1. Introduction

Under customary international law and maybe also the Genocide Convention,[1] individuals are prohibited from committing the crime of genocide, as defined in Article II of the Convention, as well as ancillary crimes, like complicity in genocide, enumerated along with it in Article III. Breach of any of these prohibitions gives rise to the criminal responsibility under international law of the individual in question. As defined in Article II of the Convention, the crime of genocide involves the commission of any of five specified acts[2] ‘with intent to destroy, in whole or in part, a national, ethnical, racial or religious group, as such’. The specific intent, or ‘dolus specialis’, to destroy in whole or in part any of the protected groups is the essence of the crime of genocide.
For their part, under customary international law and – as per the International Court of Justice (ICJ) in Application of the Genocide Convention (Bosnia and Herzegovina v Serbia and Montenegro) – also the Genocide Convention, States are prohibited from committing genocide,[3] as distinguished terminologically by the Court from the crime of genocide.[4] In the further view of the Court, they are equally prohibited under the Convention from committing the ‘acts’[5] ancillary to genocide also enumerated in Article III of the Convention. Breach of any these prohibitions gives rise to the responsibility under international law of the State in question, a responsibility which, like all State responsibility, is delictual, not criminal, in character.[6] Genocide for the purposes of State responsibility is nonetheless defined, according to the ICJ, as per the definition of the crime of genocide in Article II of the Convention. To be responsible for breach of the prohibition on genocide, a State must perform one of the five acts specified Article II ‘with intent to destroy, in whole or in part, a national, ethnical, racial or religious group, as such’. In short, it must act with a certain state of mind.
Unlike an individual, a State is a juridical, not natural, person. It is a legal construct, not a living being. As such, it has no mind in any real sense. Moreover, a State is a collective entity. It is the notionally indivisible legal personification of the myriad other entities, including the governmental organs, and the individuals that as a social and constitutional reality comprise it. If one is to ascribe to a state a mind for the purpose of the State’s commission of genocide, the problem is where to find it.
Locating the mind of a State is seldom necessary. As a matter of the secondary rules of State responsibility, breach of obligation is independent of fault and, even more so, of intention.[7] The standard of conduct required of a State depends instead on the primary obligation,[8] and it is very rare that a primary obligation can be breached by a State only with intent.
The question in the end is whether genocidal intent is to be ascribed to a State on the basis of the customary international rules on attribution of conduct to a State or on some other basis. The answer given by the ICJ in Bosnia v Serbia and again in Application of the Genocide Convention (Croatia v Serbia) is that such intent is to be ascribed on the basis of the rules on attribution of conduct to a State. But whether this answer is convincing is open to discussion. The matter is not without importance, given that the question may be aired again in, variously, the proceedings and potential proceedings on the merits in the four cases now before the Court alleging breach of the Genocide Convention’s prohibition on a State’s commission of genocide.[9]
The present article reflects on the approach to locating a State’s genocidal intent adopted by the ICJ in Bosnia v Serbia and Croatia v Serbia, as well as on an alternative to it.[10] The focus is on State responsibility for the commission of genocide, rather than for complicity in genocide or for any of the other ‘acts’ ancillary to genocide enumerated in Article III of the Genocide Convention.

2. The ICJ’s approach

In Bosnia v Serbia, the ICJ ‘affirms’:

By ‘commits any of the acts proscribed by Article III’ what the Court evidently means is ‘commits, with the requisite specific intent, any of the acts proscribed by Article III’,[12] which is to say commits the crime, not just the actus reus, of genocide, complicity in genocide, and so on. The Court elaborates – in a dictum reiterated in Croatia v Serbia[13] – that, ‘if a State is to be responsible because it has breached its obligation not to commit genocide, it must be shown that genocide as defined in the Convention’, by which the Court must be taken to intend the crime of genocide as defined in the Convention, ‘has been committed’.[14]
Subsequently, in Croatia v Serbia, the ICJ speaks – again in the context of State responsibility for breach of the obligation not to commit genocide – of ‘acts [of genocide], attributable to the State, committed’, necessarily with specific intent, ‘by a person or a group of persons whose individual criminal responsibility has already been established’ by a competent criminal court or whose commission of acts of genocide, with the necessary specific intent, the Court itself determines.[15] ‘In either of these situations’, the Court explains, it ‘applies the rules of general international law on the responsibility of States for internationally wrongful acts’,[16] the reference being to the rules on attribution of conduct to a State. In other words, ‘[i]f it is established that genocide’, evidently meaning the crime of genocide as defined in Article II of the Genocide Convention,[17] ‘has been committed, the Court will then seek to determine the responsibility of the State, on the basis of the rules of general international law governing the responsibility of States for internationally wrongful acts’,[18] specifically the rules on attribution.
In both cases, the ICJ examines the evidence before it to determine whether acts of genocide were committed with the requisite specific intent by persons or groups of persons. In the event, only in the first case and only then in relation to the massacres at Srebrenica by Bosnian Serb paramilitaries – amounting to one of the five acts specified in Article II of the Genocide Convention – does the Court find established the specific intent necessary to enable it to conclude that the massacres ‘constituted the crime of genocide within the meaning of the Convention’,[19] as referred to in Article III(a). The Court goes on to ask ‘whether the massacres committed at Srebrenica during the period in question, which constitute the crime of genocide within the meaning of Articles II and III, paragraph (a), of the Convention, are attributable, in whole or in part, to the Respondent [State]’,[20] ultimately concluding in the negative.
According, then, to the ICJ’s reasoning, the specific intent on the part of a State required for that State’s responsibility for breach of its obligation not to commit genocide is to be found in the mind of any person individually responsible for the crime of genocide whose conduct is attributable to the State in that instance. Implicitly characterizing as the ‘conduct’ for the purpose of the rules on the attribution of conduct to a State not just the person’s actus reus but also the mens rea accompanying it, the Court attributes to the State not only the person’s bare acts but also his or her mental State. A State is deemed to act ‘with intent to destroy, in whole or in part, a national, ethnical, racial or religious group, as such’ if any person whose acts are attributable to the State in that instance acts with that intent. The mind of any such person is deemed the mind of the State.

