The End of Treaty Reservations as We Know Them? Not Yet – Reflections on Sudan v. UAE and the Normative Weight of the Dissent

[Dr Veronica Botticelli is a Postdoctoral Researcher in International Law at the University of Milan, Italy. She obtained a PhD from University of Padua.]

On 5 May 2025, the International Court of Justice (ICJ, ‘the Court’) issued an order rejecting Sudan’s request for provisional measures in the case concerning the Application of the Convention on the Prevention and Punishment of the Crime of Genocide in Sudan (Sudan v. United Arab Emirates). By a 14-2 majority, the Court found it lacked prima facie jurisdiction due to the UAE’s reservation to Article IX of the 1948 Genocide Convention. By a narrower 9-7 vote, the Court also removed the case from its General List, effectively terminating the proceeding.

The factual matrix of the case has been recalled extensively elsewhere (see, e.g., here and here). Briefly, Sudan’s application alleged that during a 58-day siege of El Geneina between April and June 2023, the Rapid Support Forces (RSF) committed widespread atrocities, including extrajudicial killings, ethnic cleansing, mass displacement, rape, and the systematic destruction of villages. Sudan further claimed that the UAE materially supported the RSF by providing financial and logistical aid, deploying agents to command RSF units, recruiting mercenaries, and supplying advanced weaponry that enabled acts it characterizes as genocidal.

As some scholars had anticipated (see, e.g., here, here, and here), the Sudan v. UAE case was expected to encounter, and eventually succumbed to, significant procedural hurdles at the preliminary stage before the Court. The main obstacle was the UAE’s reservation to Article IX of the Genocide Convention, which excludes the ICJ’s jurisdiction over disputes related to the treaty. Faced with this, the Court had two options: it could either reject provisional measures while continuing to examine its jurisdiction, or – as it ultimately did – dismiss the case outright, citing a “manifest lack of jurisdiction” and, in doing so, prioritising “judicial economy” (see here). This inherent power has been invoked by the ICJ only twice before, in the Bolivia v. Chile and Marshall Islands v. United Kingdom cases.

Adding a layer of complexity to the case, Serbia intervened under Article 63 of the ICJ Statute. Formally backing the UAE’s position on jurisdictional reservations, Serbia’s move appeared to serve dual purposes. First, it aimed to protect Serbia from legal entanglement, given reports that weapons it had sold to the UAE were later used by the RSF. Second, it may have been a response to Sudan’s recognition of Kosovo. Yet Serbia’s intervention lacked substantive legal contribution, merely echoing settled jurisprudence permitting jurisdictional reservations, so long as they don’t defeat the Convention’s object and purpose.

This contribution reflects on the Sudan v. UAE decision, expanding on the legal and normative implications of the dissenting opinions (briefly skecthed here and here). It explores how the case has reignited debates surrounding sovereign consent and the object and purpose of treaties, particularly with regard to those that enshrine erga omnes partes obligations. It also considers the potential ‘spillover’ effects on third-party intervention under Article 63 of the ICJ Statute. 

(Re)Affirming the Status Quo: The Majority’s Endorsement of Reservations to Article IX of the Genocide Convention

The Court began by situating the dispute within the broader context of the armed conflict that erupted in April 2023 between the Sudanese Armed Forces and the RSF, along with affiliated militias. It noted Sudan’s serious allegations of atrocities against the Masalit ethnic group in West Darfur, including extrajudicial killings, ethnic cleansing, forced displacement, sexual violence, and ethnically motivated attacks (para. 16). While expressing deep concern over the humanitarian situation (para. 17), the Court recalled that its task at this preliminary stage was limited to assessing whether the criteria for indicating provisional measures – namely, prima facie jurisdiction, the plausibility of the rights asserted, a link between those rights and the requested measures, and the risk of irreparable harm – had been met (para. 18).

Sudan’s application relied on Article 36(1) of the ICJ Statute and Article IX of the Genocide Convention (para. 19). While both Sudan and the UAE are parties to the Convention, the Court noted that the UAE had entered a reservation to Article IX upon accession, explicitly excluding the Court’s jurisdiction (paras. 21-22). After reiterating that the erga omnes partes nature of obligations under the Genocide Convention does not override the fundamental requirement of State consent, thus echoing its position in DRC v. Rwanda (para. 28), the ICJ considered UAE’s reservation to be formulated in “clear terms” (para. 29), rejecting Sudan’s argument that the lack of reference to State responsibility rendered it ambiguous (para. 24). 

Therefore, the Court held that Article IX could not, even prima facie, establish its jurisdiction (paras. 33-34). As Sudan had invoked no alternative jurisdictional basis, the ICJ refrained to issue provisional measures while also removing the case from its General List. It underscored that, in the absence of consent, proceeding would undermine judicial economy and the integrity of its function (para. 35). Nonetheless, the Court emphasized that all States remain bound by their substantive obligations under the Genocide Convention and may still incur international responsibility for violations thereof (para. 36).

