When Guidance Becomes Fog: The ICJ and the Normative Uncertainty of the Duty to Co-operate

For a long time, the obligation to co-operate occupied an ambiguous position in international law. It appeared in diverse treaty regimes and institutional contexts, yet it was seldom recognised as an autonomous, self-standing obligation, i.e., as a general duty of customary nature, underpinning the various legal regimes, and as a genuine legal obligation in its own right. One might therefore have expected that the recent advisory opinions delivered by the International Court of Justice (‘ICJ’ or ‘Court’), most notably the 2024 opinion on Legal Consequences arising from the Policies and Practices of Israel in the Occupied Palestinian Territory, including East Jerusalem (‘Legal Consequences advisory opinion’), the 2025 opinion on Obligations of States in respect of Climate Change (‘Climate Change advisory opinion’), and the 2025 opinion on Obligations of Israel in relation to the Presence and Activities of the United Nations, Other International Organizations and Third States in and in relation to the Occupied Palestinian Territory (‘UNRWA advisory opinion’), would offer the normative clarification long lacking.

This post thus examines how the Court has articulated the duty to co-operate across its recent advisory jurisprudence and demonstrates that, rather than resolving the ambiguity, the Court’s reasoning across these decisions has created a more complex and less coherent picture. It first analyses the different models of co-operation emerging from the ICJ’s opinions, before turning to the implications of this variability for the scope, content, and practical operability of the duty.

Co-operation in the ICJ’s Recent Advisory Jurisprudence

In its 2024 Legal Consequences opinion, the Court found  that all States must co-operate with the UN in bringing Israel’s illegal presence in the occupied Palestinian territory to an end  and securing the full realisation of the Palestinian people’s right to self-determination [p. 74, para. 275]. This transformed what was formerly conceived in the 2004 advisory opinion on the Legal Consequences of the Construction of a Wall in the Occupied Palestinian Territory (‘2004 Wall advisory opinion’) as a set of negative duties into a collective [p. 196, para. 146], affirmative obligation of erga omnes character. Notably, however, while the Court had identified the right of self-determination as a peremptory norm of international law, it did not ground the duty to co-operate explicitly in the jus cogens character of that norm [pp. 65-6, para. 233]. Instead, it relied on the erga omnes nature of the obligations breached, recalling that such obligations are ‘the concern of all States’ and that, ‘in view of the importance of the rights involved, all States can be held to have a legal interest in their protection’ [p. 74, para. 274]. This reasoning is potentially confusing, as it appears to attach to the erga omnes character of the obligations the very consequences—non-recognition, non-assistance, and co-operation—that the International Law Commission (‘ILC’) has consistently associated with serious breaches of jus cogens under Articles 40 and 41 of the ILC Draft Articles on State Responsibility [p. 112, paras. 1-5; cf. also ILC Draft conclusions on identification and legal consequences of peremptory norms of general international law (jus cogens), p. 70, para. 1]. By grounding the duty to co-operate in the erga omnes dimension rather than explicitly in the peremptory character of the norm, the Court risks conflating analytically distinct categories [cf. here and here; for a different assessment, cf. here]. This ambiguity becomes more apparent when the Court’s reasoning is situated against the ILC’s own cautious approach. The ILC expressly acknowledged that ‘it may be open to question whether general international law at present prescribes a positive duty of cooperation’, observing that paragraph 1 of Article 41 may reflect an element of progressive development [p. 114, para. 3]. At the same time, the ILC emphasised that co-operation—especially within the framework of international organisations—is already practised in response to the gravest breaches of international law [ibid.]. This cautious formulation reflected the ILC’s reliance on a normative assessment of emerging practice, pointing to sufficient material to support the progressive consolidation of the obligation—an approach that has since evolved. In its 2022 Draft Conclusions on the identification and legal consequences of peremptory norms of general international law (jus cogens), the Commission observed that ‘obligation to cooperate to bring to an end serious breaches of obligations arising under peremptory norms of general international law (jus cogens) is now recognized under international law’ [pp. 70-1, para. 2]. Against this background, the Court’s approach in 2024 marks a significant departure from its position in the 2004 Wall advisory opinion, where it deliberately confined itself to collective duties of non-recognition and non-assistance [p. 196, para. 146]. Thus, the Court reflected the same caution acknowledged by the ILC. By contrast, in the 2024 Legal Consequences opinion, the Court employs unequivocally deontic language, stating that ‘all States must co-operate’ [emphasis added]. This formulation suggests a shift in the Court’s articulation of collective responsibility towards a more assertive framing of co-operation as a positive requirement. In doing so, the Court’s reasoning increasingly mirrors the structure of the ILC’s framework on collective responses to serious breaches, even though the ICJ invokes a different doctrinal trigger.

