupying Power loses the right to defend its own civilian citizens at home if the attacks emanate from the occupied territory – a territory which it has found not to have been annexed and is certainly ‘other than’ Israel.
The best explanation of the ICJ’s legal proposition would seem to be the following. A military occupation requires that the relevant territory is ‘actually placed under the authority’Footnote 86 of the occupying state’s armed forces. This means that attacks from within occupied territory must be legally assimilated, as far as the power of repression is concerned, with domestic violence. Hence, as much as a state cannot rely on self-defence against its own people, a state cannot rely on self-defence against residents living under the occupation of its own armed forces. In both situations, a state can only make use of powers of law enforcement to restore public order. In the first case, those powers result from the relevant domestic legal order; in the second case, they flow from the international law of military occupation, including, in particular, Article 43 of the 1907 Hague Regulations.Footnote 87
Whether this argument is persuasive does not need to be decided for the purposes of the present analysis. For it – and with it the legal proposition it is intended to support – applies only in the case of military occupation. On 7 October 2023, the Gaza Strip was not, however, under Israel’s military occupation. Such an occupation existed after the Six-Day War in June 1967, but Israel withdrew its armed forces from the Gaza Strip in 2005. Nevertheless, it is widely assumed that Israel’s military occupation of the Gaza Strip has not ended since then.Footnote 88 This assessment is based on the facts that (i) Israel has retained control of the airspace over the Gaza Strip, and (ii) Israel responded to the Hamas takeover of the Gaza Strip by sealing off its border crossings and later also by imposing a naval blockade in response to previous cross-border violence by Hamas. This, so the argument goes, means that Israel, despite its ‘disengagement’ from the Gaza Strip, has retained a degree of control over this territory sufficient to maintain that the military occupation of that state has been ongoing. Yet, the relevant question in the present context is whether the Gaza Strip continued to be ‘actually placed under the authority of the hostile army’ of Israel, for this is the legal standard for the existence of a military occupation according to Article 42 of the Hague Regulations and customary international law. In that respect, it is determinative that it was not possible for Israel’s army – as the events on and since 7 October 2023 have shown once again – to exercise authority in the Gaza Strip without first having to overcome considerable armed resistance from Hamas.Footnote 89 This is incompatible with the existence of actual Israeli military authority over the Gaza Strip at the material time.Footnote 90 The fact that Israel militarily occupied the West Bank on 7 October 2023 is immaterial to the question of military occupation of the Gaza Strip, as the existence of a state of military occupation depends on the factual circumstances in the territory concerned.Footnote 91 Instead of characterising the status of the Gaza Strip on 7 October 2023 as one of Israeli military occupation, it is legally more accurate to state that Israel, through the restrictions and controls it had established, had placed the Gaza Strip under a type of control displaying important features of a siege.Footnote 92
The preceding legal evaluation is not inconsistent with the position adopted by the ICJ in its 2024 advisory opinion in Legal Consequences Arising from the Policies and Practices of Israel in the Occupied Palestinian Territory, Including East Jerusalem. In this opinion, the Court found:Footnote 93
Where an occupying Power, having previously established its authority in the occupying territory, later withdraws its physical presence in part or in whole, it may still bear obligations under the law of occupation to the extent that it remains capable of exercising, and continues to exercise, elements of its authority in place of the local government. Based on the information before it, the Court considers that Israel remained capable of exercising, and continued to exercise, certain key elements of authority over the Gaza Strip, including control over the land, sea and air borders, restrictions on movement of people and goods, collection of import and export taxes, and military control of the buffer zone, despite the withdrawal of its military presence in 2005. This is even more so since 7 October 2023. In light of the above, the Court is of the view that Israel’s withdrawal from the Gaza Strip has not entirely released it of its obligations under the law of occupation. Israel’s obligations have remained commensurate with the degree of its effective control over the Gaza Strip.
This conclusion by the Court stops short of a finding that Israel’s military occupation of the Gaza Strip continued after the withdrawal of its military presence.Footnote 94 As Judge Iwasawa persuasively explained in his separate opinion, the ICJ has hereby, in effect, subscribed to what has been called a ‘functional approach’Footnote 95 to the continued application of certain rules of the law of military occupation after the withdrawal. This approach, in the fitting words of Judge Iwasawa, is not concerned with the status of the territory as such, but with the question whether a former occupying power may continue to be bound by certain obligations under the law of occupation although the occupation as such has ended.Footnote 96 For the purposes of the present analysis, it is not necessary to reach a conclusion on the merits of the functional approach to the (continued) applicability of certain parts of the law of military occupation beyond the actual display of authority by the foreign army.Footnote 97 It is sufficient to realise that the ICJ’s legal proposition in Legal Consequences of the Construction of a Wall in the Occupied Palestinian Territory – that the right of self-defence of a state is not triggered if an armed attack occurs against a state from within territory under military occupation – does not apply in a case where a former occupying power continues to be bound by certain rules of the law of occupation, commensurate with the degree of its effective control. The crucial consideration underlying the proposition that the right of self-defence cannot be exercised by an occupying power – that this power, because of its actual authority over the territory, must rely on its law enforcement capacity – does not apply in a case where the retention of a lesser degree of control may justify the application of certain rules of the law of occupation.Footnote 98 As Israel did not exercise control over Gaza in a manner that would have triggered the obligations of an occupant in full, such a legal status cannot have excluded the possibility that Israel’s right of self-defence was triggered by Hamas’ attack of 7 October 2023.Footnote 99
5.3. The absence of an international legal right of Hamas to attack Israel
If Hamas had been entitled under international law to launch its attack of 7 October 2023 against Israel, this attack could not have placed Israel in a position to exercise its right of individual self-defence as recognised in Article 51 of the UN Charter, for this right cannot be exercised in the case of an internationally lawful armed attack. Starting from this correct legal premise, Ralph Wilde has put forward the view that Israel could not invoke its right of individual self-defence in the case of Hamas' attack against it. This author is of the view that the attack of 7 October 2023 was carried out in the exercise of a ‘right to resist in international law on the part of the Palestinian people’ against Israel’s (continued) illegal use of force in the Gaza Strip in the form of its (continued) military occupation of this territory, which amounts to an aggression. Wilde acknowledges that Hamas' attack of 7 October 2023 violated, in particular, international humanitarian law, but this, in his view, did not affect the international legal entitlement of Hamas to forcibly resist Israel’s aggression.Footnote 100
Wilde’s argument fails for at least two independent reasons.Footnote 101 First, on 7 October 2023, Hamas was not the representative of the Palestinian people and thus was not entitled to exercise any supposedly existing international legal right of the Palestinian people to armed resistance.Footnote 102 Second, while it is correct to assume, as Wilde does, that violations of the law of armed conflict do not as such affect the legal situation under the jus contra bellum, Hamas’ violations of the law of armed conflict are reflective of its stated goal to annihilate the state of IsraelFootnote 103 rather than to defend the people of Palestine against an armed attack by Israel.Footnote 104 To annihilate the state concerned, however, goes beyond the legitimate goal of any supposedly existing international legal right of the Palestinian people to resistance under international law.
For each of these two reasons, it is unnecessary, for the purposes of the present analysis, to enter into the rather complex question of whether the Palestinian people possessed an international legal right to armed resistance against Israel either because of a possible unlawful armed attack by Israel against the Gaza StripFootnote 105 (in the form of, for example, its continued imposition of a naval blockadeFootnote 106) or because of a possible unlawful armed attack against the West Bank and East JerusalemFootnote 107 remainder of the Palestinian people territorial self-determination unit in the form of an unlawful military presence in the West Bank and East Jerusalem – as was determined to exist by the ICJ in its 2024 advisory opinion in Legal Consequences Arising from the Policies and Practices of Israel in the Occupied Palestinian Territory, Including East Jerusalem’.Footnote 108
5.4. ‘If an armed attack occurs against a Member of the United Nations’
In view of its severity, Hamas’ use of force against Israel certainly qualifies as an armed attack within the scope of Article 51 of the UN Charter in terms of its gravity.Footnote 109
5.4.1. The absence of an armed attack by a state
Israel’s right of individual self-defence thus would unquestionably have been triggered as a result of Hamas’ attack of 7 October 2023 if this attack constituted an armed attack by a state within the meaning of Article 51 of the UN Charter and customary international law. In the absence of a separate Gaza state at the material time,Footnote 110 Hamas' attack could not be qualified as an armed attack by such a state. If a state of Palestine existed on 7 October 2023 and if Hamas’ relevant conduct could be attributed to such a state under the applicable rules of international law, the attack could be qualified as an armed attack by the state of Palestine.Footnote 111 Here again, the question of Palestinian statehood, however, can be left undecided, for even if a Palestinian state had existed on 7 October 2023 and even if the Gaza Strip had formed part of the territory of such a state, Hamas’ attack could not have been attributed to such a state.Footnote 112 To begin with, there is no basis for attribution under the general rules of attribution that the International Law Commission (ILC)Footnote 113 has codified. The government of a state of Palestine on 7 October 2023 would have been the Palestinian Authority. This means that the members of Hamas who carried out the attack against Israel could not have been qualified as organs de jure of a Palestinian state.Footnote 114 Having not acted in complete dependence on the Palestinian Authority, the relevant members of Hamas could also not have been considered to have acted as organs de facto Footnote 115 of a Palestinian state. There is also no factual basis for maintaining that the relevant members of Hamas acted under the effective control of the Palestinian Authority.Footnote 116 Finally, the relevant conduct of Hamas could also not be seen as having occurred in the exercise, in fact, of elements of the governmental authority of a Palestinian state in the absence or default of the official authorities in circumstances such as to call for the exercise of those elements of authorityFootnote 117 for there were no circumstances ‘to call for’ this brutal attack against Israel.Footnote 118
Furthermore, assuming again the existence of a Palestinian state on 7 October 2023, Hamas' attack would also not be attributable to such a state under a rule of attribution specifically applicable to the concept of armed attack.Footnote 119 The ICJ has yet to recognise such a special rule of attribution, but scholars have occasionally admitted the existence of such a rule. The central point of reference is Article 3(g) of the UNGA definition of aggression of 1974,Footnote 120 which has been used by the ICJ for the interpretation of the concept of armed attack by a state.Footnote 121 This provision reads as follows:
The sending by or on behalf of a State of armed bands, groups, irregulars or mercenaries, which carry out acts of armed force against another State of such gravity as to amount to the acts listed above, or its substantial involvement therein.