3. A convincing approach?

The ICJ’s approach to locating the genocidal intent of a State is by no means implausible. There is no inherent reason why the ‘conduct’ of a person, within the meaning of the rules on attribution of conduct to a State set out in chapter II of Part One of the International Law Commission (ILC)’s Articles on Responsibility of States for Internationally Wrongful Acts, cannot refer to both the person’s act and its accompanying mental state. Conceptually, moreover, if a State is taken to act through the limbs, as it were, of any person whose conduct is attributable to it in that instance under the law of State responsibility, it is not immediately apparent why the State should not be taken in the same instance to intend through the mind of that person.
At the same time, nothing compels the Court’s approach. On the contrary, in the terminological hermeneutics of the rules on attribution of conduct to a State in the Articles on Responsibility of States, the fact that ‘conduct’ of a person or entity shall be considered an ‘act’ of the State could be understood to imply that the notion of ‘conduct’ is limited to the person’s bare act, which is to say the act shorn of its accompanying mental state. In the context of State responsibility for breach of the obligation not to commit genocide, this would encompass solely the actus reus, and not also the mens rea, of the crime of genocide, complicity in genocide or the like, as committed by the person. Indeed, it would seem that the reason why the ILC, in the rules on attribution in the Articles, did not provide simply that an ‘act’ of a person shall be considered be an ‘act’ of the State is that a person’s omissions, and not just acts in the narrow sense, are equally capable of being considered an ‘act’ of the State. The term ‘conduct’ in chapter II of Part One of the Articles is shorthand for ‘conduct consisting of an act or omission’, as formulated more fully in Article 2.[21] The argument for excluding a person’s mental state from the scope of the term ‘conduct’ in chapter II of Part One is lent further weight by the fact that conduct not only of natural persons but also of entities, which is to say municipal legal persons, is capable of being considered an act of the State under the rules on attribution.[22] While natural persons are possessed of real mental states, municipal legal persons – just like the international legal person of the State – are not. These points taken together might be thought to suggest that the purpose of the customary international rules on attribution of conduct is the attribution to the international legal person of the State of merely the bare acts and omissions of natural and municipal legal persons and not also, in the case of natural persons, the mental state accompanying any such act or omission. The implication could be said to be the stronger for the fact that neither intent nor legally-relevant knowledge on the part of a State is a general condition of State responsibility.[23]
Conceptually, furthermore, there is arguably a meaningful difference between a State’s arms, legs, hands, feet, fingers and toes, so to speak, and its mind. It seems more intuitive to conceive of the mind, and with it the intent, of the anthropomorphic construct of the juridical person of the State as residing in some central decision-making organ or cluster of such organs, rather than in each and every organ of State, let alone in persons or entities other than State organs whose conduct is nonetheless attributable under the law of State responsibility to the State in that instance. An intuition of the sort is arguably reflected in the customary international rules relating to the representation of the State in its international relations.[24] It is generally only the government of the State and persons authorized by it to this end, not every organ of State as understood for the purposes of State responsibility, that is taken to speak on behalf of the State in its international relations.[25] Yet an analogous intuition is not reflected in the ICJ’s approach to the genocidal intent of a State. By way of illustration, take the following hypothetical scenarios.
First, the government of a State engaged in an armed conflict issues formal instructions to its regular armed forces, instructions backed in practice by unambiguous orders from the military high command, not to kill members of a particular religious group apart from those qualifying as combatants or as civilians taking direct part in hostilities. Contrary, however, to these instructions and orders, soldiers belonging to one regular military unit of the State, each acting in that capacity and with the intent to destroy the religious group as such at least in substantial part, intentionally kill 5000 members of that group. In accordance with the customary international rules on attribution of conduct to a State codified in Articles 4 and 7 of the Articles on Responsibility of States, the acts of the soldiers are considered acts of the State, despite being contrary to instructions, and there are good policy reasons for this.[26] But intuitively we may well think that the soldiers’ acts do not reflect the intent of the State, as arguably better looked for in the instructions issued by the government, as backed by the orders from the military high command. Under the ICJ’s approach, nonetheless, the State is deemed to have acted with the intent to destroy the religious group as such in part and bears responsibility for breach of its obligation not to commit genocide.
Secondly, the government of a State directs a non-State paramilitary group to execute 5000 civilians, in retaliation for their electoral support for an opposition political party. All 5000 civilians happen to be members of the same ethnic group, but the government has no intent to destroy the group as such,[27] either in whole or in part. The paramilitaries, however, following the government’s directions to the letter, assume that the executions form part of a plan on the part of the government to destroy the group as such at least in part, and they intend their acts to do the same. Under the customary international rule on attribution of conduct to a State codified in Article 8 of the Articles on Responsibility of States, the acts of the paramilitaries are considered acts of the State. But our intuition may well be that it is not right to attribute to the State the paramilitaries’ genocidal intent when the government of the State, which directed their acts, lacked any such intent. Nonetheless, on the ICJ’s approach, the intent of the paramilitaries is deemed the intent of the State, which is consequently responsible for breach of its obligation not to commit genocide.
In neither of these two scenarios or in any analogous to them is the outcome obviated by the ICJ’s approach to establishing on the basis of circumstantial evidence the genocidal intent of a person who commits an act set forth in Article II of the Genocide Convention. In Bosnia v Serbia, the Court accepts in principle that the intent of a person to destroy a protected group as such in whole or in part can be established by recourse to the existence, ‘convincingly demonstrated’, of a ‘general plan to that end’,[28] which is to say a plan embodying ‘the specific intent (dolus specialis) of those directing the course of events’.[29] A collateral consequence is that, in the event that the conduct, which includes the mens rea established by recourse to the plan, of the person is attributable to the State, the State may end up responsible for genocide on the basis of a specific intent deriving in substance from ‘the intent of higher authority … at the level of the [g]overnment’ of that State[30] – what is later referred to succinctly, if question-beggingly, in Croatia v Serbia as ‘a State plan expressing the intent to commit genocide’.[31] But this does nothing to alter the fact that, in the converse situation of the absence of intent on the part of the government of the State to destroy in whole or in part a protected group as such, the State can be responsible for genocide, in accordance with the ICJ’s reasoning on locating the genocidal intent of a State, on the basis of a specific intent on the part of the person who commits the impugned act which owes nothing to that government.
In sum, then, according to the ICJ’s approach, the specific intent on the part of a State required for that State’s responsibility for breach of its obligation not to commit genocide is furnished by the specific intent of any person individually responsible for the crime of genocide whose conduct is attributable to the State in that instance. This approach reflects an application of the customary international rules on attribution of conduct not only to the person’s bare act but also to the mental state with which it is committed. Although certainly one way of looking at it,[32] the Court’s, however, is not the only way. Nor does it reflect what is arguably the intuition that the intent of a State as a juridical person corresponds to the intent of the competent, central decision-making organ or organs of that State.[33] An approach to locating the genocidal intent of a State that speaks to this intuition – while accepting that the bare acts specified in Article II(a) to (e) of the Genocide Convention could each be attributed to the State on any customary basis of attribution – would attribute to the State the specific intent only of that organ or those organs.
On this second, alternative approach, where the competent, central decision-making organ or organs of State do not possess the intent to destroy in whole or in part – through the acts of organs of State or of persons acting on the instructions or under the direction or control of organs of State – a protected group as such, the State would not breach the prohibition on the commission by a State of genocide. This would be so, and this is the gist of the difference, even where persons whose conduct is attributable to the State commit one or more of the acts in Article II(a) to (e) of the Genocide Convention with the requisite specific intent, thereby committing the crime of genocide. In contrast, where the competent, central decision-making organ or organs of State do possess this specific intent, which is put into action by means of acts specified in Article II(a) to (e) of the Genocide Convention carried out by organs of State or by persons acting on the instructions or under the direction or control of organs of State, the State would be responsible for committing genocide. The latter would be so regardless of whether the specific intent on the part of the competent, central decision-making organ or organs of State were shared by the persons who carry out the acts in Article II(a) to (e) through which that organ or organs intend to destroy in whole or in part the group as such. In other words, the State would breach the prohibition on the commission by a State of genocide even though the persons who carry out the relevant acts do not commit the crime of genocide. Precisely which organ or organs constitute the competent, central decision-making organ or organs of State in any instance will depend on the State and the acts in question.
The alternative approach to locating the genocidal intent of a State demands in effect, although not as a formal material element of the definition of genocide, that the relevant acts attributable to the State be performed pursuant to some sort of policy – formal or informal, explicit or implicit – reflecting the intent of the competent, central decision-making organ or organs of State ‘to destroy, in whole or in part, a national, ethnical, racial or religious group, as such’.[34] (This is not the case, in contrast, when the question is whether a person is individually responsible for the crime of genocide,[35] even if the existence of a policy may be evidentially useful in establishing the person’s specific intent.[36]) Where not otherwise established, the existence of a policy of some sort embodying the specific intent of the competent, central decision-making organ or organs ought logically to be provable by inference from a pattern of conduct where, but only where, ‘this is the only inference that could reasonably be drawn from the acts in question’.[37]

4. Conclusion

Where to locate the intent of a juridical person is a problem with which those bodies of national law that recognize corporate criminal responsibility similarly wrestle. In that context too the more or most convincing answer continues to be debated.[38] At the same time, it is not necessarily the case that solutions suited to the criminal liability under national law of, most commonly, a company will be suited to the delictual responsibility under international law of a State.
At root the difficulty in the context of State responsibility for genocide lies perhaps less in the ICJ’s approach to locating a State’s genocidal intent than in the more fundamental proposition advanced by the Court in Bosnia v Serbia, as affirmed, that the definition of genocide for the purpose of the prohibition on its commission by a State is identical to the definition of the crime of genocide in Article II of the Genocide Convention. From the perspective of the Court in its judgment, the proposition is necessitated by its doubtful[39] ruling that the Convention itself prohibits States from committing genocide. The upshot of the asserted commonality of definition is that, in order to breach the prohibition on the commission by a State of genocide, a State – despite being a juridical, not natural, person – must be said to possess intent, namely the ‘intent to destroy, in whole or in part, a national, ethnical, racial or religious group, as such’. This positing of a mental element as a condition of State responsibility sets the prohibition on the commission by a State of genocide apart not only from the overwhelming majority of primary obligations on States but also, more particularly, from those obligations on States with counterparts in crimes under international law. While intent is generally a condition under international law for individual criminal responsibility for a war crime, a State can violate the laws of armed conflict unintentionally. Similarly, although individuals cannot be responsible under international law for the crime of aggression unless they commit the relevant act with intent, no intent is required for a State to violate the prohibition on its commission of aggression. Nor is there any necessary reason why the definition of genocide for the purpose of the undoubted prohibition under customary international law, rather than the Genocide Convention, of the commission by a State of genocide need be identical to the definition of the crime of genocide, with its requirement of specific intent, in Article II of the Convention.[40] Indeed, in the pre-Convention General Assembly resolution 180 (II) of 21 November 1947, in which the Assembly declares that ‘genocide is an international crime entailing … international responsibility on the part of … States’, no definition of genocide is given and it is not clear whether specific intent is envisaged.
That said, even in the context of the prohibition on the commission of genocide by a State, it is difficult to conceive of a definition of genocide shorn of the intent ‘to destroy, in whole or in part, a national, ethnical, racial or religious group, as such’. This specific intent is seemingly inescapably the essence of genocide as a concept. If one accepts that – whether under customary international law, the Genocide Convention or both – States are prohibited from committing genocide, one appears bound to accept that they violate the prohibition only if they act with such intent. The question then inevitably becomes where to locate this intent.
According to the ICJ, the specific intent on the part of a State required for that State’s responsibility for breach of its obligation not to commit genocide is to be found in the mind of any person individually responsible for the crime of genocide whose conduct is attributable to the State in that instance. On an alternative approach, the State’s requisite genocidal intent would be sought instead solely on the part of the competent, central decision-making organ or organs of that State. While it is unlikely that the Court will change its approach, only future judgments will tell.

