On Oct. 22, the International Court of Justice (ICJ) delivered its Advisory Opinion on Israel’s obligations towards the presence and activities of third parties (the UN, other international organizations, and states) in the Occupied Palestinian Territory (OPT). The Opinion joins the flurry of international litigation on Israel-Palestine, including two others at the ICJ , last year’s Advisory Opinion on the legality of the occupation (“the Legality Opinion”) and the ongoing South Africa v. Israel case, and the ICC arrest warrants issued against Israeli leaders.
The direct impetus for the UN General Assembly’s request for this Opinion was the enactment of two Israeli laws last year, aiming to cease United Nations Relief and Works Agency for Palestine Refugees (UNRWA)) operations both in “the sovereign territory of Israel” (which, in Israel’s view, includes East Jerusalem), and in the OPT. The latter was effected by Israel’s cancellation of the 1967 “exchange of letters” between Israel and UNRWA (which regulated its continued operations in the OPT post-1967), and by prohibiting all Israeli agencies (including the military authorities in the OPT) from interacting with UNRWA (for the text of the laws, see paras 63-65 of the Opinion; for the exchange of letters, see para 50). In accordance with these laws, UNRWA was compelled to evacuate its East Jerusalem compound; its international staff lost their status and were expelled; and UNRWA has been prohibited from bringing aid into Gaza since the crossings are controlled by Israel. It should be noted that in practice, however, UNRWA continues to operate widely both in Gaza and the West Bank on the basis of its local staff, although its operations are severely impeded. This demonstrates the illogical nature of Israel’s move, since it both “bans” UNRWA while continuing to rely on it. As we shall see, this point proved crucial in the Opinion.
Nonetheless, Israel justified the laws on the basis of its allegations that UNRWA staff were involved in Hamas’s October 7 brutal attacks; that hundreds of its staff are in fact members of terrorist organizations; and by its repeated claims that armed groups are operating in its compounds. A UN investigation found that nine UNRWA employees might have been involved in the October 7 attacks. UNRWA, for its part, denied the allegation that hundreds of its employees are involved in terrorism, stated that it has indeed condemned instances where it was aware that its premises were used by armed groups, and that in general, none of the allegations support a blanket description of the “institution as a whole” – which employs 30,000 people – as infiltrated by Hamas. Critics of Israel claimed that the real motivation behind Israel’s move was its hostility towards the mere recognition of Palestinian refugees and their rights. Others were concerned with the effects of the restrictions against UNRWA on the humanitarian situation on the ground, given its central role in providing services and aid to Palestinians.
The Request for an Advisory Opinion
Against this background, the General Assembly requested the Court to provide an Opinion on
“the obligations of Israel, as an occupying Power and as a member of the United Nations, in relation to the presence and activities of the United Nations, including its agencies and bodies, other international organizations and third States, in and in relation to the Occupied Palestinian Territory, including to ensure and facilitate the unhindered provision of urgently needed supplies essential to the survival of the Palestinian civilian population.”
Three things should be pointed out about this question from the outset: first, unlike the Legality Opinion, the ICJ was not requested here to determine directly whether Israel violated its obligations. Second, the question concerns obligations towards the UN and other third parties in occupied territories, not in a state’s sovereign territory. Third, the request focused on Israel’s obligations both as an occupying power – meaning, on the basis of the laws applying in occupation – and as a UN member state per se. This opened the door to important and interesting discussions on the obligations of states to cooperate with the UN more generally.
The Opinion itself includes determinations on many legal questions, including issues such as international human rights law; obligations under international humanitarian law towards detainees; self-determination; the responsibility of the UN regarding the question of Palestine; and immunities of the United Nations. Sometimes, owing to the formulation of the General Assembly’s question, these determinations are rather general and restate well-established obligations. Moreover, some determinations are relevant to the specific context and history of Israel-Palestine, while others have more general application.