While the first alternative (the ‘sending’) can be interpreted in conformity with the effective control test under the general rules of attribution,Footnote 122 this is not possible in case of the second alternative (the ‘substantial involvement’). The latter might thus indeed articulate a special rule of attribution lowering the threshold as compared to that under the applicable general rules.Footnote 123 How far precisely the threshold is lowered remains difficult to determine with precision, and the ICJ has remained silent in that regard.Footnote 124 Helmut Aust suggests that the following rather far-reaching special rule of attribution exists:Footnote 125
[Acts of force of sufficient gravity] are also attributable to a State if they have been committed by private persons and the State has encouraged these acts, has given its direct support to them, planned or prepared them at least partly within its territory, or was unwilling to take steps which can reasonably be expected of it to prevent these acts after having received substantiated information. The same is true if a State demonstrably gives shelter to terrorists after they have committed an act of terrorism within another State in a situation in which the attack can still be regarded as ongoing.
The proposed rule covers even the ‘harbouring’ on its territory by a state of private persons who carry out acts of force (of sufficient gravity) against another state. Whether a special rule of attribution exists that reaches that far remains a matter of quite some doubt.Footnote 126 However, there is no need to take a position on this issue for the purposes of the present analysis, as the factual relationship between the Palestinian Authority and Hamas makes it impossible to say that the former harboured the latter on 7 October 2023. In conformity with the foregoing considerations, apparently no state that recognised Palestinian statehood on 7 October 2023 has qualified Hamas’ attack against Israel as an armed attack by the state of Palestine.
It may also be asked whether Hamas' attack of 7 October must legally be qualified as an armed attack by Iran. In that respect, it bears noting that Israel, in its legal explanation regarding Operation ‘Rising Lion’ against Iran (conducted between 13 and 24 June 2025), pointed out that Iran had continuously been providing support to several ‘armed groups,’ including Hamas. More specifically, Israel has stated: ‘In terms of international law rules governing the use of force, Iran’s substantial involvement in acts of armed force carried out by these groups legally constitutes an armed attack by Iran against Israel’.Footnote 127
Israel has not, however, specified its precise understanding of the standard of ‘substantial involvement’ and it has also not elaborated on the facts supporting its proposition regarding the Hamas attack of 7 October 2023. In fact, Israel has stated that the ‘exact nature of the Iranian’s regime relationship with, and degree of control over, each of these armed groups has fluctuated over time and through circumstances’.Footnote 128 It is also noteworthy that Israel had not spoken of an armed attack by Iran in its legal explanation regarding Operation ‘Iron Swords’. Against this background and without being privy to the relevant facts, it would be bold for the present author to assume that Hamas' attack of 7 October 2023 constituted an armed attack by Iran on the basis of the latter’s substantial involvement therein.Footnote 129 However, it bears stressing that even if this attack constituted an armed attack by Iran, Israel would have to provide a justification of self-defence specifically vis-à-vis either the state of Palestine (assuming such a state existed) or the Palestinian self-determination territory to the extent that its use of force was carried out in the Gaza Strip.
5.4.2. The uncertainty as regards a right of self-defence in the case of an armed attack by a non-state organisationFootnote 130
The attack by Hamas of 7 October 2023 thus being best understood as non-state conduct, Israel’s right of self-defence as a result of this attack depended on whether Article 51 of the UN Charter recognises a right of self-defence in the case of an armed attack carried out by a non-state organisation. While this topic had already received quite some attention before the Al-Qaeda attacks of 11 September 2001,Footnote 131 the wealth of academic literature on the subject has grown exponentially following this event.Footnote 132 At present, the question is one of the most controversial pertaining to the contemporaneous jus contra bellum.Footnote 133
5.4.3. The absence of a general (subsequent) practice accepted as law
In the Nicaragua case, the ICJ implied – without further explanation and without the judgment resting on this point – that the right of self-defence required an armed attack by a state.Footnote 134 Upon this reading of the UN Charter (and customary international law), which the Court has not abandoned since, a state under an armed attack by non-state actors stemming from outside that state is prevented from using force in self-defence against the non-state actors in their foreign host territory, even if the attack cannot otherwise be effectively halted and repelled. Yet, the controversy surrounding this issue is visible, not least within the Court itself, given that a number of judges have voiced their dissent on the point in question in the form of individual opinions.Footnote 135
As far as the practice of states is concerned, there was no consensus in line with the ICJ’s legal position even before 1986.Footnote 136 To begin with, and contrary to what is sometimes asserted,Footnote 137 the UNGA definition of aggressionFootnote 138 does not articulate such a consensus. It is true that Article 3(g) of that definition describes state conduct, but this form of state conduct is listed as an instance of aggression, not as an instance of armed attack. It is therefore important to correctly appreciate the relationship between these two legal concepts, as understood by the UNGA in its definition of aggression. Its fifth preambular paragraph states that ‘aggression is the most serious and dangerous form of the illegal use of force’. In view of this formulation, it is justified, as the ICJFootnote 139 has in fact done, to interpret the inclusion of a certain type of conduct in the UNGA definition of aggression as an instance of aggression, as expressed in the views of UN member states that the same type of conduct also constitutes an armed attack. However, it is not warranted to assert that the exclusion of a certain type of conduct, in particular non-state conduct, is testimony to the position of those same states that such conduct could not qualify as an armed attack. In that sense, the UNGA definition of aggression does not purport to provide, through implication, a comprehensive definition of the term ‘armed attack’ within the meaning of Article 51 of the UN Charter.Footnote 140
As far as concrete incidents during the course of the first decades after the entry into force of the UN Charter are concerned, a legal position contrary to that of the ICJ was most prominently claimed by Portugal, South Africa, and Israel. However, these states had no prospect of their legal claim being accepted internationally because of the specific contexts in which they were confronted with non-state cross-border violence: Portugal fought a frowned-upon battle for the preservation of its colonies; South Africa was an international pariah based on its racist regime; and Israel’s prolonged military occupation of a self-determination territory was criticised by many.Footnote 141 At the time, the fact that Turkey and Iran, in particular, also reacted militarily to instances of transnational non-state uses of force, and claimed that they were legally entitled to do so, remained largely unnoticed.Footnote 142
Only the devastating terrorist attacks of 11 September 2001 placed the core issue squarely on the international legal agenda.Footnote 143 Since then, it has been discussed with great intensity as to whether the exercise of the right of self-defence under international law requires, in all circumstances, that the attacker is a state. Immediately after the terrorist attacks, the United States, together with a number of allies, started its military operation ‘Enduring Freedom’ in Afghanistan, claiming to be acting in the exercise of its right of self-defence.Footnote 144 In the context of its international legal justification, the United States first took the position that a state itself would become the attacker if it purposefully harboured (as did Afghanistan – at the time ruled largely by the Taliban) a non-state organisation (e.g., Al-Qaeda) launching a cross-border attack from its territory. An essential part of the US international legal political strategy was thus to defuse the international law challenge of massive non-state transnational armed force through a generous interpretation of the concept of state armed attack.Footnote 145 But, as was shown above,Footnote 146 the international legal viability of this line of justification was, and is, doubtful. The US attempt to broaden the concept of an armed attack by a state is one important reason why – despite the wide measure of support that ‘Enduring Freedom’ received in its early years throughout the international community – it remains an open question if, and to what extent, the practice of states in connection with this use of force endorsed the view that the right of self-defence can be exercised on the territory of another state against a non-state organisation in order to halt and repel the latter’s armed attack, despite it not being one of the territorial states.Footnote 147
Neither the United States nor any other state has gone so far as to maintain that even a state that is unable to prevent a non-state actor from using force from its territory becomes an attacker. In a number of situations of the kind that arose after 2001, states invoked their right of individual self-defence: the Russian Federation proceeded in that way in 2002 to justify its use of force against Chechen rebels in Georgia; Israel acted in the same manner in 2006 to justify its use of force against Hezbollah in Lebanon; in 2017, Turkey eventually chose the same course of action to justify its use of force against the PKK in Iraq, and Colombia went down the same path in 2008 to justify its use of force against FARC in Ecuador.Footnote 148 While Israel’s exercise of the right of self-defence was, in principle, endorsed by many other states, the international reaction was far more nuanced in the other cases and negative in the case of Colombia’s use of force.Footnote 149