* Full Professor of International Law, Bocconi University. This article is based on presentations given respectively at the University of Bologna Alma Mater Studiorum on 5 April 2024 and the University of Padua on 27 May 2024.
[1] Convention on the Prevention and Punishment of the Crime of Genocide (adopted 9 December 1948, entered into force 12 January 1951) 78 UNTS 277 (hereafter ‘Genocide Convention’). For whether the Genocide Convention itself prohibits the crime of genocide or instead simply ‘confirms’, as per art I, that genocide is a crime under customary international law, see R O’Keefe, International Criminal Law (OUP 2015) 234-5.
[2] These acts are: ‘(a) Killing members of the group; (b) Causing serious bodily or mental harm to members of the group; (c) Deliberately inflicting on the group conditions of life calculated to bring about its physical destruction in whole or in part; (d) Imposing measures intended to prevent births within the group; (e) Forcibly transferring children of the group to another group.’ Although the requirement is explicit only in para (c) of art II of the Genocide Convention, these acts must be committed intentionally in order to ground individual responsibility for the crime of genocide. See eg Application of the Convention on the Prevention and Punishment of the Crime of Genocide (Bosnia and Herzegovina v Serbia and Montenegro) Judgment ICJ Rep 2007, 43 (hereafter ‘Bosnia v Serbia’) 121 para 186.
[3] See Bosnia v Serbia (n 2) 113-14 paras 166-167, 118-19 para 179.
[4] See eg ibid. See also Application of the Convention on the Prevention and Punishment of the Crime of Genocide (Croatia v Serbia) Judgment ICJ Rep 2015, 3 (hereafter ‘Croatia v Serbia’) sep op Skotnikov 199-200 para 12.
[5] The consistent reference of the Court in the context of State responsibility, taking its cue from art IX of the Genocide Convention, is to the other ‘acts’, not other ‘crimes’, in art III.
[6] See especially Bosnia v Serbia (n 2) 115 para 170.
[7] Commentary to draft articles on responsibility of States for internationally wrongful acts (2001) II/2 YB Intl L Commission 31 (hereafter ‘Commentary to articles’) 36, commentary to art 2 para 10.
[8] ibid.
[9] See ICJ, Application of the Convention on the Prevention and Punishment of the Crime of Genocide (The Gambia v Myanmar) General List No 178; Allegations of Genocide under the Convention on the Prevention and Punishment of the Crime of Genocide (Ukraine v Russian Federation) General List No 182; Application of the Convention on the Prevention and Punishment of the Crime of Genocide in the Gaza Strip (South Africa v Israel) General List No 192; Alleged Breaches of Certain International Obligations in respect of the Occupied Palestinian Territory (Nicaragua v Germany) General List No 193.
[10] For other consideration of these or closely related issues, see A Nollkaemper, ‘Concurrence between Individual Responsibility and State Responsibility in International Law’ (2003) 52 ICLQ 615, 633-5; M Milanović, ‘State Responsibility for Genocide’ (2006) 17 Eur J Intl L 553, 567-9, 574, 601-2; P Gaeta, ‘On What Conditions Can a State Be Held Responsible for Genocide?’ (2007) 18 Eur J Intl L 631, 637, 641-4; AB Loewenstein, SA Kostas, ‘Divergent Approaches to Determining Responsibility for Genocide. The Darfur Commission of Inquiry and the ICJ’s Judgment in the Genocide Case’ (2007) 5 J Intl Criminal Justice 839; B Bonafé, The Relationship between State and Individual Responsibility for International Crimes (Brill 2009) 119-45; A Seibert-Fohr, ‘State Responsibility for Genocide under the Genocide Convention’ in P Gaeta (ed), The UN Genocide Convention. A Commentary (OUP 2009) 349, 367-8; WA Schabas, Genocide in International Law. The Crime of Crimes (2nd edn, CUP 2009) 518; R Pisillo Mazzeschi, E Carli, ‘Proof of Specific Intent in the Crime of Genocide. The Case of South Africa v. Israel Before the International Court of Justice’ (2024) 22 J Intl Criminal Justice 429, 440-1, 443.
[11] Bosnia v Serbia (n 2) 119 para 179. See also ibid 113-14 paras 166-167.
[12] This is evident from the fact, inter alia, that the first ‘act’ listed in art III of the Genocide Convention is genocide itself, and genocide is defined in art II as requiring specific intent. Consider also, inter alia, the identity drawn between the finding, ibid 166 para 297, that the massacres at Srebrenica constituted ‘acts of genocide’, as well as the reference, ibid 199 para 379, 201 para 384, to the same, and the reference, ibid and ibid 214 para 413, to the same acts as ‘the crime of genocide’.
[13] See Croatia v Serbia (n 4) 61 para 128.
[14] Bosnia v Serbia (n 2) 119 para 180.
[15] Croatia v Serbia (n 4) 61 para 128.
[16] ibid.
[17] This is clear not only from the preceding context but also from the Court’s immediately subsequent highlighting, ibid 61-2 para 130, of the definition of the crime of genocide in art II of the Genocide Convention, in which, furthermore, the Court states that, ‘[a]ccording to that Article, genocide contains two constituent elements: the physical element, namely the act perpetrated or actus reus, and the mental element, or mens rea.’
[18] ibid 61 para 129.
[19] Bosnia v Serbia (n 2) 199 para 379. See also ibid 214 para 413.
[20] ibid 201 para 384.
[21] Articles on Responsibility of States for Internationally Wrongful Acts, UNGA Res 56/83 (12 December 2001) Annex (hereafter ‘Articles on Responsibility of States’) art 2 (‘There is an internationally wrongful act of a State when conduct consisting of an act or omission …’). See also Commentary to articles (n 7) 35 commentary to art 2 para 4.
[22] See Articles on Responsibility of States (n 21) arts 4(2), 5, 7. See also Commentary to articles (n 7) 42 commentary to art 4 para 12.
[23] See Commentary to articles (n 7) 36 commentary to art 2 para 10.
[24] Consider also in this light L Condorelli, Claus Kreß, ‘The Rules of Attribution: General Considerations’ in J Crawford et al (eds), The Law of International Responsibility (OUP 2010) 221, 223: ‘[C]onduct giving rise to international responsibility entails legal consequences which in no way correspond to the aims of the State to which the conduct is attributable, while conduct which manifests the consent of a State to be bound on the international plane produces, at least in principle, precisely this effect.’
[25] See generally N Pavlopoulos, The Identity of Governments in International Law (OUP 2024) 24-30 and references therein.
[26] See eg Commentary to articles (n 7) 45, commentary to art 7, paras 2, 3.
[27] The qualification ‘as such’ connotes that the intent must be to destroy the group in its quality as that group. It is insufficient to intend to destroy members of the group only in their capacity as specific or random individuals. See eg Prosecutor v Niyitegeka, ICTR-96-14-A, Appeals Chamber Judgment (9 July 2004) para 53.
[28] Bosnia v Serbia (n 2) 196-7 para 373. See also Croatia v Serbia (n 4) 66 para 145.
[29] Bosnia v Serbia (n 2) 195 para 371. It is worth underlining that the apparent object of ascertaining the existence of a general plan embodying the specific intent of ‘those directing the course of events’ is to establish that the persons who committed relevant acts pursuant to this plan did so themselves with specific intent, with the consequence that they committed the crime of genocide. (The question is then whether these persons’ ‘conduct’, viz acts accompanied by specific intent, is attributable to the State so as to render the State responsible for genocide.) The object of ascertaining the existence of a plan appears not to be to ascertain directly the specific intent of the State itself. This is sufficiently evident from the overall context, especially as framed ibid 142 para 242, as well as more specifically from the introduction, ibid 194 para 370, to the pertinent passage, from the reference, ibid 199 para 379, to ‘the crime of genocide’, and from the fact (see n 30) that the relevant plan could be one on the part of the leadership of a non-State group or entity. The Court nonetheless muddies the waters by noting, ibid 195 para 371, that the objectives defined in the plan of the leadership of the non-State Republika Srpska were shared by the head of State of the respondent State and by referring, ibid 198 para 376, to specific intent ‘on the part of the Respondent’, as opposed to ‘on the part of the government of the Respondent’ or the like. The apparent object of ascertaining the existence of a general plan is similarly evident in Croatia v Serbia (n 4) 61-2 paras 128-130, 118 para 402, as is the blurring of this object, ibid 65-6 paras 143, 145, and 148-50 paras 504-507.
[30] Bosnia v Serbia (n 2) 195 para 371. Note that a general plan capable of furnishing circumstantial evidence of the genocidal intent of any person who commits an act specified in art II of the Genocide Convention need not be one on the part of the government of the State but can instead be one on the part of, for example, the leadership of any non-State paramilitary group to which the person may belong or of any non-State political entity on whose behalf that group may fight. See eg ibid. In the latter two situations, in the event that the conduct of the person is attributable to the State, the State will end up responsible for genocide on the basis of a specific intent deriving in substance from the intent of the leadership of the non-State paramilitary group or political entity, rather than of the government of the State.
[31] Croatia v Serbia (n 4) 66 para 145.
[32] For support for the approach taken, in the event, by the Court, see eg Milanović (n 10) 567-9, 574, 601-2.
[33] For support for an approach along these lines, see Pisillo Mazzeschi, Carli (n 10) 440-1, 443.
[34] See similarly Gaeta (n 10) 643; Schabas (n 10) 518.
[35] See, among others, Prosecutor v Jelisić, IT-95-10-A, Appeals Chamber Judgment (5 July 2001) para 48; Prosecutor v Simba, ICTR-01-76-A, Appeals Chamber Judgment (27 November 2007) para 260.
[36] Jelisić (n 35) para 48.
[37] Croatia v Serbia (n 4) 67 para 148, 122 para 417, adding a gloss to Bosnia v Serbia (n 2) 197 para 373. Note, however, that in these cases the inference sought was of the specific intent of the persons who committed the relevant acts, which, if proved, would have established their commission of the crime of genocide. The question was then whether this ‘conduct’ was attributable to the State in question. The argument advanced in the text relates to proof of the existence of a policy reflecting – and this is the ultimate question – the specific intent of the State, in the form of the specific intent of its competent, central decision-making organ or organs of State. For the latter, see eg Croatia v Serbia (n 4) separate opinion Gaja 396 para 3.
[38] See eg, among many others, BP Edwards, ‘What Is a Corporate Mind? Mental State Attribution’ in MJ Durkee (ed), States, Firms, and Their Legal Fictions. Attributing Identity and Responsibility to Artificial Entities (CUP 2024) 197; PS Abril, A Morales Olazábal, ‘The Locus of Corporate Scienter’ (2006) Columbia Business L Rev 81; ME Diamantis, ‘Corporate Criminal Minds’ (2016) 91 Notre Dame L Rev 2049.
[39] See eg O’Keefe (n 1) 80; Gaeta (n 10) 632-40.
[40] See also Gaeta (n 10) 637, 641-4.