In this piece, I will focus on two key aspects that are both central to the issue and received detailed treatment in the Opinion: the occupying power’s obligations to provide essential needs, and how these interact with the activities of international organizations; and the general duty of states to cooperate with the UN. The latter question spawned a debate between the judges in their individual declarations, which gives rise to much broader theoretical and jurisprudential questions on the role and nature of the UN. I will not address the question of immunities here. Because of its rather technical nature, it should be addressed in a separate and detailed piece.
Indeed, the Opinion displays consensus on most issues. The sole operative dissenting opinion, on some points, is by Vice-President Sebutinde – an opinion that requires a separate analysis. I would also like to call attention to Judge Cleveland’s important declaration which includes significant points concerning the duties of non-state actors such as Hamas in relation to the issues in question, while emphasizing the non-reciprocal nature of these obligations.
The Status of Gaza
The scope of a state’s duties concerning humanitarian aid in armed conflict depends on whether a territory is occupied or not. In hostilities absent an occupation, belligerents are only required to allow and facilitate aid, subject to their right to prescribe technical arrangements (see the new ICRC commentary here). However, once a territory is under the effective control of the adverse party, the law of occupation kicks in, and “positive” duties to provide aid apply (see here). More generally, the occupant must ensure public order and security in the occupied territory. The key question that the Court had to grapple with, in this context, was how these duties interact with the activities of third parties in the territory, including the UN, other international organizations, and states.
First, the Court had to determine the status of the relevant territories. Regarding the West Bank and East Jerusalem, there is a longstanding international consensus on their status as occupied territory – a status which the Court had already validated in two previous Advisory Opinions – and which it only had to briefly repeat in the current Opinion (para 83). Regarding Gaza, the situation became more complicated after Israel’s “disengagement” in 2005. In the Legality Opinion, the Court briefly held that “Israel’s withdrawal from the Gaza Strip has not entirely released it of its obligations under the law of occupation” and these “remained commensurate with the degree of its effective control over the Gaza Strip (Legality, para 94);” and moreover, that this is “even more so since 7 October 2023” (Legality, para 93).
In the new Opinion, the Court holds that in the months since the Legality Opinion, and owing to its military presence on the ground and other actions, “Israel’s effective control over the Gaza Strip has increased significantly,” and that accordingly “Israel’s obligations under the law of occupation have also increased significantly, commensurate with the increase in its effective control over the territory.”(Para 86)
This is yet another manifestation of the “functional approach” to occupation (espoused also by the ICRC), which views occupation not as a binary set of obligations but as a sliding scale of obligations, the existence and scope of which are derived from the belligerent’s actual control over governmental functions (for the definitive theoretical treatment, see Aeyal Gross here).
This determination prompted a partial dissent from judge Gómez Robledo, who opined that the Court should have flat out ruled that Gaza “is now under full-scale occupation” (para 3). Considering the scope of Israel’s ground presence in Gaza at the time, this dissent is understandable. It was also my own opinion (along with several colleagues) that at least in July 2025 “Israel is considered an occupying power in all or nearly all of Gaza.”
Nonetheless, the functional approach has its own non-escalatory merits, particularly considering the current fluctuating situation on the ground. For instance, if we take seriously the all-or-nothing approach to occupation, we might argue that Israel, as the occupant, now has to re-enter parts of Gaza to prevent retribution by Hamas against Palestinian citizens, notwithstanding the ceasefire in effect. Indeed, like all of the laws of war, the law of occupation has a facilitative aspect which can be used to legitimate force that we should not lose sight of.
Be that as it may, since Israel undoubtedly controls the flow of humanitarian aid into Gaza, whether one adopts the functional or binary approach to occupation is less consequential here.
Humanitarian Obligations under Occupation, Third Parties and UNRWA
Having reiterated that the West Bank and Gaza (functionally) are occupied, the Court moved to analyze the specific provisions of international humanitarian law (IHL) regarding humanitarian access in occupied territories, as they relate to activities by third parties. Here, the Court’s reasoning provides much welcome clarity on a complex set of norms.