The Pandora’s Box of Treaty ‘Object and Purpose’: Analyzing the Dissenting Perspective on Genocide Convention Reservations

The dissenting and partly dissenting opinions attached to the Court’s order in the case of Sudan v. UAE offer a sharp critique of the majority’s formalistic approach to treaty reservations. They also expose deeper tensions within the ICJ’s case law on compromissory clauses and the limits of state consent. These opinions not only consider the procedural fate of a single case, but also the coherence and enforceability of the Genocide Convention itself as an instrument intended to protect the most fundamental values of the international community.

Judge Yusuf’s Dissenting Opinion is based on procedural grounds and challenges the ICJ’s authority to remove a case from its General List after a preliminary finding of a lack of prima facie jurisdiction at the provisional measures stage. Adopting a strict textualist reading of the Court’s Statute and Rules, he highlights their silence on the permissibility of such removal (para. 25). However, this position arguably set aside the Court’s long-established jurisprudence affirming its inherent power to manage its caseload, including the discretion to discontinue proceedings where jurisdiction is manifestly absent. The principle of “judicial economy” (see here), articulated in precedents such as the Nuclear Tests cases and the Legality of Use of Force (Serbia and Montenegro v. Belgium) proceeding, allows the ICJ to safeguard its institutional integrity and procedural efficiency by declining to proceed with cases lacking any plausible basis for adjudication. Furthermore, Judge Yusuf’s opinion conspicuously avoids addressing one of the core issue underlying the dispute: namely, the legal effect and broader implications of reservations to Article IX of the Genocide Convention.

In contrast, the Joint Partly Dissenting Opinion of Judges Bhandari, Charlesworth, Gómez Robledo, Cleveland, Tladi and Judge ad hoc Simma is arguably the most substantial judicial reconsideration of treaty reservations since the ICJ’s ruling in DRC v. Rwanda, notably drawing on the influential Joint Separate Opinion of Judges Higgins, Kooijmans, Elaraby, Owada and Simma in that earlier case. Central to both arguments is the proposition that Article IX of the Genocide Convention is not merely a peripheral procedural clause but, rather, the Convention’s only mechanism for inter-state enforcement. Lacking any other institutional supervisory body or alternative dispute resolution mechanism, Article IX thus functions as the sole means by which states can enforce the Convention’s erga omnes partes obligations. In this context, a reservation to Article IX not only restricts procedural access, but also undermines the structural coherence of the Convention and frustrates its object and purpose.

This critique aligns with evolving trends in general treaty law, as reflected in the International Law Commission’s (ILC) Guide to Practice on Reservations to Treaties. ILC Commentary on Guideline 3.1.5.7’s cautions against reservations to dispute settlement provisions where such clauses are indispensable to the effective implementation of a treaty. As posited by the ILC, “is the “raison d’être” of the treaty, its “fundamental core” that is to be preserved in order to avoid the “effectiveness” of the treaty as a whole to be undermined” (ILC’s Commentary, p. 442). In this regard, the Genocide Convention stands out as a particularly case. Indeed, unlike other ‘core’ UN human rights treaties (see Lixinski), such as the International Convention on the Elimination of All Forms of Racial Discrimination (CERD), the Genocide Convention lacks alternative dispute settlement or monitoring mechanisms. Furthermore, instruments such as the CERD establish compulsory interstate procedures (Articles 11-13) and contain provisions that explicitly prohibit reservations which would “undermine” their institutional architecture. In contrast, the Genocide Convention relies solely on Article IX for interstate enforcement. Therefore, entering a reservation to this provision is not just problematic; rather, it fundamentally undermines the structural coherence of the treaty regime. Although the ILC’s position is framed in general terms, it suggests a broader normative shift from a purely voluntarist approach grounded in state consent towards a more ‘axiological’ or value-oriented model. 

In this regard, the dissenting judges’ analysis reflects a broader conceptual insight: that the object and purpose of a treaty may go beyond the articulation of substantive obligations, thus encompassing also the procedural mechanisms essential for their enforcement. The Genocide Convention’s primary object and purpose – i.e. to prevent and punish the crime of genocide – cannot be separated from the legal framework that gives effect to these obligations. Procedural clauses such as Article IX therefore assume a functional and substantive dimension in treaties whose effectiveness hinges entirely on judicial enforcement. This has also emerged in the Joint Dissenting Opinion appended to the 1951 Advisory Opinion, which recalled the Preamble and UNGA Res. 96(1) and provocatively affirms as follows:

“What is the “object and purpose” of the Genocide Convention? To repress genocide? Of course, but is it more than that? Does it comprise any or all of the Convention’s enforcement articles? That is the heart of the matter. One need only look at them to realise the importance of this question” (p. 44).