The Climate Change advisory opinion is notable for one of the Court’s clearest acknowledgements of a customary duty to co-operate, albeit in a carefully circumscribed form. The ICJ observes that international co-operation is indispensable in the field of climate change and that ‘the customary duty to co-operate for the protection of the environment is reflected in several provisions of the climate change treaties’ [p. 70, para. 215]. At the same time, the Court adds that ‘State co-operation is a paramount principle in solving global problems’ [p. 82, para. 261]. However, this duty is presented not as a general obligation of international law, but rather as a sector-specific rule grounded in the particular interests at stake, i.e., erga omnes, thereby reflecting the Court’s caution in recognizing a customary duty only where State practice and opinio juris are especially dense, while avoiding any claim of a general duty to co-operate. Accordingly, while the Court’s language may signal openness to a more substantial role for co-operation in addressing global challenges, it ultimately falls short of recognizing a general obligation under international customary law. This restrained approach stands in sharp contrast to the Court’s reasoning in its 2024 Legal Consequences advisory opinion. If the Court‘s reasoning in the Legal Consequences opinion—attaching the duty to co-operation to the erga omnes character of the breach—were taken at face value, the Court’s identification of erga omnes obligations in the Climate Change advisory opinion should logically have entailed a corresponding duty to co-operate [cf. Declaration of Judge Tladi, p. 10, paras. 36-7]. Yet, the Court did not articulate any such consequence in that context, nor does it explain why the duty to co-operate does not arise [ibid.]. The result is a jurisprudence in which the triggering conditions and legal basis of the duty to co-operate remain uncertain, oscillating between a binding consequence of erga omnes breaches and a context-specific principle confined to particular regulatory domains [ibid.].

In its 2025 UNRWA advisory opinion, the Court grounded the duty to co-operate in Article 2(5) of the UN Charter, holding that ‘Israel may not obstruct the functions of the United Nations and must provide every assistance in any action taken by the Organization’ [p. 55, para. 179]. This formulation expressly combines a negative obligation (non-obstruction) with a positive obligation (the provision of  ‘every assistance in any action’), reflecting the categorical wording of Article 2(5) itself [p. 54, para. 172]. At the level of principle, the Court thus appears to endorse a broad and absolute duty of co-operation with UN organs and agencies. However, the Court does not translate this abstractly broad obligation into concrete operational requirements. Owing to the wording of the UN General Assembly’s request, the Court confined its analysis to the obligations of Israel alone, without addressing the corresponding duties of UN member States more generally, and without determining whether Israel had in fact breached its obligations [p. 33, para. 77]. As a result, the opinion leaves unresolved how Charter-based duties of co-operation relate to a broader framework of collective responsibility applicable to serious breaches of jus cogens. This silence is especially striking when contrasted with the approach taken by the ILC, which expressly addressed the relationship between institutional co-operation and serious breaches of peremptory norms. The ILC found that, ‘where an international organization has the discretion to act, the obligation to cooperate imposes a duty on the members of that international organization to act with a view to the organization exercising that discretion in a manner to bring to an end the breach of a peremptory norm of general international law (jus cogens)’ [pp. 75-6, para. 11; emphasis added]. This omission may reflect a reluctance on the part of the Court to articulate the full systemic implications of Charter-based co-operation for collective responses to jus cogens violations.