The Court constructed these obligations in the following way: Articles 55 and 56 of the Fourth Geneva Convention (GCIV) impose positive obligations on occupying powers to ensure that the population is supplied with essential needs. In cases where, for whatever reason, the population is “inadequately supplied,” additional obligations – namely those included in Article 59 – come into play (para 92). Article 59 holds that in such cases, the occupying power must agree, in addition to its positive obligations to ensure the needs itself, to “relief schemes” and facilitate them. These may be undertaken either by states or impartial humanitarian organizations.
As the Court ruled, when the population is inadequately supplied, the obligation to allow relief schemes is “unconditional,” and it triggers direct obligations between the occupying power and third states and international organizations that seek to provide aid. However, this obligation applies only if these actors are impartial (paras 94-95). Importantly, although Article 59 does not mention this explicitly, the Court held that owing to its control over the territory, the occupant has a right to inspect aid and be satisfied that it will indeed be delivered to the population (para 97). This makes sense, since the occupant is generally entitled to take security measures in the territory. Moreover, since a belligerent is entitled to prescribe technical arrangements regarding aid to non-occupied territory, it would have been strange if it were not permitted to do so in territory under its de facto control.
In this context, the Court makes a crucial determination that might be overlooked. Throughout the Gaza conflict, Israel has claimed that it is permitting a significant amount of aid into Gaza, but it does not reach the population since the UN fails to collect it from the Gazan side of the border. Whatever the factual merits of these claims, the Court clarified that the duty to provide aid does not end simply when it crosses the border, and that in cases of occupation,
“an occupying Power must do more than simply allow the passage of essential items into the occupied territory. It must also use all means at its disposal so that these items are distributed in a regular, fair and non-discriminatory manner, including by facilitating access to them” (para 101).
Having determined that, the Court turned to analyze the application of Article 59 in the case at hand. In Gaza, it relied heavily on international reports such as by the World Food Programme, Integrated Food Security Phase Classification (IPC), World Health Organization and the UN’s Office for the Coordination of Humanitarian Affairs (OCHA), and noted Israel’s complete block of aid from March to May 2025, as well as the inadequate operations of the Gaza Humanitarian Foundation, to conclude that the population in Gaza has indeed been inadequately supplied (paras 102-109) (the Court does not elaborate on the question of the prohibition on starvation, beyond recalling it briefly. See paras. 143-145).
As a side note, I was a bit surprised by the centrality of the “inadequately supplied” standard in the Court’s analysis. This is both because too much emphasis on this test facilitates abuse by occupying states (and belligerents in general – since similar language is found also concerning non-occupied territories, see Art. 70(1) of Additional Protocol I), as these might exploit the indeterminacy of the standard; and, more fundamentally, the standard is problematic because in most cases a belligerent would object to third party aid precisely because it wants to cause scarcity.
Nonetheless, as the Court found that the people of Gaza are inadequately supplied, and Israel thus must agree to relief schemes by third parties, the question turned to UNRWA. Here, Israel claimed that it has no obligations to work with UNRWA since it is not “impartial,” as Article 59 requires (para 111).
While “impartiality” is usually understood as non-discrimination in the distribution of aid, the Court, rightfully in my view, read into the concept of impartiality also the requirement of “neutrality,” meaning, that the organization is not “taking sides” in the conflict (para 113-116). Here, the Court looked at Israel’s allegations against UNRWA, and reached the conclusion that these cannot establish UNRWA’s lack of neutrality “as an organization.” Simply put, the Court’s view was that in quantitative terms, there are no findings that can implicate an entire organization of 30,000 employees (para 118).
In this context, Israel also argued that even if Article 59’s requirement to permit “relief schemes” applies, it has no obligation to cooperate with a specific organization such as UNRWA. The Court agreed that in general an occupying power is free to choose the organization it works with (paras 119-120) (a finding that is in tension, in my view, with its later findings regarding the duty to cooperate with the UN, and leaves many open questions concerning the treatment of third states that seek to offer aid). However, it held that this “discretion” must ultimately be assessed in light of the occupant’s obligation to ensure the welfare of the population. In the specific context of Gaza, the Court noted the widescale and crucial humanitarian operations by UNRWA, and coupled with the fact that Israel has not moved to replace it with any credible mechanism, led to the conclusion that Israel cannot ban UNRWA from Gaza without violating its duties as an occupant (120-124). The Court also extended this reasoning, without much elaboration, to other international organizations and third states (paras 125-127).