1. Introduction

Genocide cases make headlines. The provisional measures hearings in South Africa v Israel, held on 11 and 12 January 2024, were watched by an audience in the tens of thousands. The hashtag #InternationalCourtofJustice was even trending on social media platforms such as Instagram, TikTok, and Twitter.[1] Shortly thereafter, on 26 January,^[2] the International Court of Justice (‘ICJ’ or ‘the Court’) held another public sitting for the reading of its first provisional measures Order,[3] in which it declared that ‘the right of the Palestinians in Gaza to be protected from acts of genocide’ was ‘plausible’[4] and that ‘there is a real and imminent risk that irreparable prejudice will be caused to the rights found by the Court to be plausible’.[5] The Court ordered Israel to, inter alia, take all reasonable measures to prevent the commission of acts falling within art II of the Genocide Convention,[6] and to ensure that its military did not engage in any such acts.^[7] It also directed Israel to take ‘immediate and effective measures to enable the provision of urgently needed basic services and humanitarian assistance’ in Gaza.^[8]
In the weeks that followed, as conditions in Gaza worsened and as Israel appeared to expand its military offensive, South Africa requested the Court to amend its provisional measures orders three times. However, there was only one more public hearing conducted in the Great Hall of Justice.
First, on 12 February, South Africa requested the Court to exercise its power under art 75(1) of the Rules of Court to indicate provisional measures proprio motu against Israel.^[9] The Court declined to do so, while nevertheless reiterating the necessity of ‘immediate and effective implementation’ of the provisional measures previously indicated.^[10] Then on 6 March, South Africa sought further provisional measures and/or the modification of the provisional measures order of 26 January, on the basis that ‘widespread starvation’ in Gaza constituted a change in circumstances necessitating modification of the January Order.^[11] On 28 March– without holding a public hearing – the Court acceded to the request and indicated measures particularly directed at ensuring the ‘unhindered provision at scale of urgently needed basic services’.^[12] Third and finally, South Africa submitted another request on 10 May. This time the Court sat in oral proceedings (on 16 and 17 May) but they were poorly attended. Israel’s external legal counsel were not available within the specified timeframe leaving only its Co-Agents appearing in the Great Hall of Justice.[13] The Court, in an Order dated 24 May,^[14] directed Israel to, inter alia:

This order was passed with 13 votes in favour and 2 against, an apparently strong majority. News outlets shortly afterwards declared that the Court had ordered Israel to ‘immediately halt its military assault on the southern Gaza city of Rafah’.[16] However, it was (and remains) controversial whether this was an order to halt all military action in Rafah or only that which could risk creating conditions of life that might lead to physical destruction of the Palestinian people.[17]
But rather than take up questions relating to the substance of the Court’s Orders, the present article is addressed to an aspect of the Court’s procedure in this case; specifically, the oral proceedings. The Court indicated provisional measures in South Africa v Israel without holding an oral hearing. This has hitherto been an exceedingly rare event; the Court has issued provisional measures without oral proceedings on only two prior occasions.[18] More broadly, decisions on the papers seem to be something of a growing trend and have featured in all three of the Court’s recent Genocide Convention cases.[19]
While it might seem trite to focus on issues of procedure in the face of a ‘catastrophic humanitarian situation’[20] in Gaza which has only grown vastly more significant over time,[21] it will be beneficial to the Court’s future practice to identify the value of oral hearings and principles that might guide their non-use.
The article proceeds in 5 parts. Part 2 commences with a summary of the procedural regime governing oral proceedings; observing that in respect of provisional measures and other interlocutory phases, an oral hearing is not compulsory. Querying why oral proceedings are nevertheless considered an essential part of the court process, Part 3 explains that their unique importance lies in the expressive manifestation of procedural fairness values. There are three core value elements of procedural fairness – equality, dignity, and impartiality – and oral proceedings perform and embody these values. Part 4 observes that the provisional measures proceedings in South Africa v Israel gave rise to several procedural fairness concerns, which could negatively impact on the Court’s perceived legitimacy and authority. Part 5 concludes.

2. Oral proceedings for provisional measures: Implicit rules and inconsistent practices

2.1. The Statute and Rules

The Court’s Statute and Rules provide only limited guidance in respect of oral proceedings generally. Article 43(1) of the Court’s Statute requires that proceedings ‘consist of two parts: written and oral’.^[22] Article 43(5) indicates that the oral proceedings may include the hearing by the Court of ‘witnesses, experts, agents, counsel, and advocates’.^[23] Once the oral proceedings are concluded, the case is closed for the time being,^[24] and the Court retires to deliberate on its judgment or order.
The oral proceedings are governed principally by the Rules in Articles 54 to 72, which guide the Court in respect of matters such as the time and place of the hearings,^[25] the languages used in Court,^[26] the order in which the parties will be heard, and the examination of any witnesses or experts.^[27] The use of oral witness evidence, factual or expert, remains very unusual before the Court.^[28]
Article 59(1) of the Rules makes clear that the decision to conduct hearings in private is a broad power that rests with either the Court or the parties:

However, oral hearings held in private are one thing; sometimes publicity might disturb the proceedings,^[29] or certain evidence might need to be redacted to protect victims.^[30] But this is a quite separate question from whether to conduct oral proceedings at all, and this issue is not explicitly governed by either the Statute or the Rules.
Turning to the Rules governing the conduct of provisional measures, some additional insights regarding oral proceedings may be gleaned. Article 74(1) requires that requests for the indication of provisional measures have priority over all other cases – this includes in respect of oral proceedings. However, provisional measures hearings are considered to be so urgent that parties are only afforded ‘an opportunity of being represented’,[31] and the Court ‘shall receive and take into account any observations that may be presented to it before the closure of the oral proceedings’.[32] This implies that provisional measures may be indicated following a purely written procedure, a conclusion further supported by the terms of Article 76 which provide that:

To the extent that the Court is empowered to revoke or modify any decision concerning provisional measures proprio motu and need only ‘afford the parties an opportunity of presenting their observations’, it would seem that the Court has a wide discretion to decide whether or not to hold oral proceedings.
The Court’s practice also suggests that it is only in respect of the merits phase before the full Court that an oral hearing is mandatory.^[33] For interlocutory phases, the oral proceedings are optional,^[34] although usually held.^[35] The Court also held hearings by videoconference during the height of the COVID-19 pandemic,^[36] although this now seems to be in abeyance.
However, the Court has only very rarely issued an order or judgment without holding oral hearings. These cases are addressed next.

2.2. Judgments and Orders issued without oral proceedings

The first series of cases can be treated collectively; several Advisory Opinions in which ‘appeals’ from international administrative tribunals were in issue. These are discussed further in Part 3. But outside of this context, there are very few interlocutory decisions in which the Court has issued a judgment or orders without holding a hearing – two involving requests for provisional measures, three concerning intervention, and two addressed to interpretation.

a) Provisional Measures

The first is the well-known case of LaGrand, in which Germany instituted proceedings against the United States in the evening of 2 March 1999 in a dispute concerning alleged violations of the Vienna Convention on Consular Relations.[37] Karl and Walter LaGrand were German nationals convicted of murder and sentenced to the death penalty; Karl had been executed on 24 February and Walter was due to be executed on 3 March 1999. At 9am that morning (The Hague time) Vice-President Weeramantry received the representatives of the parties in chambers, at which point the German Agent asked the Court to indicate, without holding any hearing, provisional measures proprio motu pursuant to Article 75 of the Rules.
This was the first (and so far, only) time the Court has made use of this provision. Ordering that Walter LaGrand should not be executed pending a final decision in the case,[38] the Court observed that in the event of ‘extreme urgency’, it may order provisional measures without holding oral hearings.[39]
In his separate opinion, Judge Schwebel indicated his ‘profound reservations’ regarding the procedure followed by the Court.[40] In particular, he was concerned not only with the lack of an oral proceeding but also with the fact that the US was denied the opportunity to present written observations.[41] He moreover considered that it was not a genuine application of the rule in Article 75, as it could not be an order proprio motu since Germany had initiated the order via its request for provisional measures.[42]
More recently, in Armenia’s proceedings against Azerbaijan provisional measures were indicated without a hearing on two occasions. The case was instituted on the basis of alleged breaches of the Convention for the Elimination of Racial Discrimination in the context of an ongoing conflict in the Nagorno-Karabakh region.[43] Azerbaijan has issued proceedings against Armenia likewise alleging breaches of the same convention.[44] These cases, while not formally joined,[45] mirror one another.
The Court first indicated provisional measures in both cases in December 2021, following which in September 2022, Armenia, referring to Article 76(1), requested a modification of the Court’s Order. Azerbaijan filed written observations, but no oral hearing was conducted before the Court, two weeks later, found that the circumstances did not require the modification of the original December 2021 Order.[46]
Armenia then filed a new request for provisional measures in December 2022. Hearings were conducted, and in February 2023 the Court indicated a new provisional measure related to movement through the Lachin Corridor. This was again followed by a request for modification; the Court received written communications from the parties before concluding (in July, several weeks after receiving the request in May) that the circumstances did not require the modification of the February 2023 Order.[47]