In 2014, a situation of inability of the territorial state arose in Syria, a situation that became central to the international legal debate about self-defence and non-state armed attacks.Footnote 150 Syria had already been tormented by an armed conflict when the so-called ‘Islamic State’ gained a foothold there, before bringing neighbouring Iraq to dire straits militarily. At no time was the ‘Islamic State’ a state under international law. However, as of 2014, the United States, along with a number of other states, claimed the right to defend Iraq, at the latter’s request, against cross-border attacks by the ‘Islamic State’ originating on Syrian territory. In doing so, the states concerned made clear to the UNSC their legal conviction that, under certain circumstances, a non-state armed attack could justify the exercise of the right of self-defence. As regards those circumstances, most of the states concerned adopted the test of the unwillingness or inability of the territorial state.Footnote 151 Germany added a legal nuance in highlighting the fact that the ‘Islamic State’ had ‘occupied a certain part of the Syrian territory’.Footnote 152 It bears noting that the case of Syria and the ‘Islamic State’ is the first instance in the practice of states, following the entry into force of the UN Charter, in which a group of states unambiguously invoked the right of collective self-defence in the case of a non-state armed attack. The use of force against the ‘Islamic State’ in Syria has not, however, given rise to a new consensus among states. On the contrary, it would seem that this case has sparked an international legal debate among states, in the course of which those states that categorically reject the idea of a right of self-defence in the case of a non-state armed attack have become increasingly more vociferous. In 2016, 120 states belonging to the Non-Aligned Movement adopted a statement to the effect that ‘Article 51 of the UN Charter is restrictive and should not be re-written or re-interpreted’.Footnote 153 While this formulation is too general to reach firm conclusions, on 24 February 2021, in a UNSC meeting convened by Mexico under the Arria Formula, a number of UN member states went on record with the unambiguous legal position that the exercise of the right of self-defence requires an armed attack by a state.Footnote 154 A statement to the same effect is included in a document issued by the African Union in January 2024.Footnote 155 In January 2025, Mexico restated the same position in a detailed letter.Footnote 156 At the same time, however, a considerable number of states have made use of the 2021 Arria meeting to state their legal conviction that non-state armed attacks can give rise to the exercise of the right of self-defence under certain conditions.Footnote 157 The underlying legal question received less specific attention when the United States and Great Britain used force on Yemeni soil in January 2024 and April 2025 to deter repeated attacks by Houthi forces against vessels in the Red Sea, which had commenced in November 2023. Both the United States and Great Britain invoked the right of self-defence, without going into any detail concerning the non-state character of the armed attack to which they responded.Footnote 158 The military operation was condemned as unlawful by a few states, but here again the focus was not on the non-state character of the Houthis.Footnote 159 Over the course of the ensuing debate in the UNSC it was made clear by Switzerland, in particular, that the reference in UNSC Resolution 2722 (2024) to the ‘right of Member States, in accordance with international law, to defend their vessels from attacks’Footnote 160 was not tantamount to a recognition of a right of self-defence that would have justified forcible action on Yemeni soil.Footnote 161
All in all, the pertinent (subsequent) practice of states therefore currently illustrates a fairly deep-seated division. On the whole, the reactions of states to Israel’s military operations in the Gaza Strip since 2008, including operation ‘Iron Swords’, reveal a greater amount of support for Israel’s entitlement to exercise its right of self-defence, but there is no consensus. While all three major previous military operations – ‘Cast Lead’ (2008–09), ‘Pillar of Defense’ (2012), and ‘Protective Edge’ (2014) – gave rise to widespread international criticism, the latter was based predominantly on the alleged disproportionality of Israel’s recourses to the use of force, and not on a denial of a right of self-defence as such.Footnote 162 As far as operation ‘Iron Swords’ is concerned, again, a significant number of states have initially taken the view that Israel was in a position to exercise its right of self-defence, hence not considering the non-state character of Hamas to be a legal obstacle.Footnote 163 At the same time, the Non-Aligned Movement has called Israel’s military operation an ‘illegal aggression’.Footnote 164 Again, the explanation given for rejecting Israel’s right to self-defence did not appear to rest on the basis of Hamas’ non-state character. Instead, the critical states often alleged Israel’s military occupation of the Gaza Strip and then referred to the ICJ dictum in the Wall case in order to object to an exercise of the right of self-defence.Footnote 165 Nevertheless, the wide measure of opposition to Israel’s Operation ‘Iron Swords’ makes it impossible to say that the international reaction has moved the practice of states decidedly in the direction of a consensus about a right to self-defence in cases of non-state armed attacks.Footnote 166
5.4.4. The ordinary meaning and the preparatory work: A case of genuine legal uncertainty
For Israel, as for all UN member states, the question of whether the right of self-defence extends to a non-state armed attack is primarily a matter of interpretation of the UN Charter rather than of the identification of the applicable rule of customary international law. Having established the absence of a general subsequent practice of UN member states accepted as law, in the following the issue will be revisited in applying the remaining customary rules of treaty interpretation, as embodied in Articles 31 to 33 of the Vienna Convention on the Law of Treaties (VCLT).Footnote 167 The general rule is reflected in its Article 31(1) and stipulates that ‘(a) treaty shall be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose’.Footnote 168
The wording of Article 51 of the UN Charter qualifies the requisite of an attack by way of its modality (‘armed’) and its target (‘a Member of the United Nations’), but it does not say anything about the author of the attack. In particular, none of Article 51’s authentic versions posit the requirement that the armed attack must be carried out by ‘a Member of the United Nations’ or ‘a state’. Prior to the UN Charter, states had also not used the term ‘attack’ exclusively in connection with state conduct. Quite the contrary, they had also repeatedly described violent cross-border action by private persons as ‘attacks’.Footnote 169 Accordingly, the ordinary meaning of Article 51 appears to cover state and non-state armed attacks alike. It has been argued, however, that the context within which the term ‘armed attack’ is used in Article 51, as well as the object and purpose of the UN Charter, require a state-centric reading of the term ‘armed attack’ in Article 51.
An initial contextual argument starts with the observation that the prohibition of the use of force in Article 2(4) of the UN Charter applies only to states. This is taken to suggest that the concept of ‘armed attack’ in Article 51 should be understood in symmetric terms, hence requiring a state authorship.Footnote 170 While such an approach is certainly arguable, it is not compelling. This is because of the different legal contexts in which the concepts of ‘use of force’ and ‘armed attack’ are used in Article 2(4), on the one hand, and in Article 51, on the other. The term ‘use of force’ denotes conduct that is prohibited by virtue of Article 2(4). In this context, it is natural that the conduct must be carried out by a state, only the latter being the addressee of that rule. Conversely, the term ‘armed attack’ in Article 51 does not describe the conduct of the rule’s addressee, but it constitutes a condition for a legal permission. Such a condition may consist of non-state action without this entailing any inconsistency with Article 2(4). Instead, also in the case of a non-state armed attack, the addressee of the permissive rule contained in Article 51 would be a state. There is thus, in any event, complete structural symmetry between Articles 2(4) and 51. A variant of the ‘symmetry argument’ starts from the observation that Article 2(4) prohibits the use of force by a state in its international relations. It is argued that, hence, Article 51 must possess an international scope of application as well, and that such an international dimension would be lacking if Article 51 was applied in the case of a non-state armed attack. However, the last part of this reasoning is inaccurate, for Article 51 does contain a permissive rule for use of force by a state in its international relations. The exercise of the right of self-defence involves international relations not only in the case of an armed attack by another state, but also in the case of a non-state armed attack coming from abroad. In the latter case, Article 51, if applicable, provides the defending state with permission to use force either in relation to the state from whose territory the non-state armed attack was carried out, or, exceptionally, in relation to the people entitled to exercise their right to self-determination on the territory from where the non-state armed attack emerged.Footnote 171 In fact, it is this legal effect vis-à-vis the territorial state that constitutes the actual bone of contention regarding the application of Article 51 to non-state armed attacks,Footnote 172 while the defensive use of force by the attacked state is generally considered unproblematic vis-à-vis the non-state attacker, unless the latter can – exceptionally – claim to act on the basis of an international legal permission.Footnote 173
The second contextual argument emphasises that Article 51 constitutes an exception from Article 2(4) and maintains that the inclusion of non-state armed attacks in the term ‘armed attack’ in Article 51 would run counter to the principle that exceptions should not be interpreted so broadly that they render the primary rule meaningless.Footnote 174 In responding to this argument, it is not decisive whether one agrees with the proposed relationship of rule and exception between Articles 2(4) and 51, or whether one prefers the position taken by the ILC that the right of self-defence in Article 51 forms part of the prohibition of the use of force so that the exercise of the right of self-defence ‘is not, even potentially, in breach of Article 2(4)’.Footnote 175 It is sufficient to state that while the application of the right of self-defence to non-state armed attacks would indeed effectively entail a restriction of the prohibition of the use of force, this prohibition would still retain a scope of application to be considered very far from meaningless.
The third contextual reason given in support of a state-centric interpretation of the term ‘armed attack’ is derived from the second part of the first sentence of Article 51. Herein, the exercise of the right of self-defence is subordinated to the adoption, by the UNSC, of ‘measures necessary to maintain international peace and security’. A non-state armed attack, so the argument begins, does not concern international – understood as interstate – peace and security and, accordingly, the UNSC cannot activate the collective security system under Chapter VII of the UN Charter. This would mean that the exercise of the right of self-defence in the case of a non-state armed attack could not be subordinated to the adoption of ‘measures necessary to maintain international peace and security’ by the UNSC, contrary to the normative scheme of Article 51.Footnote 176 Also, however, this third contextual consideration turns out to be unconvincing for, even if one construes ‘international’ within Chapter VII narrowly in the sense of ‘interstate’, it is perfectly possible to consider a non-state armed attack emanating from the territory of one state and directed against another state as a threat to international peace and security within the meaning of Article 39 of the UN Charter. Therefore, it was always open for the UNSC to activate its powers under Chapter VII in the case of a cross-border non-state armed attack.Footnote 177 In the meantime, as many observers have noted,Footnote 178 the UNSC has repeatedly acted accordingly.