This position is further reinforced by Judge Gómez Robledo’s dissenting opinion and Judge ad hoc Simma’s declaration, both of which call for the Court’s reservations doctrine to be examined again through a more dynamic, context-sensitive lens. These opinions emphasise the conflict between traditional voluntarist interpretations based on state consent and the need to protect fundamental human rights. They argue that the ICJ’s prevailing formalism fails to recognise the transformative nature of treaties such as the Genocide Convention, where strict adherence to consent-based reservations could render the treaty’s core protections ineffective. When states can opt out of the only available forum for adjudication, the Convention’s erga omnes partes obligations become illusory.

The Unintended ‘Spillover’ Effects: The Consequences of Treaty Reservations on the Admissibility of Future Third-Party Interventions

The minority’s evolving stance on reservations may signal a significant development in general treaty law, either as a departure that is aspirational, or as the foundation for rethinking long-established principles, most notably as concerns conventions establishing erga omnes partes obligations. 

Should this interpretive shift gain traction, particularly given the Court’s growing caseload of proceedings enforcing collective obligations under the Genocide Convention, its implications could extend well beyond traditional bilateral contexts. One under-explored consequence is the admissibility of third-party interventions under Article 63 of the ICJ Statute.

Historically, reservations to compromissory clauses and third-party interventions under Article 63 were approached relatively conservatively. Reservations were assessed either through a formalist, consent-based model under the Vienna Convention on the Law of Treaties (VCLT), or through the more purposive – albeit dated – framework of the ICJ’s 1951 Advisory Opinion. In both cases, reservations were tolerated as a compromise to facilitate broad participation at the expense of legal coherence. For this reason, they are often described as a ‘necessary evil’ (see ex multis Klabbers, p. 150): instrumental to universality, yet detrimental to enforceability.

Similarly, third-party intervention under Article 63 has always been interpreted as enabling a state to intervene when it is a party to a treaty and the treaty’s interpretation is in question. This approach remained unchanged until the Court’s 2023 order in the Ukraine v. Russia Genocide Case introduced a new procedural ‘barrier’ for intervening, which is not provided for in the ICJ Statute nor in the Rules of Court: States that have reserved certain rights are now excluded from interpretive proceedings that directly affect their treaty obligations. According to this ‘creative’ reasoning, the ICJ – thus upholding Russia’s observation – held that United States of America was barred from intervening despite being a party to the Genocide Convention due to its reservation to Article IX. By linking the right to intervene to acceptance of provisions whose interpretation or (re)construction is at stake in the proceeding, the Court has limited the capacity of such states to influence the interpretation of treaties, particularly those concerning collective enforcement and erga omnes partes obligations.

This reasoning raises questions about the future admissibility of interventions under Article 63 by States that have entered reservations. In Sudan v. UAE, Serbia’s intervention under Article 63 offered the Court a chance to address this issue, although it is not procedurally obligated to rule on the admissibility of an intervention if it dismisses the main case for lack of jurisdiction, thereby rendering the intervention moot (as occurred in the Request for an Examination case concerning the Solomon Islands, Marshall Islands, Micronesia, and Samoa interventions under Articles 62 and 63). In Sudan v. UAE, the situation would have turned to be equally complex: Serbia’s intervention might have been undermined by its own affirmation of its reservation to Article IX. Had the Court followed its recent reasoning in Ukraine v. Russia, even a disputed reservation might have barred intervention concerning reserved provisions. Paradoxically, such reservations both weaken jurisdictional accountability and exclude States from interpretive proceedings, silencing them despite their continued treaty obligations. Conversely, a more functional, teleological reading of Article 63 could empower non-reserving States to strengthen treaty coherence. However, if this emerging trend gains momentum, it risks deepening procedural asymmetries and raising concerns about fairness and representational imbalance by excluding reserving States from key interpretive processes.

Conclusion: Not the End, But the Beginning of a Shift

Although Sudan v. UAE did not overhaul the legal framework on treaty reservations, it may signal a significant interpretive change. The majority upheld the validity of reservations to Article IX of the Genocide Convention, but the dissent raised strong normative concerns about their compatibility with the Convention’s object and purpose. 

The decision’s implications extend far beyond this case. With several genocide-related cases still pending – including, but not limited to, those brought against Myanmar and Israel –, the validity of reservations is likely to face increasing scrutiny. Whether the dissent’s purposive, teleological reasoning will gain wider acceptance remains uncertain. What is clear is that the debate over reservations, consent, and collective enforcement has moved from the margins to the center of international legal discourse.

The procedural spillover effects on third-party interventions highlight broader stakes. Excluding States with jurisdictional reservations from interpretive proceedings risks evolving treaty regimes without their input, raising concerns about legitimacy, inclusiveness, and the long-term integrity of these regimes. Conversely, non-reserving States committed to treaty coherence may continue to influence emerging legal norms through strategic interventions under Article 63.

Ultimately, the dissenting judges in Sudan v. UAE illustrate how procedural issues can shape international law’s substance and trajectory. While treaty reservations are unlikely to disappear soon, their unquestioned acceptance appears increasingly fragile, especially when collective interests and fundamental values are involved.