b) Intervention

Decisions relating to the admissibility of Article 63 declarations of intervention have also been handed down without an oral hearing. On 9 June 2023, the Court uploaded to its website an order regarding the admissibility of the many Article 63 interventions in the Ukraine v Russia genocide case.[48] Despite Russian objections, all of the 32 interventions from 33 States were admitted bar one; the United States intervention was inadmissible at the preliminary objections phase due to its reservation to Article IX of the Genocide Convention.[49] Russia’s objection triggered Article 84(2) of the Rules, which requires the Court to ‘hear the State seeking to intervene and the parties’ before deciding on admissibility. However, the Court proceeded in writing alone. Russia had in fact requested an oral round of hearings,[50] but this was denied for reasons not explained by the Court. It is possible that the large number of rather repetitive declarations motivated the Court to avoid oral hearings in the name of efficiency,[51] certainly this might be implied from the Court’s emphasis on ‘procedural efficiency’ when it set an early deadline for States to intervene.[52] But the lack of oral hearings may also have been a contributing factor in Russia electing to re-litigate some of the issues relating to the interventions at the preliminary objections phase.[53]
The Court again utilised a purely written procedure in respect of the admissibility of the declarations of intervention submitted by the Maldives and jointly by Canada, Denmark, France, Germany, the Netherlands and the United Kingdom in The Gambia v Myanmar.[54] It may be that decisions on the papers are the new normal for situations of ‘mass intervention’,[55] although it remains to be seen what procedure the Court will follow in South Africa v Israel, where at the time of writing 10 States have applied to intervene using a combination of Articles 62 and 63 of the Statute. This adds a layer of complication not seen in the previous cases, as Article 62 enables intervention as a party.[56]
The Court also did not hold oral hearings in respect of El Salvador’s attempted intervention in Military and Paramilitary Activities in and against Nicaragua (Nicaragua v United States of America),[57] but in this instance Nicaragua did not formally object to the intervention, and as such El Salvador had no right to a hearing under Article 84(2) of the Rules. Judges Ruda, Ago, Mosler, Jennings and Lacharrière jointly opined that it would have been ‘more in accordance with judicial propriety’ for an oral hearing to have taken place,^[58] since Nicaragua had drawn the Court’s attention to ‘certain deficiencies both in form and substance’;^[59] an objection de facto if not de jure.

c) Interpretation

Finally, the Court twice rejected applications for interpretation of a judgment without oral hearings. In Land and Maritime Boundary between Cameroon and Nigeria,[60] the Court made a decision to reject Nigeria’s request for interpretation without holding a hearing – despite having had to do so for the swearing in of the judges ad hoc and again for the reading of the judgment.[61] The Court suggested that it had sufficient information in the written dossier to reach its decision and as such did not require further explanations by way of oral proceedings.[62] Similarly in the Asylum case,[63] Colombia’s request for interpretation (submitted the same day as the judgment was handed down) was rejected by the Court following an exchange of letters between the parties Agents but no oral hearings.[64]

2.3. Decisions on the papers: Fast and efficient?

So far it is difficult to discern a clear pattern justifying the Court’s decisions to not conduct oral proceedings. Two possible arguments could be made – the need for a decision is so urgent that there is no time to conduct oral proceedings, or a judgment on the papers is a more economical use of the Court’s limited resources.[65] However, neither provides a complete explanation due to the Court’s inconsistent practices.
Urgency might provide a rationale – certainly the situation in LaGrand was one of extreme urgency and the case may well be sui generis on that front. But the Court has held oral proceedings in other time-sensitive cases, albeit less pressing. In Paraguay v United States, a Paraguayan national was due to be executed on 14 April 1998. Paraguay applied for provisional measures on 3 April, alleging in its application breaches of the Vienna Convention on Consular Relations.[66] The Court conducted oral hearings on 7 April and indicated measures on 9 April 1998.[67] Similarly in the Jadhav case, an Indian national had been sentenced to death in Pakistan while allegedly being denied access to consular assistance.[68] According to Pakistani law, Mr Jadhav had only 40 days in which to file an appeal – India submitted its request for provisional measures 28 days into this timeline on 8 May 2017, following which the Court held public hearings on 15 May and issued its order on 18 May 2017. It would appear that oral proceedings for provisional measures can be held very rapidly when circumstances demand.
The more significant indication that urgency is not the driving force behind decisions to dispense with oral hearings is the longer time frames involved in Armenia v Azerbaijan: two weeks and one month and 24 days, respectively. Thus, the Court elected to not conduct oral proceedings despite there being plenty of time in which to do so. Urgency may be a motivation in some cases, but it is not the only factor motivating the Court.
Another rationale could be that the Court, faced with a patently unmeritorious application, is being efficient by not holding a hearing. Efficiency in this sense means the resolution of a case ‘at the earliest possible date and at the lowest possible cost to the parties’.^[69] This was perhaps implicit in the Cameroon v Nigeria Interpretation decision;^[70] the Court deemed that it had the information required and did not need to waste resources on further hearings. Yet in that case there were three dissenting opinions, which suggests that the request for interpretation was far from groundless. Judge ad hoc Ajibola in particular was concerned that the lack of oral hearing, or at least the lack of two rounds of written procedure, had impinged on the principle of audi alteram partem.[71] Moreover, this does not explain the March Order in South Africa v Israel, where measures were indicated despite the lack of a hearing.
As the case law does not provide a ready explanation, we must return to first principles. The following Part explains that oral proceedings expressively manifest procedural fairness values. There are three core value elements of procedural fairness – equality, dignity, and impartiality – and oral proceedings perform and embody these values. Oral proceedings are thus significant in and of themselves,^[72] and not merely an inefficient step in the production of a judgment. As such, the Court should be conscious of these values when deciding whether to dispense with oral proceedings and overall adopt (and articulate) a more consistent and principled approach.

3. The value(s) of oral proceedings

What is the point of oral proceedings? While grand, this phase is often perceived as being a ‘critical weakness’ in the Court’s procedure.^[73] Scholars, judges, and counsel have noted that the oral proceedings can be repetitive and dull,^[74] unnecessarily prolix,^[75] underproductive,^[76] a ‘dialogue of the deaf’,^[77] and even irrelevant.^[78] There are regular calls for reform, including suggestions for a reduction in the repetition of the written material and greater engagement from the bench.^[79] Despite such persistent criticism, the oral proceedings are equally considered ‘the highpoint to proceedings, when the parties will have the Court’s attention in full’.^[80] A former member of the Court has written that the oral arguments ‘have served to stimulate deep and careful consideration of the issues by the Court’.^[81] In the end, the oral phase seems to ‘matter’.^[82]
This is because the Court during oral proceedings is not only hearing the arguments of the parties,^[83] but is also operating in an affective register using ritual,^[84] and symbolic communication.^[85] Through the oral proceedings, the Court not only performs justice, but also performatively enacts it.^[86]
In respect of the former, scholars regularly note the parallels between performance and courts;^[87] Bentham famously called the courtroom a ‘theatre of justice’.^[88] Oral proceedings rely upon the same key elements as a theatrical performance: ‘actors, stages, scripts and audiences’.^[89] State representatives can even engage dramatic rhetoric,^[90] such as James Crawford’s mock declaration of the independence of South Australia during the Kosovo Advisory Opinion.^[91] The parties perform for two audiences: the Court itself, and the general public,^[92] relying on both verbal and nonverbal forms of communication.^[93] The latter was aptly demonstrated during the 2022 Arbitral Award of 3 October 1899 (Guyana v Venezuela) oral proceedings, during which Venezuela’s Agent wore a pin in the geographical shape of Venezuela including the Essequibo, the region in dispute between the parties.^[94]
But the oral proceedings can also be read performatively,^[95] as ‘reiterated and referential practices that produce social effects over time’.^[96] The formal, ritual-like characteristics of oral proceedings, such as the black robes symbolizing judicial authority,^[97] or the use of ‘stylized and ritualistic dialogue’,^[98] can improve the effectiveness of the actions and statements made during the proceedings.^[99] These symbols and rituals are ‘constitutive’ of the Court and its judicial process.^[100] Empirical studies have shown that the use of judicial symbols during oral proceedings strengthens institutional legitimacy and increases the chances that the decision will be accepted by the parties,^[101] even when the decision is disappointing.^[102] The Court therefore relies on the symbolism and ritual which take place during oral proceedings in order to foster its own legitimacy and authority.^[103]
The activities taking place during the oral proceedings can thus ‘entrench certain practices and modes of thought, naturalizing and making them seem self-evident through each repeated performance’.^[104] The Court’s oral proceedings serve to demonstrate (as performance) and constitute (as performatives) procedural fairness.^[105]

3.1. Defining procedural fairness

There is sufficient evidence of a set of common core values constituting procedural fairness across international courts and tribunals.^[106] Ruiz Fabri and Paine identify as elements of procedural fairness ‘ensuring equality of arms between the parties, principe du contradictoire, and expectations around the independence and impartiality of adjudicators and the transparency of proceedings’.^[107] To this list may be added the idea that equal participation in open court promotes the value of ‘dignity’;^[108] in particular that there is an independent value in adjudicative practices that afford parties ‘the opportunity to tell their story in a meaningful way’.^[109] As Lahav notes, permission to tell one’s story may be considered ‘the irreducible minimum of just procedure’.^[110] Individual judges have also made reference to particular concepts such as equality as elements of procedural fairness.^[111]
In sum, it appears that the core values constituting procedural fairness are equality between the parties, participant dignity, and the impartiality of the Court. Even if no more than a narrow overlapping consensus,^[112] this is sufficient for the purposes of situating the concept of fairness in respect of the oral proceedings of the Court.