The fourth consideration in support of a state-centric construction of the term ‘armed attack’ in Article 51 combines contextual with purposive reasoning, and it is the most important one. In his capacity as Special Rapporteur to the ILC on State Responsibility for Internationally Wrongful Acts, Roberto AgoFootnote 179 correctly argued that, in Article 51, the term ‘armed attack’ constitutes a condition of the right of self-defence and must therefore be interpreted in conformity with the very essence of this notion. The concept of self-defence, so the second step of the argument goes, requires prior wrongful action. From there, the conclusion is derived that an armed attack within the meaning of Article 51 must constitute an internationally wrongful use of force by a state rather than non-state action escaping the prohibition of the use of force. According to this line of reasoning, if a state wishes to defend itself against a cross-border non-state armed attack on the territory of the host state of the non-state actor, the defending state would find itself in a state of necessity in relation to the host state rather than in a state of self-defence.Footnote 180
This finely construed argument carries considerable weight but, again, it is not compelling. First of all, a host state of the non-state attacker may still act wrongfully, despite the state not being the attacker itself. This is obviously so where the host state, while being capable of preventing the non-state armed attack from occurring, does not do so because it is either unwilling to act or fails to exercise due diligence to that effect. In those two scenarios, the host state commits an internationally wrongful act which is connected with the non-state armed attack.Footnote 181 Therefore, even assuming that a prior wrong done is indeed the essence of the concept of self-defence as part of the jus contra bellum, it is perfectly possible to say that the attacked state finds itself in a situation of self-defence vis-à-vis the host state of the non-state attacker.Footnote 182
The issue becomes considerably more complicated, however, where the host state is unable to prevent the non-state attack from occurring,Footnote 183 for in that case the host state does not commit an internationally wrongful act within the meaning of the law on state responsibility. It is therefore unsurprising that those who object to the inclusion of non-state armed attacks in the term ‘armed attack’ in Article 51 often focus on the scenario of the unable host state. Dire Tladi, a particularly articulate proponent of this view, has called the situation the ‘innocent non-consenting third state scenario’.Footnote 184 Yet, the term ‘innocent’ does not do justice to the complexity of the matter. While the unable host state does not act wrongfully under the law of state responsibility, this state is not just a ‘bystander’ with respect to the non-state armed attack, for it is the territory under the host state’s sovereign sphere of control from where the armed attack – and the significant danger this attack entails for the victim state – emanates. In such circumstances, the fact that the host state does not incur state responsibility for wrongful conduct with respect to the non-state attack does not mean that the host state could not be liable to tolerate, on its territory, forcible action by the victim state defending against the attack. Quite on the contrary, it is justifiable, on the basis of a fair allocation of the costs involved in halting and repelling the non-state attack, that the host state would incur liability upon which its right to the inviolability of its territory suffers. After all, the forcible action by the victim state, to which the host state is liable, amounts to essentially the same type of action that the host state, if capable, would itself have been duty-bound to carry out.Footnote 185
This consideration does not, by itself, answer the question, however, of whether the situation of the victim state vis-à-vis the unable host state can properly be characterised as one of self-defence or whether it can indeed only be called a state of necessity, as Roberto Ago has suggested. The answer to this question is far less straightforward than is often suggested. In fact, the unable host state scenario lies at the borderline between self-defence and a state of necessity in a way that makes it impossible to reach a decision by recourse to the ‘essence’ of either of those concepts.Footnote 186 This is so even if it is assumed that a prior wrong is the essence of self-defence under international law, given that the concept of ‘wrongfulness’ is itself full of complexity as it can carry different meanings depending on the context in which it is used. It is thus conceivable to consider the conduct of the host state as wrongful for the specific purposes of self-defence as distinct from the purposes of state responsibility. In the unable host state scenario, wrongfulness in the sense of state responsibility is absent because the content of the relevant international obligation – flowing from the ‘no harm principle’– is to exercise due diligence to prevent a cross-border non-state attack rather than to prevent such a result categorically.Footnote 187 To construe the international obligation in this way makes sense in a system of state responsibility, which does not contain a category of ‘fault’ as a general condition in view of the specific legal consequences flowing from a finding of state responsibility. These consequences may reach from an obligation to cease the wrongful conduct to, possibly, that of reparation. The legal consequence of the exercise of the right of self-defence, however, is entirely different. It includes liability to tolerate the intrusion into the right that the defensive action entails. Such liability does not require a failure to exercise due diligence; it may arise also from a wrong in a purely objective sense, such as the non-prevention by the host state of a non-state armed attack emanating from its sovereign territory.Footnote 188 It already follows, for this reason, that the position taken by the ILCFootnote 189 – that the United States and Great Britain used ‘incorrect’ legal terminology in the historic Caroline caseFootnote 190 – is far too categorical.Footnote 191 Both states had concurred in characterising the position of Great Britain vis-à-vis the United States as one of self-defence, without premising this on the charge that the US had failed to exercise due diligence in preventing the occurrence of the non-state armed attack in question.Footnote 192
The real question, as far as the contextual and purposive interpretation of Article 51 is concerned, is therefore which of the two possible legal characterisations – ‘self-defence’ or ‘state of necessity’ – provides for a normatively more convincing jus contra bellum framework to govern the relevant scenarios of a cross-border non-state armed attack. In searching for the answer, it is important to consider, first, that under customary international law, as reflected in Article 25 of the ILC Draft Articles on Responsibility of States for Internationally Wrongful Acts, a state of necessity does not constitute a circumstance precluding wrongfulness where it seriously impairs an essential interest of the state to which the obligation exists. This is to mean, as the ILC explains further in its commentary, that a state of necessity can be invoked only where the interest to be protected ‘outweighs all other considerations’. Where the state that suffers the intrusion is a mere bystander, such a stringent standard of proportionality is perfectly apposite. However, it is questionable, at the very least, whether this is also the case where the danger for the interest to be protected emanates from the sovereign sphere of control of the state that is to suffer the intrusion, as is the case in the scenarios of a cross-border non-state armed attack under consideration. Second, it must be borne in mind that, contrary to what some proponents of a categorisation of our scenarios as situations of a state of necessity have suggested,Footnote 193 the latter is not available as a circumstance precluding the wrongfulness of a use of force within the meaning of Article 2(4) as a matter of proper interpretation of the UN Charter.Footnote 194 To categorise the unable host state scenario as a state of necessity is thus tantamount to suggesting that the victim of the non-state armed attack is legally deprived of the ability to halt and repel that armed attack at its source in the host state. In the light of what was set out before, this does not seem a fair allocation between the victim and the host state of the costs involved in halting and repelling the non-state attack. Two considerations may nevertheless be advanced in support of a categorisation as a state of necessity also in view of this consequence. First, it might be argued that an exception (the right of self-defence) from a rule (the prohibition of the use of force) should be construed narrowly. Second, it might be maintained that the unavailability of a state of necessity as a circumstance precluding the wrongfulness of a use of force is reflective of the overarching purpose of the UN Charter to restrict the instances of a lawful use of force in the interest of the preservation of international peace. One could wonder whether this decision made by the drafters of the UN Charter would be undermined by allowing a situation to come under the legal umbrella of self-defence where the characterisation of the same situation as a state of necessity is equally possible.Footnote 195 Neither of those considerations, however, is compelling. The rule–exception argument is unduly formalistic because it is a matter of pure construction whether one considers the prohibition of the use of force and the right of self-defence as two separate rules or, as the ILC prefers, as two parts of one single conduct rule. The second – purposive – consideration ultimately rests on the idea that the maintenance of international peace, given its superior status in the Preamble and in Article 1(1) of the UN Charter, must trump the security interest of the state under non-state attack at least in the unable host state scenario. However, it is problematic to use the overarching purpose of maintaining international peace and security in such a way, for the right of states to exercise self-defence has also found recognition in the UN Charter. If exercising the right of self-defence is therefore an integral component of the legal edifice erected by the Charter in the pursuit of international peace and security, it is somewhat circular to decide borderline questions about the scope of the right of self-defence by direct reference to the goal of maintaining international peace and security. Furthermore, it should be recalled that it is now firmly established in the practice of the UNSC that a non-state cross-border armed attack may constitute a threat to international peace and security. If this is so, it is no legal stretch at all to regard forcible action to halt and repel such an attack as a measure to maintain, rather than to disturb, international peace and security.Footnote 196 This is not to suggest that it is compelling to characterise the scenarios of a cross-border non-state attack, and in particular that of the unable host state, as situations of self-defence rather than of a state of necessity, for the purposes of a contextual and purposive interpretation of Article 51. But it is to demonstrate that it is also not compelling – in fact, far from compelling – to adopt an interpretation of Article 51, based on a narrow understanding of the concept of self-defence, that would leave a state suffering from a non-state armed attack, which may well be of no lesser gravity than that of an armed attack carried out by a state, without the right to defend itself where the attack originates.