In my view, the Court could have reached a similar conclusion through the general rules of the law of occupation, in particular the duty to maintain public order (Article 43 Hague Convention). Arguably, the decision to abruptly ban an agency so fundamental to the welfare of the local population simply does not fit with the occupant’s duty to maintain public life.
Concerning the West Bank, the Court added an additional important clarification. In the West Bank, since there was no claim that the population is “inadequately supplied” in the acute sense of Article 59, it could be argued – as per the Court’s own understanding of Article 59 – that Israel is under no obligations to allow third parties to provide aid there. The Court, however, rejected this possibility, based on the specific and longstanding practice of the Israeli occupation. As it noted, in the OPT, Israel “left the responsibility” to provide the needs enumerated in Articles 55-56 of GCIV “to the United Nations acting through UNRWA, as well as other international organizations and third States.” Since it chose to do so rather than to provide these services itself, it cannot now restrict their activities (para. 132).
In other words, Israel can’t have it both ways: it cannot delegate its positive responsibilities to international bodies, while at the same time arguing for an absolute power to restrict them. The Court’s key finding here is that since the reliance on international bodies has been “central” to Israel’s performance of its obligations, it is obligated either to “facilitate those operations or to otherwise ensure that these obligations are fully met” (para 133). Since Israel has no credible plans to take over UNRWA’s operations, and considering the extent of the operations of UNRWA and other international bodies on the ground, their activities cannot be banned or impeded.
Again, this conclusion can be reached also by applying the general obligations of occupying powers (as per Article 43, Hague Convention), and in fact reflects a broader, “public law” way of thinking, stressing the actual reliance of the local population on the services of international bodies in determining the scope of the legal obligation.
Crucially, the analysis above is strictly concerned with the law of occupation, before the Court addressed the specific duties of Israel towards the UN as a member state. As we shall see, in the view of the Court, these duties further restrict its discretion in terms of external aid. However, as opposed to the Court’s clear analysis of the former, its treatment of the duty to cooperate leaves much space for speculation.
The Duty to Cooperate with the UN
The Court analyzed Israel’s duty to cooperate with the UN in light of specific aspects of the Israeli-Palestinian conflict. It noted the special responsibility of the UN towards the question of Palestine, inter alia through the operation of UNRWA; as well as the catastrophic war in Gaza which also affected UNRWA’s premises and personnel, killing hundreds of its employees and people sheltering in its facilities (para. 166-69). Nonetheless, the question of the scope of States’ obligations to cooperate with the UN is much wider. In a sense, it goes to the root of fundamental questions on the nature of the UN, and whether it is a quasi-constitutional body representing the international community as whole, or, rather, an international organization strictly empowered by States to perform particular roles.
In this context, Article 2(5) of the UN Charter establishes the duty to cooperate with the UN, by providing that in pursuit of the UN’s purposes, “All Members shall give the United Nations every assistance in any action it takes in accordance with the present Charter.” In situations of occupation, for instance, one could argue that the duty to cooperate with the UN delimits the occupant’s discretion while fulfilling its obligations under the law of the occupation, such as in Article 59 of GCIV. This would limit, for instance, the ability of the occupant to refuse to allow UN bodies to participate in aid operations.
The question, of course, is the scope of this duty to cooperate. Does it arise only when the organization seeks to fulfill a legally binding UN decision (such as a Security Council resolution)? Does it apply only in conjunction with other pre-existing binding rules, such as the rules of IHL? Or rather, is it an all-encompassing duty to assist the UN?
The Court left this question a bit muddled, owing perhaps to the disagreements revealed in some of the judges’ individual declarations. As it held, the duty to cooperate “must be read together with the provisions of the Charter relating to the powers of various organs of the United Nations” (para 172), a formulation that can be understood in numerous ways.