a) Party equality

The first value inherent in procedural fairness is party equality – the correlative ideas of the right of each party to be heard, and the procedural equality of the parties^[113] – captured in the maxim audi alteram partem or principe du contradictoire.^[114] Indeed, Strong suggests that equality and the right to be heard are so well accepted that they constitute a form of procedural ‘jus cogens’.^[115]
Procedural equality is intended to ‘guarantee that neither party derives any special and material procedural advantage over the other’.^[116] In its contentious proceedings, the Court has endorsed this value, in particular noting that procedural equality requires that each party have an equal opportunity to present its case and to comment on its opponent’s case.^[117] Equality is, in the words of the Court, an ‘integral’ constituent of the rule of law.^[118] The Court’s Statute embeds equality of treatment; for example in Article 35(2) which states that even in the case of ad hoc procedural arrangements for States that are not parties to the Statute, in no case shall the conditions established for access to the Court ‘place the parties in a position of inequality before the Court’.^[119]
The Court has addressed procedural equality in some detail in the Advisory Opinions mentioned in Part 2, above. These are an unusual feature of the Court’s docket, as the tribunals under review decide cases between individuals and international organisations, usually related to the employment of the former by the latter. In this context, the Court has considered the ‘inherent inequality’ between the entities involved, as individuals have no right to appear before the Court,^[120] and has dispensed with oral proceedings as a means of countering this inequality.^[121] The Court’s solution to party inequality in the Complaints Made against UNESCO Advisory Opinion was to accept written statements from individuals provided that they were transmitted to the Court via UNESCO.^[122] UNESCO undertook to make the transmission ‘directly to the Court, without checking the contents’.^[123] The Court considered this workaround sufficient to meet the requirements of equality.^[124] However, a number of the judges considered that the Court should have declined to issue an Opinion, on the basis that ‘the absence of oral proceedings constituted either an insuperable or a serious obstacle’.^[125] The Court noted in a later Opinion that it ‘took the view that any absence of equality … is capable of being cured by the adoption of appropriate procedures which ensure actual equality in the particular proceedings’,^[126] albeit that this was not a matter ‘free from difficulty’.^[127]
In 2012, the Court faced a particularly critical problem of unequal parties, as the International Fund for Agricultural Development not only attempted to insist upon its right to be heard orally,^[128] but also failed to transmit its opponent’s documents to the Court in a timely fashion.^[129] While the Court proceeded to give an Opinion, as it had done in the previous similar cases, it was particularly critical of the inequalities created by the procedures used.^[130] Judge Cançado Trindade said ‘[t]he procedural acrobatie is not to hold oral hearings: this is not a solution either, as the Court thereby ends up depriving itself to instruct better the dossier of the case, by imposing such limit to the freedom of expression of the “parties” concerned’.^[131] Likewise, Judge Greenwood stated that such inequality between the parties ‘is incompatible with modern notions of justice and due process’.^[132]
The Court’s practice in respect of these Advisory Opinions suggests that the Court holds the equality of the parties as one of its highest order values, and also that it perceives the oral proceedings as having a particular importance related to equality. Not only was there a practical concern that the Judges would not have access to sufficient information without the oral proceedings,^[133] but the inability of one party to stand before the Court amounted to an inherent inequality as well as an imposition on party freedom.
The Court’s practice in these cases does not make explicit, however, the relationship between oral proceedings and party equality, particularly given that the Court was able to proceed to render Advisory Opinions purely on the papers.^[134] Moreover, the general principle of audiatur et altera pars can arguably be complied with by proceeding in writing.^[135] But while oral proceedings are not the only means through which equality can be pursued, they are its optimal manifestation. Viewed through an expressive lens, the Court’s oral proceedings enact equality both physically and symbolically, and for that reason give an important message to State parties, underscoring the fact that they will be treated fairly.^[136]
Oral proceedings express and make real equality; placing the parties in a position of adversarial equality despite any power differences that may exist outside of the courtroom.^[137] Parties are placed on an equal footing, physically, and are allocated equal speaking time.^[138] The Court actively listens to both parties in an open and transparent manner. Moreover, the parties are required to engage with the Court and one another using the same language of legal forms.^[139] Parties are judged on the quality of their argument alone. No matter how powerful a party is outside of the courtroom, inside they ‘must operate as equals of their adversaries’.^[140] Inside the highly formal, ritualised space of the Great Hall of Justice, the treatment of States as procedural equals contributes to possibilities of actual equality.^[141] While in other fora informal disputing can magnify power imbalances, the Court’s oral proceedings are able to give a ‘material presence’^[142] to sovereign equality and permit States to ‘escape power relationships’^[143] temporarily.

b) Dignity

The second value identifiable within the broader concept of procedural fairness is dignity. Fairness, Kennedy tells us, ‘is a feeling, the result of participation in a discourse’.^[144] Procedural fairness requires that ‘those affected by a decision have the option to participate in the process by which the decision is made’.^[145] The value of participation is in turn connected to the dignity of the participants.^[146] It is based on the almost instinctive notion that ‘there is a difference between losing and being treated unfairly’.^[147] It is possible to leave the courtroom victorious ‘and yet dissatisfied’,^[148] on the basis that one’s voice was not heard.^[149]
Dignity is a unique concept, albeit closely related to equality.^[150] Khaitan argues that dignity is fundamentally expressive, insofar as it takes seriously ‘the expression of disrespect/insult/humiliation etc to a cherished person, object or value’.^[151] Franck, drawing on Kant, explains that individuals possess dignity ‘[b]ecause persons are ends, and may not be used merely as means, they are above all price, and have no equivalent’.^[152] In general terms, participation in litigation affirms the value of dignity by allowing those harmed to ‘receive direct recognition’.^[153] Giving the parties a ‘voice’ recognises their dignity and grants them respect,^[154] and ensures the party has an opportunity to participate in a decision that may affect them ‘profoundly’.^[155] Thus, the practice of hearing both sides is based on dignity,^[156] as well as in party equality. Litigant stories have an independent value that is not merely expressed vis-à-vis equality with their opponent; the Court will hold oral proceedings even where one party elects not to appear.[157] The idea of one’s right to a ‘day in court’ is powerful.^[158]
Dignified participation also signals that the participants are considered fully-fledged members of their society.^[159] Ensuring the dignity of participants is also a key consideration in the physical design of courtrooms,^[160] because a ‘dignified setting’ conveys the message that the participants are being treated with respect.^[161] The act of listening attentively during oral proceedings reinforces the expression of the participants’ dignity.^[162] Thus the ‘courtroom performance of human dignity becomes real, at least for the duration of the performance itself’.^[163]
The dignitary theory is premised upon the innate moral value of humanity. Can this theory apply to the Court’s oral proceedings? Importing a metaphorical notion of ‘humanity’ to States parties as litigants before the Court is problematic.^[164] However, dignity inheres in sovereign States as well,^[165] and oral proceedings expressively perform and constitute State dignity in a manner analogous to individual dignity in domestic courts. States have ‘feelings’, as Jessup once said.^[166] States can be friends,^[167] or enemies.^[168] The manner in which States relate to one another is governed by ‘a world of feeling and sentiment’,^[169] concerned largely with the question of respect for their inherent dignity as equal sovereigns.^[170]

c) Impartiality

In addition to fostering the inherent value of dignity, ensuring that the parties are able to tell their story and have their day in court also contributes to the third and final value of procedural fairness by expressively demonstrating the Court’s own impartiality and in turn, its legitimacy. If the parties involved, and the public, can witness that arguments have been fully presented to the Court, they will be more likely to accept and comply with the final judgment.^[171] Impartiality is in many respects an internal judicial ethic. If decisions are made behind closed doors on the basis of written submissions, the judge may well be impartial, but the parties (and public) cannot see their impartiality.^[172] Here, oral proceedings expressive functions are indispensable. The place, performance, and symbolism of oral proceedings together operate as ‘social signs that the judge has put aside her individuality and assumed the role of an authority acting “under law” – that is, deciding according to neutral principles’.^[173] A visible performance of impartiality is essential, because there is no other opportunity in the judicial process for certifying that the judges are engaged in a fair assessment of the parties’ respective cases.^[174] Crawford and McIntyre point out that ‘accountability is a necessary counterpoint to judicial independence and impartiality; indeed it is, in many ways, merely the other side of that coin’.^[175] The Court’s oral proceedings are open to the public,^[176] bringing ‘into the public light what would otherwise remain a relatively private and obscure administration of justice’.^[177] It is a key part of the symbolic expression of the Court, permitting the public audience to bear ‘witness to the fairness of proceedings’.^[178] Elimination of oral proceedings removes the one occasion on which the judges can be seen at work.^[179]

3.2. Oral proceedings and the Court’s legitimacy

Taken together, the procedural fairness values expressively performed during oral proceedings make an important contribution to the Court’s legitimacy.^[180] Courts operate in accordance with a principle of publicity,^[181] which in turn sustains the foundation of their authority and power.^[182] While the Court still relies upon party consent as a source of its authority, as the principal judicial organ of the United Nations^[183] it is also engaged in an exercise of public power.^[184] Publicity and transparency are therefore essential attributes of the Court’s work.^[185] The repeated public enactment of equality, dignity, and impartiality during the oral proceedings ‘sustains and perpetuates public and juridical belief’ in the Court as ‘a source of justice’.^[186]
The legitimising performance of oral proceedings is particularly important to balance the Court’s lack of formal enforcement. Luban colourfully argues that international courts ‘bootstrap themselves into legitimacy by the quality of justice they deliver; their rightness depends on their fairness’.^[187] Parties are more likely to submit to judicial resolution before,^[188] and more likely to comply with a judgment that is issued by,^[189] a Court that is perceived as fair and legitimate.^[190] This is so even when the final decision is disappointing.^[191] As Stahn states, expressivism ‘is an instrument to gain “buy-in” for authority’.^[192] For the Court, lacking as it does any supporting enforcement mechanism, the expression of procedural fairness that is performed during oral proceedings is an essential contribution to its legitimacy,^[193] and social authority.^[194]