At this juncture, the specific situation of the Gaza Strip on 7 October 2023 calls for one additional reflection. As Yoram Dinstein has astutely observed, the unable host state scenario is an ‘extraordinary case’.Footnote 197 If Palestinian statehood at the material time is assumed, the Gaza war constitutes an instance of such a case. If Palestinian statehood is denied, however, the situation becomes even more extraordinary and perhaps unique, for we are then confronted not with an unable host state, but with an unable people entitled to exercise their right to self-determination over the territory from where the non-state armed attack emanates. This gives rise to the question of whether the preceding contextual and purposive considerations apply in essentially the same manner in what may be called an ‘unable host people scenario’. By definition, a self-determination territory is not – yet – under the sovereign control of the people entitled to exercise the relevant right to self-determination. It is therefore impossible to apply the ‘no harm principle’ to the ‘host people’ and to consider the people’s conduct with respect to the non-state attack as wrongful, even in a purely objective sense. This could be taken to suggest that the exercise of the right of self-defence is inapposite in this ‘very exceptional’ scenario. Then, however, the inclusion in Article 2(4) of a use of force by a state on a self-determination territory beyond its borders, as supported above,Footnote 198 would have the result that a ‘host people’ finds itself in a stronger position under the jus contra bellum than a ‘host state’. Such a legal situation does not appear to be convincing and it can be avoided. It must be borne in mind that the extension of the scope of application of Article 2(4) to the use of force by a state on a self-determination territory beyond its borders effectively means that people entitled to create a state on a given territory are given legal protection against forcible intrusion into the relevant territory equivalent to the protection that a state enjoys. However, to be normatively consistent, this equivalence in legal protection must carry with it as a corollary the same liability to tolerate a forcible intrusion into that territory by a victim state of a non-state armed attack emanating from the self-determination territory. It follows that the contextual and purposive considerations set out above with respect to the extraordinary case of an unable host state must apply by analogy to the even more extraordinary case of an unable host people.Footnote 199
As a result of the foregoing application of the customary rule of treaty interpretation reflected in Article 31(1) of the VCLT, whether or not the ordinary meaning to be given to the term ‘armed attack’ in Article 51 requires authorship by a state in its context and in the light of the object and purpose of the UN Charter is a matter of quite considerable ambiguity. Because of the customary rule of interpretation embodied in Article 32(a) of the VCLT, this ambiguity suggests turning to the preparatory work of the UN Charter for possible further elucidation of the matter.
Article 51 was included in the UN Charter only in San Francisco, at a rather late stage of the negotiations.Footnote 200 When countries from Latin America voiced concern that the right to take measures of collective defence under the (inter-American) Act of Chapultepec could be curtailed by the envisaged monopoly of the UNSC in adopting measures of collective security, the idea was born within the US-delegation to address the problem by spelling out the right of self-defence in the UN Charter. The final version of Article 51, the inclusion of which in the Charter was decided by the competent Conference Committee III/4 on 23 May 1945, stems from a joint proposal submitted by the United States and Great Britain on 12 May 1945. The United States had presented an initial draft on the same day in which a distinction was drawn between the rights of individual and collective self-defence. Only the latter was subjected to an ‘armed attack’, while the former was made dependent on prior ‘aggression by any state’. Although the author of the ‘armed attack’ was not explicitly specified in this draft, the need for an armed attack by a state was implicit for two reasons. First, the passage referred explicitly to the Act of Chapultepec, and therein the right of third states to take action in collective defence was made dependent on ‘every attack of a State against … an American State’.Footnote 201 Second, it was the United States’ intent to make the scope of application of the right of collective self-defence narrower than that of individual self-defence. In the joint US-American and British draft, the distinction between the conditions of individual and collective self-defence was eliminated because of the concern voiced by Great Britain that ‘aggression’ was a term too hard to define with precision. As a result, the term ‘armed attack’ became the prior conduct in the cases of both individual and collective self-defence. Yet, nothing suggests that the understanding of the term ‘armed attack’ as action by a state had undergone a change during the course of that day.Footnote 202 The preparatory work of the UN Charter thus supports an understanding of the term ‘armed attack’ in Article 51 in the sense of an ‘armed attack by a state’. This is clearly of considerable significance in respect of the meaning given to the term ‘armed attack’ in the context of collective self-defence, for it was the right of collective self-defence that was not yet clearly recognised in international law at the time, and that the Latin American states thus saw in need of receiving an explicit mention. The right of collective self-defence they wished to see confirmed in the UN Charter consisted of an armed attack by a state.Footnote 203 That a right of collective self-defence in the case of a non-state armed attack was not an issue in San Francisco is unsurprising because such a right had also not been an issue in international law before then.Footnote 204
A right of individual self-defence in the case of a non-state armed attack, however, had been a long-standing issue in international law, to which the Caroline case is testament. Therefore, the question arises as to the significance to be attributed to the fact that the drafters of Article 51 understood the term ‘armed attack’ in the sense of ‘armed attack by a state’ in the context of the right of individual self-defence. This is a far more complex matter because the right of individual self-defence had been an issue in the negotiations of the UN Charter not only in the context of Article 51, but also – and primarilyFootnote 205– in the course of drafting Article 2(4). Here, starting from Dumbarton Oaks, it was the common understanding of the negotiators, even after the decision was made to include Article 51, that the prohibition of the use of force left the ‘use of arms in legitimate self-defense … admitted and unimpaired’.Footnote 206 It follows that the (likely) understanding of the term ‘armed attack’ as ‘armed attack by a state’ by those who formulated Article 51 does not necessarily suggest that the preparatory work of the UN Charter supports the view that no right of individual self-defence was admitted in the case of a non-state armed attack. For it is possible that the drafters of Article 2(4) considered the exercise of such a right to be ‘unimpaired’ by that prohibition.
The latter certainly was the understanding of the US delegation at San Francisco, as Tadashi Mori has revealed on the basis of a meticulous analysis of the relevant internal discussions. Yet, as this author emphasises as well, the question of a right of individual self-defence in the case of a non-state armed attack was not explicitly taken up in the actual negotiation process.Footnote 207 This process was directly concerned only with the right of individual self-defence in the case of (state) aggression; a right of self-defence that had become more and more firmly entrenched in international law after the First World War, alongside the diplomatic process towards the outlawing of aggressive war that paved the way to the prohibition of the use of force in Article 2(4).Footnote 208 However, the explicit focus of the negotiations in relation to Article 2(4) on what Mori calls the conception of individual ‘self-defence as defensive war’, as developed in the interwar period, does not exclude the implicit reception of what the same author calls the nineteenth century ‘policing concept of the right of individual self-defence’ as a second category of self-defence, which was equally admitted and unimpaired by Article 2(4). The emergence of this type of the right of individual self-defence can be traced back to the West-Florida dispute of 1818 between the United States and Spain, to the diplomatic exchanges between the United States and Mexico over their 1836 border incident, and to the classic diplomatic correspondence between the United States and Great Britain in the Caroline case.Footnote 209 While it is true – and important – that Mexico placed a strict limit of necessity and proportionality on the right of self-defence, it did not go so far as to reject the latter altogether.Footnote 210 Sometimes the nineteenth century state practice regarding the right of self-defence is considered to be legally irrelevant because of the absence, at the time, of a prohibition of war,Footnote 211 but this does not rest on an accurate understanding of the historic evolution. What Mori calls the policing concept of individual self-defence in the nineteenth century was not understood by states and scholars at the time as an exception from a prohibition of war, but as a ground for precluding the wrongfulness of forcible entry, not amounting to war, into the territory of another state. The nineteenth century right of individual self-defence, having been applicable in the scenario of an ‘unable host state’, in modern parlance constituted a ground for precluding the wrongfulness of forcible action in breach of the inviolability of state territory in peacetime.Footnote 212 In the interwar period, this nineteenth century type of the right of individual self-defence retreated into the background while a right of individual self-defence as the exception from the emerging prohibition of (aggressive) war entered into the foreground of the international legal discourse. Upon closer inspection, however, it turns out that the new type of individual self-defence in the case of aggression by a state did not replace the old nineteenth century type of individual self-defence; instead, the two coexisted with their respective scopes of application.Footnote 213 Therefore, a strong argument can be made that the statement according to which the right of individual self-defence remained ‘admitted and unimpaired’ concerned not only the right of individual self-defence in the case of state aggression, but also in the case of non-state armed attacks.Footnote 214
The examination of the preparatory work of the UN Charter thus leads to a nuanced conclusion. As far as the right of collective self-defence is concerned, the genesis of the treaty text supports an understanding of ‘armed attack’ in Article 51 in the sense of ‘armed attack by a state’. However, the same is not true concerning the right of individual self-defence; here, the matter is genuinely complicated. While there is no suggestion that the term ‘armed attack’ was given a broader meaning in the drafting process leading to Article 51, a strong argument can be made that the drafters of Article 2(4) did not consider forcible action in conformity with the long-established right of individual self-defence in cases of non-state armed attacks as falling within the scope of the prohibition of the use of force.