Scope and Content of the Duty to Co-operate

The Court’s jurisprudence further confirms that the scope and content of the duty to co-operate vary significantly depending on their legal basis. The divergent scope and content of the duty to co-operate across these contexts may arguably not be accidental and reflect the distinct legal bases and functions from which the obligation arises. However, this functional differentiation is accompanied by a pronounced asymmetry in normative clarity, particularly in cases where co-operation is grounded in serious violations of jus cogens and in constitutional obligations of UN membership.

Where co-operation is invoked in response to an internationally wrongful act, as in the 2024 Legal Consequences opinion, the duty is conceived primarily as an enforcement-oriented collective response aimed at bringing an unlawful situation to an end. Yet, precisely because of the gravity of the norms at issue, the absence of articulated standards is especially problematic. The Court elevates co-operation to a binding, affirmative obligation incumbent upon all States, while at the same time failing to specify its modalities, limits, or evaluative criteria. Unlike in the climate change context, no benchmark—such as due diligence, reasonableness, or proportionality—is offered against which State conduct might be assessed. The result is a duty that is normatively maximalist but operationally opaque, raising doubts as to whether States are expected to take concrete positive measures, to act through specific institutional channels, or merely to support collective processes in a general sense. 

A comparable lack of normative clarity characterizes the duty to co-operate derived from institutional obligations under the UN Charter, as illustrated by the UNRWA advisory opinion. While the Court grounds the obligation firmly in Charter provisions, it does not translate this abstractly broad obligation into concrete operational requirements. It invokes the language of ‘every assistance’, yet it refrains from specifying what forms such assistance must take, the degree of effort required, or the criteria by which compliance is to be assessed. As a result, the Court’s reasoning appears to be centered on the prohibition of obstruction, thereby leaving the positive component of the duty largely undeveloped. The UNRWA advisory opinion also raises a further issue concerning the scope of co-operation. In deriving the duty from Article 2(5) of the UN Charter, the Court implicitly interprets the phrase “any action” to encompass UNGA resolutions [p. 55, paras. 175, 177]. Notably, however, the Court does not engage with the non-binding character of such resolutions. This omission prompted disagreement among ICJ judges. Vice-President Sebutinde [p. 29, para. 91] and Judges Abraham and Cleveland [p. 2, para. 6] rejected this broader interpretation, arguing that the duty to co-operate is limited to binding “actions” of the UN, such as Security Council resolutions adopted under Chapter VII. In their view, extending Article 2(5) to non-binding GA resolutions would effectively transform such resolutions into binding obligations. This restrictive reading, however, conflates the implementation of resolutions with the distinct duty to co-operate, thereby depriving Article 2(5) of independent normative content. As the Court itself emphasises, the duty to co-operate is designed to ensure the effective fulfilment of the UN’s mandate by requiring States to refrain from obstructing its operations. This broader interpretation was endorsed by Judge Xue [p. 3, para. 11] and is implicitly supported by the Court’s reasoning [ibid.]. Yet, by failing to address the disagreement explicitly or to clarify the limits of Article 2(5), the Court leaves unresolved a central question regarding the scope of Charter-based co-operation, reinforcing the broader pattern of normative ambiguity that characterises its recent jurisprudence on the duty to co-operate.

By contrast, in the climate change context, the Court articulates the duty to co-operate with considerably greater precision. There, co-operation serves a preventive and managerial function, aimed at protecting a global common good through sustained collective action. This functional orientation allows the Court to frame co-operation as an obligation of conduct, subject to a due diligence standard, and to identify concrete forms of co-operation—financial assistance, technology transfer, and capacity-building—while accommodating differentiated capacities. The relative clarity of this model only accentuates the indeterminacy of the co-operation duties articulated in the jus cogens and UN Charter contexts.

Co-operation without Coherence?