4. South Africa v Israel: Procedural fairness unperformed?

The importance of understanding the values inherent in oral proceedings and the need to articulate principles regulating their use or non-use is aptly demonstrated by the South Africa v Israel case.
To begin, the Court’s two non-hearing decisions in February and March led to completely different results. The first, rejecting South Africa’s request for an Article 75(1) order, seems logical enough. The Court was declining an invitation to use its proprio motu powers; essentially, the Court was simply acting as it would otherwise have acted. Nothing changed.
But the second decision was fundamentally different in scope. The Court was convinced to modify the provisional measures indicated in the 26 January Order, a course of action which, as Judge Nolte pointed out, is rare in the Court’s jurisprudence.[195] Moreover, it did so on the basis of evidence that neither party submitted nor had the opportunity to comment upon.[196] By deciding on the basis of written submissions alone, the Court arguably deprived the parties of full procedural fairness (the dignity value of the right to be heard) and also lost the opportunity of reinforcing its authority to reach the decision that it did. This is particularly serious when the jurisprudence is underdeveloped; Judge Nolte indicated his concern that the Court’s ruling might be taken to mean that ‘the threshold for modifying, adding or specifying a provisional measure is low’.^[197]
It can be asked whether the Court, having conducted a public hearing in January, felt that the procedural fairness benefits of oral proceedings had been satisfied; particularly given that the March Order was a modification rather than a new application for provisional measures. This would be consistent with the Court’s approach in Armenia v Azerbaijan, where Armenia’s two requests for modification were adjudicated on the papers while oral proceedings were conducted for its new request for provisional measures in December 2022.[198] Although the Court has not expressly articulated any reasons for its decisions to dispense with oral hearings, treating requests for modification as an extension of the original Order (and therefore as part and parcel of the original oral proceedings) seems both logical and appropriately economical. It corresponds with the rule in Article 61(1) that allows the Court to indicate the issues ‘on which it considers that there has been sufficient argument’ (thus limiting the scope of the oral argument[199]) and with the approach taken by the Court to dealing with questions from the bench, whereby parties respond in writing following the conclusion of the oral proceedings.[200]
However, this neat explanation comes undone by the procedure followed in respect of the Court’s final Order in South Africa v Israel. South Africa’s ‘urgent Request for the modification and indication of provisional measures’ dated 10 May referenced both Articles 75 and 76 of the Rules,[201] but the Court elected to treat it as ‘a request for the modification of the Order of 28 March’.[202] Despite this, oral proceedings took place. It appears that initially, the Court intended to proceed by way of a written procedure, as Israel was invited to present written observations on the Urgent Request by 15 May.[203] But for reasons unknown, on 13 May the Court informed the parties that an oral proceeding would take place on 16 and 17 May.
Given the enormous attention focussed on the case, was the Court consciously electing to perform its role in public in order to bolster the effectiveness of its resulting orders? If so, the result was mixed at best. Rather than performing procedural fairness effectively, the oral proceedings gave rise to new concerns about a lack of fairness.
As noted in Part 1, above, during the May hearings South Africa was represented by a full team, while Israel’s external legal counsel was unavailable. The Court did not accede to Israel’s request to postpone the hearings ‘in light of the circumstances’,[204] which may be taken as a coded reference to the fact that the ground offensive in Rafah, first announced in February, had actually begun on 7 May and as such matters were reasonably urgent.[205] The Court’s decision to hold oral hearings in spite of the unbalanced representation available to the parties was criticized by Judge Sebutinde in her Dissenting Opinion, where she noted that:

Indeed, the manner in which the Court conducted the May hearing appears to run contrary to the fundamental values underlying oral proceedings. While Israel was afforded the opportunity to be heard, impeding their access to external counsel arguably negatively impacted their perceptions of dignified treatment and may have degraded the Court’s performance of impartiality. This in turn may have reduced the Court’s overall perceived legitimacy and authority. As pointed out by Olivia Flasch,

Only fairness makes losing tolerable.^[208] In a case such as this one, involving as it does the most serious allegations of wrongdoing and underscored by a factual background of truly apocalyptic human suffering, the Court must do everything in its power to ensure that justice is not only done but is at the same time seen to be done. The expressive capacities of oral proceedings to both perform and performatively enact procedural fairness need to be given the closest attention; at the very least where the Court elects not to hold oral proceedings, it should articulate clearly the reasons behind the decision. Parties should not be left to guess whether or not they will receive a public hearing, nor be left stumbling in response to last minute changes.