5.4.5. A grey area of the jus contra bellum
The foregoing treaty exegesis has brought to light a grey area of the jus contra bellum where international lawyers can reasonably disagree about the preferable interpretation of the current state of the law. A particular layer of complexity is the fact that recourse to the preparatory work of the UN Charter has yielded indications that point in different directions concerning the right of individual self-defence, on the one hand, and the right of collective self-defence, on the other. While the idea of construing the right of collective self-defence more narrowly would be in line with the historic evolution of the right of self-defence in international law,Footnote 215 and while the enhanced risk of escalation connected with a collective exercise of the right of self-defence is a consideration in support of a differentiated understanding of both forms of self-defence, such a reading of the UN Charter is strongly counter-intuitive in view of the conjoint explicit regulation that both forms of the right of self-defence have received in Article 51. In his recent monograph on the right of collective self-defence, James Green has expressed what, in the abstract, is clearly the predominant position among states and scholars:Footnote 216
The inherent ‘conjoining’ of the two manifestations of self-defence, however, means, that distinguishing two differing understandings of ‘armed attack’ – one applicable to individual and one, at a higher threshold, to collective self-defence – would be unworkable, and contrary to any reasonable interpretation of Article 51.
Yet, the fact remains that the uncertainty surrounding the question of a right of self-defence in the case of a non-state armed attack is of a different kind in respect of either of the two articulations of that right.
Regarding the right of individual self-defence, in the view of the present author, better arguments suggest that this right can be exercised in the case of a cross-border non-state armed attack; the most important two considerations being that a cross-border armed attack may threaten the security of the victim state in the same manner as an armed attack by a state would, and that the host state is no mere bystander in respect of an armed attack emanating from its territory. In view of the preparatory work, one might ask whether this form of the right of individual self-defence should rather be conceived as an inherent limitation on the prohibition of the use of force rather than an exception to it, as articulated in Article 51. However, such a divorce of the right of individual self-defence in the case of a non-state armed attack from Article 51 would seem implausible for three reasons. It would, first, be artificial in view of the fact that the ordinary meaning of this provision covers all relevant scenarios. Second, it appears to be teleologically implausible, as the reporting requirement contained in Article 51 should apply equally. Third, it would be out of touch with the practice of states where the issue today is considered as a matter of interpreting Article 51. The preparatory work of the UN Charter, however, provides a strong argument not to construe the prohibition in Article 2(4) and the permission in Article 51 as two separate norms in the form of a rule–exception relationship, but to read both provisions together as one – admittedly, somewhat complex – primary rule.
Regarding the right of collective self-defence and in view of the preparatory work of the UN Charter, it is submitted by the present author that the better arguments currently repel such a right of collective self-defence in the case of a non-state armed attack – a right that did not exist in pre-UN Charter international law and that is presently not (yet?) supported by general state practice.Footnote 217 It is to be conceded that the matter is not unambiguous either: first, because the purposive consideration underlying the right of individual self-defence in the scenarios under consideration also supports the existence of the right in its collective dimension; second, because one may differ as to the weight to be attributed to the preparatory work of the UN Charter;Footnote 218 and, third, because the conjoint regulation of both dimensions of the right of self-defence has generally animated states no longer to differentiate between the right of individual self-defence and its collective extension.
5.4.6. Situating Operation Iron Swords in the grey area
Israel’s right of self-defence in question can only be individual in nature. Considering the above, on 7 October 2023 Israel had a strong case for exercising that right,Footnote 219 despite the fact that the armed attack by Hamas was one of a non-state organisation.Footnote 220 As we have seen above, while Operation Iron Swords has widely been condemned as an aggression, it is difficult to identify statements that deny Israel the exercise of its right of self-defence specifically because of the non-state nature of the armed attack. It is impossible to explain this fact with certainty. One reason may be that Israel’s critics have found it more convenient to rely on an alleged military occupation to explain their position.Footnote 221 Another possible reason is the peculiarity characterising the factual situation in the Gaza Strip on 7 October 2023 with a non-state organisation controlling the relevant territory more effectively than quite a number of state governments are exercising authority over their respective territory. In such a situation it would seem even harder than in other cases of non-state armed attack to deny the victim state the possibility of exercising its right of individual self-defence.Footnote 222 This consideration appears to have guided the drafters of the Report of the Secretary-General’s Panel of Inquiry on the 31 May 2010 Flotilla Incident, which contains the following key finding: ‘The Panel notes … that the uncertain legal status of Gaza under international law cannot mean that Israel has no right to self-defence against armed attacks directed toward its territory’.Footnote 223
6. On the modalities of Israel’s exercise of its right of self-defenceFootnote 224
The ICJ has consistently taken the view that the exercise of the right of self-defence is subject to the conditions of necessity and proportionalityFootnote 225 and, in its advisory opinion in the Nuclear Weapons case, the Court even considered those conditions to be ‘inherent in the very concept of self-defence’.Footnote 226 This position is essentially uncontested in international legal scholarship.Footnote 227 Yet, as we shall see, the consensus ends when it comes to the contents of those two conditions in question and, in particular, concerning the principle of contra bellum proportionality. In addition, it must be asked whether the inability of the host state – or host people – which situates Israel’s use of force in self-defence in proximity to a state of necessity, has a bearing on how to understand the specific demands flowing from the conditions of necessity and/or proportionality in this instance.
6.1. Legitimate goal
Necessity and proportionality judgements cannot be made without a point of reference. It is therefore imperative, first, to define the legitimate goal which may be pursued through the exercise of self-defence in the given case. Importantly, however, it is not the formulation of an illegitimate goal which, by itself, makes the exercise of the right of self-defence by a state unlawful. What matters is the actual pursuit of such a goal by the use of force under the false pretence of self-defence. If an armed attack has occurred and is ongoing, it is undisputed that forcible action in self-defence may be directed at halting and repelling this attack,Footnote 228 but it is controversial whether the legitimate goal of self-defence invariably stops there. In a seminal article on the topic, David Kretzmer has put forward, based on an analysis of the relevant practice of states, that the victim state may also use force ‘to reduce reasonably foreseeable future threats’, ‘unless the armed attack is limited, localized, and unconnected to a previous “accumulation of events” or war-threatening situation’.Footnote 229 While Kretzmer’s standard has been criticised by Georg Nolte for being too permissive, the latter has conceded that the test of halting and repelling must be applied with some flexibility. ‘Repelling’, as Nolte has written, ‘also extends to measures which bring an end to the armed aggression in a somewhat broader, socially constructed sense’.Footnote 230 It would not appear that these formulations necessarily lead to different results in all too many cases. The area of genuine controversy probably consists of determining how proximate in time the next expected attack must be if it is not imminent.Footnote 231 In that respect, Israel’s use of force under consideration does not constitute a genuinely controversial case. Hamas’ rocket attacks from within the Gaza Strip continued after the ground assault on Israel’s territory had been halted, and to the extent that these rocket attacks stopped for a moment in time, the circumstances were such that Israel had to assume that their continuation was imminent.Footnote 232 This assessment holds true – irrespective of whether holding Israeli nationals as hostages by Hamas, by itself, would qualify as an armed attack within the meaning of Article 51.Footnote 233
6.2. Necessity of recourse to force in the Gaza Strip
At its generally accepted core, the condition of necessity refers to the question of whether resorting to force was without a reasonable non-forcible alternative in order to achieve the legitimate goal of self-defence in the first place.Footnote 234 In the circumstances, Israel could consider prior negotiations with Hamas a futile exercise. Israel also had no reason to expect that the host state or host people were in a position to effectively handle Hamas’ armed attack. Instead, it was obvious that the host state, or host people, was unable to curtail Hamas’ scope of action.Footnote 235
The necessity for Israel’s recourse to force in the Gaza Strip has been questioned, however, on the basis of two considerations. First, it has been pointed out that Hamas' attack on 7 October 2023 demonstrates how Israel’s previous exercises of its right of self-defence in the Gaza Strip have failed to end the violence; instead, it is argued that Israel has heightened the probability of renewed violence against it by fuelling hatred among Palestinians, causing young Palestinian men to join Hamas’ military wing.Footnote 236 There are two reasons, though, why the necessity for Israel’s exercise of self-defence cannot be successfully challenged on this basis. The first reason is that the consideration in question is too speculative in empirical terms to be taken into account as a matter of law. The second reason is that the consideration in question concerns a (possible) wider future, while the point of reference for the necessity judgement is the threat that Hamas’ military capacity constituted for Israel’s immediate future.
Second, it has been argued that Israel could have relied on the non-forcible alternatives of regrouping its armed forces and on strengthening the Iron Dome in order to neutralise the military threat posed by Hamas,Footnote 237 but it remains seriously doubtful, based on past experience, that such an approach alone could have completely neutralised the threat posed by Hamas.Footnote 238 Also, the condition of necessity does not require the victim state to incur a more than negligible risk that the armed attack it has suffered continues in the future. Therefore, it cannot be denied that Israel’s exercise of its right of self-defence was necessary in the first place.
6.3. Proportionality
While the existence of a requirement of proportionality for the lawful exercise of the right of self-defence is not seriously controversial, the precise meaning of this requirement very much is, as was indicated before. In essence, the debate is about whether what is often called ‘strategic proportionality’ sets a limit on forcible defensive action beyond what is reasonably required in order to achieve the legitimate goal of the exercise of the right of self-defence in a given case. As is set out in the following, the position taken in that respect can have very practical consequences.