Taken together, the Court’s jurisprudence suggests that functional differentiation according to the purpose of the co-operation may help explain why the content varies across legal regimes. In particular, the Court appears to distinguish between co-operation as collective enforcement; co-operation as institutional facilitation, inherent in membership of the UN; and co-operation as preventive and managerial governance, aimed at protecting global common goods. While differing functions may justify variations in the modalities of co-operation, they do not explain why the Court grounds this duty in different legal bases without clarifying how these bases relate within a coherent framework. Yet, from the perspective of systemic integration—reflected in Article 31(3)(c) of the Vienna Convention on Treaty Law—there is at least a presumption that where the same obligation is invoked across different contexts, obligations grounded in different legal bases should be capable of operating coherently within a single, unified legal order, rather than giving rise to markedly divergent scopes and contents, a presumption the Court leaves largely unexplored. In the two contexts where the stakes are arguably high—serious breaches of peremptory norms and the constitutional functioning of the UN—the obligation is framed in emphatic, even absolute, terms, yet remains strikingly under-specified in its scope, modalities, and legal consequences. As a result, the duty to co-operate emerges not as a unified legal obligation capable of systematic application, but as a context-dependent construct whose legal basis, intensity, and operational requirements fluctuate across regimes.

This divergence can be rationalized by reference to the distinction between co-operation as a secondary obligation, triggered by an internationally wrongful act, and co-operation as a primary obligation, flowing directly from treaty law, custom, or institutional membership. On this view, the duty articulated in the 2024 Legal Consequences advisory opinion would be secondary in nature, whereas the duties identified in the Climate Change and UNRWA opinions would operate as primary obligations. Yet, this distinction ultimately concerns the mode of triggering the obligation, not its substantive content. It does not explain why co-operation articulated as a response to an internationally wrongful act—arguably the context in which the need for legal guidance is most acute—is framed in emphatic yet under-specified terms.

A further explanation may lie in the political sensitivity of the contexts in which the Court is operating. In the Palestine-related advisory opinions, including UNRWA and Legal Consequences, the articulation of concrete positive duties of co-operation would risk being perceived as prescribing specific courses of action in an ongoing and highly contested situation. The Court’s reluctance to specify the content and limits of the duty in these cases may therefore reflect a form of judicial self-restraint, aimed at preserving institutional legitimacy and avoiding accusations of political overreach. By contrast, the climate change context, while politically salient, concerns a structurally different problem, namely a diffuse, long-term global risk governed by established treaty frameworks and technocratic modes of co-operation. In that setting, the Court appears more willing to articulate co-operation with doctrinal precision, framing it as an obligation of conduct subject to due diligence and identifying concrete modalities of compliance. 

If political sensitivity does influence the Court’s willingness to specify the content of the duty to co-operate, however, this only reinforces the normative concern. Allowing the clarity of legal obligations to fluctuate according to political context risks transforming co-operation into a variable standard whose legal content depends less on principle than on institutional prudence. 

Conclusion 

Across its advisory opinions, the Court invokes co-operation as a norm of considerable significance, yet leaves its source, scope, and content insufficiently specified, rendering the obligation susceptible to competing and potentially inconsistent interpretations. This lack of clarity is particularly striking given the normative weight the Court itself attributes to co-operation, not merely as a programmatic aspiration but as a constitutive element of what is often described in the literature as the transition from a traditional ‘law of co-existence’ to a more integrated ‘law of co-operation’ [n 1].

If co-operation is to function as a genuine legal obligation, it requires a more coherent and internally consistent doctrinal framework. Absent such clarification, co-operation risks remaining an open-ended concept whose legal force varies unpredictably across contexts, undermining both its normative authority and its practical utility. The fragmentation observed in the Court’s recent advisory opinions thus does not merely reflect doctrinal nuance, but raises broader concerns about the operability of co-operation as a binding legal standard.

A more holistic and systematizing engagement with the duty of co-operation would have been preferable, as the ICJ instead permitted fragmented formulations to emerge across legal contexts devoid of doctrinal coherence. The consequences of this uncertainty for State compliance should not be underestimated. Where the content and scope of an obligation remain indeterminate, States are afforded significant latitude to interpret their duties narrowly, particularly in politically sensitive contexts. This risk is amplified in a period marked by resurgent unilateralism, protectionism, and strategic disengagement from multilateral frameworks. In such a climate, a duty to co-operate articulated in indeterminate terms may struggle to guide and constrain State behavior.