The specific application of the duty to the Israeli-Palestinian context was similarly vague in its reasoning. The Court emphasized that since Israel is not sovereign in the OPT, it cannot invoke any sovereign prerogative not to cooperate with the UN in these territories (paras. 176, 184). While this is hardly objectionable, it is still a far cry from a positive theory on the scope of such cooperation. Ultimately, the Court held that since UNRWA has been entrusted by the General Assembly to provide relief to Palestinians, and considering its crucial role on the ground, Israel cannot decide unilaterally to withhold its cooperation with the UN. On this basis, it held that Israel must, under Article 2(5), cooperate with the UN and not obstruct its activities (paras 177-79). So, it seems that in the view of the Court, the fact Israel is not sovereign in the territories, coupled with the special legal and factual context of UNRWA’s operations, results in an obligation to cooperate with it.
The individual declarations and opinions by the judges highlighted the scope of disagreement within the Court on this issue. Judge Gómez Robledo, in a partially dissenting opinion, criticized the Court for adopting a “timid and excessively formalist” approach, by not explicitly specifying that the obligation to cooperate entails an obligation (in certain circumstances) “to lend assistance to the General Assembly in any action undertaken by it” (para 7). He further asserted that the Court actually limited the duty to cooperate to apply “only within the narrow framework of strictly binding obligations imposed by certain bodies, first and foremost the Security Council” (para. 8). While I’m doubtful about this reading of the Court’s opinion (this reasoning is more explicitly found in Sebutinde’s dissent, para. 91), it is true that the Court stopped much short of what Gómez Robledo suggested: to recognize that the General Assembly “embodies the will of the international community as a whole” and that the United Nations “synthesizes the civitas maxima ideal, the ultimate source of international law” (para. 8).
Of course, some would immediately object that this in effect confers binding power to the General Assembly. Judges Abraham and Cleveland, for their part, also lamented the lack of clarity in the Court’s reasoning on the duty to cooperate (para. 11). However, they advanced a much narrower view of the duty, warning that a broad view would “bestow on the Article [article 2(5)] an extraordinary power to transform non-binding ‘actions’ of United Nations organs into legally binding ones” (para 6). They thus held that the duty “necessarily operates in conjunction with the legal obligations that a Member State has undertaken.” These can emanate from UN Security Council obligations, and also from other sources such as rules of IHL, IHRL, and other applicable law(paras. 5, 11). Abraham and Cleveland, however, do not specify the technique through which these “external” obligations are transformed into duties to cooperate under Article 2(5). To be clear: there are ways to make such an argument, but the judges do not address it here.
Judges Charlesworth and Xue take an intermediate path. Judge Charlesworth decries the Court’s “cryptic” reasoning on the duty (Art. 3), and reads certain passages of the Court’s seminal Reparation for Injuries Suffered in the Service of the United Nations case as imposing a duty to cooperate that potentially goes “beyond the four corners of the Charter” and the confines of binding Security Council resolutions (para 5). However, Judge Charlesworth stresses that General Assembly Resolutions cannot be made binding through this interpretation; in her view, the obligation under Article 2(5) is simply one of open-ended “good faith cooperation” with the activities of the UN (para 6). Judge Xue, in an impressive purposive and historical analysis of Article 2(5), similarly understands it as a “general obligation” to give assistance to the UN, even absent binding decisions or pre-existing external legal obligations (paras. 4- 24).
There is much space for discussion here. To me, the duty to cooperate can be conceptualized as an obligation not to arbitrarily refuse to assist the UN in its activities under the Charter, in areas which are not under a State’s sovereign prerogative. But one can think of many other formulations close to that effect.
In sum, the Advisory Opinion provides much clarity on some aspects, chiefly relating to humanitarian obligations under the law of occupation, while maintaining ambiguity on others, such as the scope of the duty to cooperate with the UN. One thing is for certain: the Opinion will surely fuel debates for years to come.