6.3.1. Proportionality in the functional sense (or necessity in terms of choice of targets and duration)
At a minimum, self-defence proportionality requires that the use of force will not go beyond what is reasonably required in order to achieve its legitimate goal. It has been observed that the proportionality condition, if considered in this way (which is often called the ‘functional approach’), ‘actually becomes an element of the broader necessity requirement, so that the two assessments largely merge into one’.Footnote 239 Indeed, according to the functional approach, the proportionality condition effectively covers both the necessity for the use of force in terms of chosen targets and necessity as far as its duration is concerned. In short, proportionality in the functional sense ensures that a use of force in self-defence, in addition to its having been necessary in the first place, keeps within this necessity for as long as it lasts.Footnote 240 In this sense, the proportionality condition is indeed ‘inherent in the very concept of self-defence’ because this concept could never justify a use of force that over time has transformed into aggressive military action. As has already been noted,Footnote 241 Israel’s apparent legal position – that once a right of self-defence is exercised in a given case, no continuous proportionality assessment in a functional sense is required – is untenable. It is disturbing to read that Israel refers to the functional approach to self-defence proportionality in an official document as proportionality ‘in a colloquial sense’.Footnote 242
In view of the intensity of Hamas' attack on 7 October 2023, the continuation of attacks thereafter, and the stated intent to annihilate Israel, right after the initial attack one could legitimately argue that in order to end this armed attack it was necessary to destroy Hamas’ military capability.Footnote 243 In its application to the ICJ, South Africa, however, submitted that Israel’s use of force in the Gaza Strip has been reflecting, since soon after 7 October 2023, a policy that goes beyond the goal of neutralising Hamas’ military capacity and is, in fact, genocidal.Footnote 244 To substantiate its claim, South Africa submitted that Operation Iron Swords was displaying a ‘clear pattern of conduct’, consisting of, inter alia, ‘the targeting of family homes and civilian infrastructure’.Footnote 245 Assuming that Israel has been pursuing a policy of targeting Palestinian civilians and civilian infrastructure, or that Israel has deliberately chosen a policy of indiscriminate targeting, then its use of force in the Gaza Strip has indeed gone beyond what was necessary to achieve the legitimate goal of self-defence. Under these circumstances, Israel would have violated the condition of self-defence proportionality in the functional sense from early on. This is not inconsistent with the separate existence of the law contra bellum and the jus in bello, for it is one thing to say that each military attack must comply with the targeting rules under the jus in bello and quite another to recognise that a use of force, which follows a strategy of targeting civilians or of indiscriminate targeting, overall exceeds the degree of force necessary to achieve a legitimate self-defence goal.Footnote 246 Yet, Israel has denied South Africa’s accusation on factual grounds. It has stated its determination to respect the law governing the conduct of hostilities in armed conflict and it has attributed the heavy death toll among civilians to the fact that ‘Hamas has systematically and unlawfully embedded its military operations, militants and assets throughout Gaza within and beneath densely populated civilian areas’.Footnote 247 At the time of writing, a judicial determination of the factual situation has not been made, and it appears premature for an outside observer to reach a firm conclusion in this respect.
Whatever the correct assessment of proportionality is in functional terms until that point, a fresh assessment was required as a result of the ceasefire having come into effect as of 19 January 2025. Irrespective of the status of the ceasefire under international law (if any), Hamas had now stated its intent to cease hostilities and to release the hostages on condition that Israel complies with the relevant stipulations as well. While Israel certainly had no reason to consider a mere declaration made by Hamas to be credible, Hamas agreed to the ceasefire conditions under the influence of heavy losses and Israel’s clear intent to resume its military operations should Hamas not follow up on its stated intent. In those circumstances, it is arguable that the complete destruction of Hamas’ military capacity was no longer necessary to end its armed attack. On this assumption, one must further interrogate why the ceasefire collapsed. If Hamas gave Israel reason to believe that it would no longer abide by the terms of the ceasefire agreement, the legitimacy of the goal of destroying Hamas’ (remaining) military capacity would have been revived. However, if the reason was that Israel had unilaterally decided no longer to abide by the terms of the ceasefire – without prior violation by Hamas – a strong argument can be made that Israel’s resumption of military operations in the Gaza Strip as of 18 March 2025 was disproportionate in the functional sense. Obviously, no firm conclusion can be reached without knowledge of the relevant facts.
Also assuming, finally, that after the collapse of the ceasefire, Israel could legitimately pursue the goal of destroying Hamas’ military capabilities, there are, from that moment in time onwards, far stronger indications than before that Israel’s use of force has gone beyond this goal. This is so irrespective of the question of whether Israel pursued a genocidal policy. For the purposes of analysing Israel’s continued use of force under the self-defence proportionality condition in the functional sense, it is relevant that – alongside the resumption of hostilities – the passage of humanitarian aid was completely suspended for some time. Additionally, leading Israeli cabinet members made statements in connection with the so-called (first) ‘Trump Plan’. This strongly indicated the possibility that Israel then was pursuing the goal of effectively deporting the Palestinian population through the creation of inhospitable living conditions in, and permanent Israeli control over, the Gaza Strip.Footnote 248
6.3.2. Proportionality in the quantitative sense
If self-defence proportionality is to be understood in quantitative terms, it requires a balancing between the weight of the damage that the forcible action in self-defence is designed to prevent from occurring and the weight of the damage that the same use of force involves. Such a quantitative understanding of proportionality is often presented as an alternative to the functional approach to self-defence,Footnote 249 but, as we have seen, the latter approach effectively constitutes a necessity condition in terms of chosen targets and duration. As such, it is inherent in the very concept of self-defence and is, in any event, applicable as a minimum condition for a lawful exercise of self-defence. The right question to ask, therefore, is whether there is a proportionality requirement in the quantitative sense which applies in addition to the functional approach to proportionality. The practical consequence can be shown in view of Israel’s use of force under consideration. If understood exclusively in the functional sense, Israel satisfied the proportionality requirement for as long as its use of force was necessary to (further) weaken Hamas’ (remaining) military capability. Incidental civilian damage resulting from such forcible action would be immaterial to the contra bellum proportionality test and relevant only to the application of the proportionality requirement in bello. If, however, use of force in self-defence must also be proportionate in a quantitative sense, it is conceivable that the overall incidental damage has become or will become so extensive that the use of force, though still proportionate in the functional sense (that is, necessary to achieve the legitimate self-defence goal), could no longer be lawfully continued.
Whether self-defence proportionality also has a quantitative dimension in current international law is uncertain.Footnote 250 In his capacity as Special Rapporteur to the ILC on State Responsibility for Internationally Wrongful Acts, Roberto Ago has become the most prominent voice against such a requirement. In 1980, he stated as follows:Footnote 251
The requirement of the proportionality of the action taken in self-defence … concerns the relationship between that action and its purpose, namely – and this can never be repeated too often – that of halting and repelling the attack, or even, in so far as preventive self-defence is recognized, of preventing it from occurring. It would be mistaken, however, to think that there must be proportionality between the conduct constituting the armed attack and the opposing conduct. The action needed to halt and repulse the attack may well have to assume dimensions disproportionate to those of the attack suffered. What matters in this respect is the result to be achieved by the ‘defensive’ action, and not the forms, substance and strength of the action itself. … In fact, the requirement of the ‘necessity’ and ‘proportionality’ of the action taken in self-defence can simply be described as two sides of the same coin.
Conversely, in 2018, the Use of Force Committee of the International Law Association affirmed a self-defence proportionality condition in the quantitative sense in the following terms:Footnote 252
Proportionality is a formula requiring the balancing of two elements against each other. … (T)he proportionality assessment should weigh the relative interests against each other, thereby assessing whether the harmful effects of the force taken in self-defence are outweighed by achieving the legitimate aims.
In order to clarify the matter, the rules of treaty interpretation are of no assistance. The ICJ has occasionally touched upon the matter and, in doing so, has displayed a certain inclination to embrace a possible quantitative dimension of the proportionality condition; however, the Court has yet to articulate a clear and considered line of thinking regarding the issue.Footnote 253 Finally, the practice of states is also not free from ambiguity. The conversation among the 35 elected state representatives in the fourth Special Committee charged with the definition of aggression clearly supports an exclusively functional understanding of self-defence proportionality, which appears to have influenced Roberto Ago’s view of the matter.Footnote 254 In the more recent practice of states, however, one also finds traces of proportionality assessments influenced also by quantitative considerations. Among the factors taken into account by states that made comments to that effect, civilian damage caused by the forcible self-defence action figures prominently.Footnote 255 To the extent that Israel’s use of force under consideration has been criticised for being disproportionate, the relevant statements mirror this more recent tendency. An initial assessment going in this direction was made by Norway’s Prime Minister, who already at the end of October 2023 qualified Israel’s military operation as having gone ‘beyond proportionality’, pointing specifically to the ‘catastrophic humanitarian consequences for civilians’.Footnote 256 Ireland followed suit along the same lines and with remarkable specificity:Footnote 257
(I)nternational law limits the use of force in self-defence to no more than is necessary and proportionate. In our view, these limits are being exceeded, as evidenced by the unacceptable death toll, the massive destruction of property, including homes, throughout Gaza, the displacement of as many as 2 million people and the ensuing humanitarian catastrophe that we have referred to.
On 19 May 2025, the United Kingdom, France and Canada issued a statement that reiterated the support of these countries for Israel’s ‘right to defend Israelis against terrorism’, while dismissing the latest Israeli course of action as ‘wholly disproportionate’, calling on the state ‘to stop its military operations’.Footnote 258 It remains open to argument, however, whether this practice of states qualifies as evidence of their embracing quantitative considerations in assessments of self-defence proportionality in a manner that is sufficiently consistent to justify the conclusion that the self-defence proportionality test now includes a quantitative dimension beyond its recognised functional one. In fact, it is quite apparent that states are hesitant to clearly articulate a standard of self-defence proportionality that is more stringent than the functional approach. There are two likely reasons for this reluctance.Footnote 259 First, it remains controversial which kind of damage should enter into the balance and, more often than not, there will be epistemic challenges to making accurate factual predictions. Second, and perhaps even more importantly, a proportionality condition in quantitative terms involves the possibility that a state must end its use of force before having achieved its legitimate self-defence goal. This is not a prospect that any state will accept lightly. All this explains why quite a few scholars still adhere to Roberto Ago’s reservation, questioning the existence of a proportionality condition in a quantitative sense as a matter of current international law.Footnote 260 Others are of the view that such a condition does exist, but must be defined with great caution. According to this view, disproportionality in quantitative terms can be established only if the harm involved in the use of force taken in self-defence is excessive in the sense that it greatly outweighs the harm to be expected if the armed attack were to continue unabated.Footnote 261
A weighty argument of principle can be advanced to embrace such a cautiously worded text of quantitative proportionality in all cases of self-defence for the benefit of civilians. For even in cases of an armed attack by a state, more often than not the civilian population within that will not bear moral responsibility for the decision made by its government, so that it will be morally questionable to burden that population with a sort of collective responsibility for the aggressive course of action taken by its state.Footnote 262 For present purposes, however, it can be left undecided whether international law has developed to such a point. A sufficiently strong case can be made that such a condition of quantitative proportionality does – at least – apply in the special case of self-defence under consideration vis-à-vis an unable host state/host people of a non-state attacker.Footnote 263 Three principal reasons support this view. First, as we have seen,Footnote 264 in such a case the legal relationship between the victim state of the armed attack and the host state/host people is situated at the borderline between self-defence and a state of necessity because the host state/host people are not responsible for internationally wrongful conduct in respect of the armed attack. The greater proximity to a state of necessity strongly suggests deriving normative inspiration from the fact that the principle of proportionality in the quantitative sense forms part of the regulation under customary international law of a state of necessity as a ground for precluding wrongfulness.Footnote 265 Second, the condition of proportionality in the quantitative sense forms part of the nineteenth century acquis of the law on self-defence, which remains relevant to the current state of law. More specifically, the Mexican note of 15 October 1836, relating to the above-mentioned Mexican–US border incident of that same year,Footnote 266 encapsulates a particularly clear articulation of a condition of proportionality in the quantitative sense. Mexico’s Ambassador Gorostiza wrote:Footnote 267
[self-defence] fixes upon us the obligation of preserving and defending ourselves, (it) equally prohibits us from so doing to the evident injury of a third party, unless in a case of absolute necessity, when the danger is imminent, when it cannot be avoided by any other means, and when the injury apprehended is infinitely greater than that which we are about to occasion’.
Third, the body of more recent state practice, in which references to a requirement of quantitative proportionality are most clearly visible, pertains chiefly to situations of an unable host state/host people,Footnote 268 including the fairly explicit Irish statement regarding Israel’s use of force under consideration, cited above.Footnote 269 Regarding the precise formulation of the proportionality condition in the quantitative sense, Mexico’s statement of 1836 supports a most demanding standard, broadly in line with governing the state of necessity under customary international law. However, this appears to be overly stringent in view of the fact that the unable host state/host people are no mere bystanders, but instead are liable to tolerate the forcible intrusion. In such a situation, the condition of quantitative proportionality is best conceived as requiring that the overall damage caused by the forcible self-defence action does not substantially outweigh the damage caused by the non-state armed attackFootnote 270 and the harm expected to occur in the event of the continuation of the armed attack. In a particularly lucid treatment of the subject, this has been called a principle of ‘soft proportionality’.Footnote 271
It is beyond the purpose of this study to conduct a comprehensive proportionality analysis under the legal standard explained above, especially as such an endeavour could only be tentative at a moment in time when the relevant figures remain a matter of controversy and doubt. In view of the publicly available data, however, it is possible to give some cautious directions. On Israel’s side, all types of damage count, including those suffered in the course of hostilities, as long as the underlying use of force could be considered to be a lawful exercise of the right of self-defence. On the side of the Palestinian host state/host people, all types of civilian damage count, including damage to public institutions. Particular weight is to be attributed to deaths, physical and mental harm (including that stemming from malnutrition) as well as displacements from home. The core of the damage to be compared is civilian in nature, including deaths, injuries, displacement and destruction of property.Footnote 272 Regarding the future damage that Israel had to expect in the event of a continuation of Hamas’ armed attack by use of its full military capabilities, it is to be assumed that Israel would at least have undertaken certain meaningful measures of precaution, such as the regrouping of military forces at the borders with a view to be in a better position than it was on 7 October to deal with possible future attacks. With that in mind, the damage that Israel suffered from Hamas attacks before 7 October provide indicative figures as to possible future damage in the event of a continuation of the Hamas attack.Footnote 273 If these parameters are duly considered, a serious question mark had to be posed behind the proportionality of Israel’s use of force early on, as the number of deaths, injuries and displaced persons on the Palestinian side soon reached a level far beyond the losses that Israel had already suffered and reasonably had to expect in the event of a continuation of the armed attack.Footnote 274 Yet, for two reasons, one could still be hesitant in drawing the conclusion that the losses inflicted by Israel on the civilian population in the Gaza Strip by that time substantially outweighed the past and future harm on Israel’s side of the balance. First, though difficult to measure with precision, it would have been unfair not to include into the equation the deep-seated and widespread sense of vulnerability and helplessness among Israelis as a result of the past attack, and in view of its continuation and the effect of this sentiment on public life in the country.Footnote 275 Second, it cannot be ignored that Hamas’ attack was designed also to serve as a stimulus for Hezbollah and the Houthis to attack Israel as well.Footnote 276 In such circumstances, it is warranted to add this threat to the calculation that Hezbollah and the Houthis posed to Israel in the case of an unabated continuation of Hamas’ attack on that same state.Footnote 277 However, the longer that Israel’s use of force has continued, the higher the losses among the Palestinian population became, all of this alongside a further deterioration in an already dire general humanitarian situation in the Gaza Strip, the more clearly the balance has shifted to Israel’s detriment. Israel’s resumption of military operations in March 2025, in combination with the complete suspension of the passage of humanitarian relief, marks the point in time at which a finding of quantitative disproportionality became inevitable, even for the most cautious outside observer.Footnote 278
7. Conclusion
The Gaza war, which started on 7 October 2023 and which has come to a close only about two years later, is a challenging case as far as the application of the prohibition of the use of force under international law is concerned. As the preceding analysis has shown, this has arisen chiefly from the genuine legal uncertainty concerning the applicability of the right of self-defence when an armed attack by a non-state organisation emanates from the territory of a state that has proven unable to prevent the said armed attack. Arguably, the situation in the Gaza Strip on 7 October 2023 presents the rare variation of such an ‘unable host state scenario’ in which the non-state armed attack (by Hamas) against a state (Israel) originated from a territory (the Gaza Strip) destined for the realisation of the right to self-determination of a people (the Palestinian people). In such a case, the dilemmatic conflict that underlies the uncertainty about the applicability of the right of self-defence is between the legally protected interests of the state that is the victim of the armed attack and the ‘host people’ of the non-state attacker. The present study has shown that Israel, notwithstanding weighty arguments to the contrary, had a strong legal case to exercise its right of individual self-defence through forcible action against Hamas’ military wing in the Gaza Strip. It has also turned out, however, that, from fairly early on, Israel’s actual use of force has given rise to questions relating to the conditions of necessity and proportionality. As Israel’s military operation has been continuing, the relevant doubts have deepened. Since Israel, after the collapse of the ceasefire (which was effective as of 19 January 2025), resumed its military operations in March 2025, it has been highly doubtful whether Israel’s use of force has still been proportionate in a functional sense (by staying within the necessity of achieving the legitimate goal of its self-defence). Finally, it has been set out in the foregoing that, in the case of a non-state armed attack, the exercise of the right of self-defence must satisfy the condition of proportionality in a quantitative sense vis-à-vis an unable host state or host people. With the resumption of its military operations in March 2025 at the latest, Israel’s use of force in the Gaza Strip has been violating this condition. Since then, at least, Israel has exceeded even the outer limits of the right of self-defence and has thus been in breach of the prohibition of the use of force.
Acknowledgements
The essay builds on presentations given on 17 December 2024 in Tel Aviv as part of the 19th Annual Minerva Conference on International Humanitarian Law (digitally) and on 18 March 2025 at the Institut de Recherche en Droit International et Européen de la Sorbonne in Paris. I wish to thank Gal Cohen, David Kretzmer, Eliav Lieblich, Yuval Shany and Svenja Raube for commenting on draft versions of the essay as well as Evelyne Lagrange for numerous illuminating exchanges about the legal issues at stake. I am grateful to Gal Cohen, Olivier Corten, Evelyne Lagrange and Nicholas Tsagourias for allowing me to read in draft form the thoughtful articles they wrote in connection with the subject-matter of this study. I am greatly indebted to my research associate Rosa-Lena Lauterbach for her formidable assistance in the collection of relevant materials. Many thanks go to the following members of my team for critical reading of a draft version of the essay: Elena Barth, Alexander Elfgen, Teresa Fachinger, Amira Kan’an, Keiichiro Kawai, Rosa-Lena Lauterbach and Paulina Rob. It is with great pleasure that I finally acknowledge the impressive work done by the Review’s editorial team in the preparation of the manuscript for publication.
Funding statement
Not applicable.
Competing interests
The author declares none.
Open access