* BA LLB/LP Hons (Flin) LLM International Law (Cantab) PhD (Melb); Senior Lecturer in Law, University of South Australia.
[1] J McIntyre, A Plan, ‘The ICJ Goes Viral: Transparency and Sensationalism in South Africa v Israel’ Opinio Juris (31 January 2024).
[2] Application of the Convention on the Prevention and Punishment of the Crime of Genocide in the Gaza Strip (South Africa v Israel) (Provisional Measures) (26 January 2024) (‘South Africa v Israel, Order of 26 January 2024’).
[3] Also covered extensively in the media. See eg: <www.aljazeera.com/news/liveblog/ 2024/1/26/live-icj-to-issue-preliminary-ruling-in-south-africa-genocide-case-against-i>.
[4] South Africa v Israel, Order of 26 January 2024 (n 2) para 54.
[5] ibid 22 para 74.
[6] Convention on the Prevention and Punishment of the Crime of Genocide (adopted 9 December 1948, entered into force 12 January 1951) 78 UNTS 277.
[7] South Africa v Israel, Order of 26 January 2024 (n 2) para 86 [1]-[2].
[8] ibid 86 [3]-[4]. For a detailed analysis see MA Becker, ‘Crisis in Gaza: South Africa v Israel at the International Court of Justice (or the Unbearable Lightness of Provisional Measures)’ (2024) Melbourne J Intl L (forthcoming).
[9] ‘Urgent Request for Additional Measures Under Article 75(1) of the Rules of Court of the International Court of Justice’, Application of the Convention on the Prevention and Punishment of the Crime of Genocide in the Gaza Strip (South Africa v Israel) (‘Urgent Request’).
[10] International Court of Justice, Press Release no 2024/16 (16 February 2024) ‘Decision of the Court on South Africa’s Request for Additional Provisional Measures’.
[11] Application of the Convention on the Prevention and Punishment of the Crime of Genocide in the Gaza Strip (South Africa v Israel) (Provisional Measures) (28 March 2024) (‘South Africa v Israel, Order of 28 March 2024’) para 16.
[12] ibid para 51(2)(a).
[13] ‘Verbatim Record 2024/28’, Application of the Convention on the Prevention and Punishment of the Crime of Genocide in the Gaza Strip (South Africa v Israel) (Provisional Measures) (17 May 2024) 8-9.
[14] Application of the Convention on the Prevention and Punishment of the Crime of Genocide in the Gaza Strip (South Africa v Israel) (Provisional Measures) (24 May 2024) (‘South Africa v Israel, Order of 24 May 2024’).
[15] ibid para 57(2)(a).
[16] See eg: <www.aljazeera.com/news/2024/5/24/icj-orders-israel-to-halt-its-offensive-on-rafah-gaza-in-new-ruling>.
[17] J McIntyre, ‘Consensus, at What Cost?’ Verfassungsblog (25 May 2024).
[18] Discussed in Part 2 infra.
[19] In addition to South Africa v Israel, Ukraine instituted proceedings against Russia in 2022 (Allegations of Genocide under the Convention on the Prevention and Punishment of the Crime of Genocide (Ukraine v Russia) (International Court of Justice, General List No 182)) and The Gambia instituted proceedings against Myanmar in 2019 (Application of the Convention on the Prevention and Punishment of the Crime of Genocide (The Gambia v Myanmar) (International Court of Justice, General List No 178)). In both, orders in respect of intervention were made without an oral hearing. See discussion in Part 2.
[20] South Africa v Israel, Order of 26 January 2024 (n 2) para 72.
[21] South Africa v Israel, Order of 28 March 2024 (n 11) (Judge Nolte) para 1.
[22] Statute of the International Court of Justice (opened for signature 26 June 1945, entered into force 24 October 1945) 33 UNTS 99316 (‘Statute’) art 43(1).
[23] ibid art 43(5).
[24] ibid art 52; International Court of Justice, Rules of Court (adopted 14 April 1978, entered into force 1 July 1978) (‘Rules’) art 72 provides for the reopening of proceedings to facilitate the presentation of further evidence should both parties consent, but although the Court has at times considered it, a reopening has never happened. See Nuclear Tests (Australia v France) (Judgment) [1974] ICJ Rep 253, 263-5; Military and Paramilitary Activities in and against Nicaragua (Nicaragua v United States of America) (Merits) [1986] ICJ Rep 14, 39; JJ Quintana, Litigation at the International Court of Justice (Brill 2015) 378-379. Cf Permanent Court of International Justice, Seventh Annual Report of the Permanent Court of International Justice (1930-1931) [1931] PCIJ Ser E No 7, 301.
[25] Rules (n 24) arts 54, 55, 59.
[26] Rules (n 24) art 70.
[27] Rules (n 24) arts 56, 57, 58, 62-69.
[28] J Crawford, ‘Advocacy Before International Tribunals in State-to-State Cases’ in D Bishop, EG Kehoe, The Art of Advocacy in International Arbitration (2^nd edn, Juris 2010) 303, 323.
[29] R Kolb, The Elgar Companion to the International Court of Justice (Edward Elgar, 2014) 321. This was the case in United States Diplomatic and Consular Personnel at Tehran: International Court of Justice, Annuaire–Yearbook No 34 (1979-1980) 127.
[30] Eg Certain Questions of Mutual Assistance in Criminal Matters (Djibouti v France), where a document was required to be kept secret by French law: International Court of Justice, Annuaire–Yearbook No 62 (2007-2008) 199, 338.
[31] Rules (n 24) art 74(2).
[32] ibid.
[33] For proceedings in Chambers, an oral proceeding is not mandatory: Rules (n 24) art 92(3). Cf Treaty of Neuilly, Article 179, Annex, Paragraph 4 (Interpretation) (Judgment) [1924] PCIJ Ser A No 3, 5.
[34] S Talmon, ‘Article 43’ in A Zimmermann et al (eds), The Statute of the International Court of Justice: A Commentary (3^rd edn, OUP 2019) 1088, 1129-30; Quintana (n 24) 350.
[35] Shabtai Rosenne is critical of not holding oral proceedings for incidental phases: ‘Controlling Interlocutory Aspects of Proceedings in the International Court of Justice’ (2000) 94(2) AJIL 307, 308.
[36] Rules (n 24) arts 59 and 94, amended on 25 June 2020. See also G Pinzauti, P Webb, ‘Litigation before the International Court of Justice during the Pandemic’ (2021) 34 Leiden J Intl L 787.
[37] Vienna Convention on Consular Relations (adopted 24 April 1963, entered into force 19 March 1967) 596 UNTS 261.
[38] LaGrand (Germany v United States of America) (Provisional Measures) [1999] ICJ Rep 9, 16 para 29(1)(a).
[39] ibid 14 para 21.
[40] ibid (Judge Schwebel) 22.
[41] ibid. Judge Oda also expressed concerns with the Order, but on the basis that in his view the Order was not directed to the preservation of the rights of Germany (ibid 19).
[42] ibid (Judge Schwebel) 22.
[43] Application of International Convention on the Elimination of All Forms of Racial Discrimination (Armenia v Azerbaijan) (International Court of Justice, General List No 180, 16 September 2021).
[44] Application of International Convention on the Elimination of All Forms of Racial Discrimination (Azerbaijan v Armenia) (International Court of Justice, General List No 181, 23 September 2021).
[45] Rules (n 24) art 47. See eg Certain Activities Carried Out by Nicaragua in the Border Area (Costa Rica v Nicaragua) (Joinder) [2013] ICJ Rep 166 and Construction of a Road in Costa Rica along the San Juan River (Nicaragua v Costa Rica) (Joinder) [2013] ICJ Rep 184.
[46] The full procedural history is recounted in Application of the International Convention on the Elimination of All Forms of Racial Discrimination (Armenia v Azerbaijan) (Request for the Modification of the Order of 22 February 2023) [2023] ICJ Rep 403, 404-405.
[47] ibid.
[48] J McIntyre, K Wigard, O Pomson, ‘A Hot Potato: The ICJ’s Order on the Admissibility of 32 Declarations of Intervention in Ukraine v Russia’ EJIL:Talk! (13 June 2023).
[49] Genocide Convention (n 6).
[50] Allegations of Genocide under the Convention on the Prevention and Punishment of the Crime of Genocide (Ukraine v Russia) (Admissibility of the Declarations of Intervention) (5 June 2023) 6.
[51] K Wigard, O Pomson, J McIntyre, ‘Keeping Score: An Empirical Analysis of the Interventions in Ukraine v Russia’ (2023) 14 J Intl Dispute Settlement 305; J McIntyre, ‘Less a Wave Than a Tsunami: Procedural Implications for the ICJ of the Article 63 Interventions in Ukraine v Russia’ Völkerrechtsblog (11 October 2022).
[52] Ukraine v Russia (Admissibility of the Declarations of Intervention) (n 50) 5.
[53] ‘Verbatim Record 2023/13’, Allegations of Genocide under the Convention on the Prevention and Punishment of the Crime of Genocide (Ukraine v Russia) (Admissibility of the Declarations of Intervention) (18 September 2023) 96-98.
[54] Application of the Convention on the Prevention and Punishment of the Crime of Genocide (The Gambia v Myanmar) (Admissibility of the Declarations of Intervention) (3 July 2024) 4 [18].
[55] B McGarry, ‘Mass Intervention?: The Joint Statement of 41 States on Ukraine v. Russia’ EJIL:Talk! (30 May 2022).
[56] Statute (n 22) art 62. See further Land, Island and Maritime Frontier Dispute (El Salvador/Honduras) (Application to Intervene by Nicaragua) [1990] ICJ Rep 92, 135 para 99.
[57] Military and Paramilitary Activities in and against Nicaragua (Nicaragua v United States of America) (Order on Declaration of Intervention by El Salvador) [1984] ICJ Rep 215.
[58] Nicaragua (n 57) 219 (Judges Ruda, Mosler, Ago, Sir Robert Jennings and de Lacharrière).
[59] ‘Written Observations on the Declaration of Intervention’, Military and Paramilitary Activities in and against Nicaragua (Nicaragua v United States of America) (10 September 1984) 465.
[60] Interpretation of the Judgment of 11 June 1998 in the Case concerning the Land and Maritime Boundary between Cameroon and Nigeria (Cameroon v Nigeria) (Preliminary Objections) (Nigeria v Cameroon) [1999] ICJ Rep 31.
[61] Rosenne (n 35) 308.
[62] Interpretation of the Judgment of 11 June 1998 (n 60) 33.
[63] Interpretation of the Judgment of November 20^th 1950 in the Asylum Case (Colombia v Peru) [1950] ICJ Rep 395.
[64] ibid 399-401.
[65] A Sarvarian, ‘Procedural Economy at the International Court of Justice’ (2019) 18 L & Practice Intl Courts Tribunals 74, 90-1. See also B Coleman, ‘The Efficiency Norm’ (2015) 56 British Columbia L Rev 1777, 1813-14.
[66] Vienna Convention on Consular Relations (n 37).
[67] Vienna Convention on Consular Relations (Paraguay v United States of America) (Provisional Measures) [1998] ICJ Rep 248.
[68] Jadhav (India v Pakistan) (Provisional Measures) [2017] ICJ Rep 231.
[69] Arbitral Award of 31 July 1989 (Guinea-Bissau v Senegal) (Judgment) [1991] ICJ Rep 53, 124 (Judges Aguilar Mawdsley and Ranjeva).
[70] Interpretation of the Judgment of 11 June 1998 (n 60) 33.
[71] ibid 56 (Judge Ajibola).
[72] MS Ball, ‘The Play’s the Thing: An Unscientific Reflection on Courts under the Rubric of Theater (1975) 28(1) Stanford L Rev 81, 82.
[73] C Rose, ‘Questioning the Silence of the Bench: Reflections on Oral Proceedings at the International Court of Justice’ (2008) 18 J Transnational L Policy 47, 48.
[74] R Kolb, The International Court of Justice (Hart 2013) 958, 971.
[75] S Rosenne, ‘The 1972 Revision of the Rules of the International Court of Justice’ (1973) 8(2) Israel L Rev 197, 227.
[76] Rose (n 73) 48.
[77] Eg ‘Verbatim Record 2002/17’, Land and Maritime Boundary Between Cameroon and Nigeria (Cameroon v Nigeria) (12 March 2002) 31 (Sinclair).
[78] M Lachs, ‘The Revised Procedure of the International Court of Justice’ in F Kalshoven, PJ Kuyper, JG Lammers (eds), Essays on the Development of the International Legal Order, in Memory of Haro F van Panhuys (Sitjoff and Noordhoff 1980) 21, 35.
[79] A Keene (ed), ‘Outcome Paper for the Seminar on the International Court of Justice at 70: In Retrospect and in Prospect’ (2016) 7 J Intl Dispute Settlement 238, 251-3; International Law Association, Committee on the Procedure of International Courts and Tribunals Final Report (1 May 2020) 33-4.
[80] S Ugalde, JJ Quintana, ‘Managing Litigation before the International Court of Justice’ (2018) 9 J Intl Dispute Settlement 691, 717.
[81] S Oda, ‘The International Court of Justice viewed from the Bench (1976-1993)’ (1993) 244 Recueil des Cours de l’Académie de Droit International 118. See also M Bedjaoui, ‘The “Manufacture” of Judgments at the International Court of Justice’ (1991) 3 Pace Intl L Rev 29, 46.
[82] Ugalde, Quintana (n 80) 718. See also RY Jennings, ‘The Role of the International Court of Justice’ (1997) 68(1) British YB Intl L 1, 14: ‘The oral argument is, in this writer’s opinion and experience, an essential part of the adversarial system of trial’; Pinzauti, Webb (n 36) 793: ‘The prevailing view is that “oral hearings make a difference, even at the ICJ”, even if they take place via video conference’.
[83] Indeed, Tommaso Soave suggests that oral proceedings ‘have little or no value in acquiring additional information from the litigants’, T Soave, The Everyday Makers of International Law: From Great Halls to Back Rooms (CUP 2022) 267-8.
[84] E Craig, ‘The Inhospitable Court’ (2016) 66(2) U Toronto L J 197, 218; L Schirch, Ritual and Symbol in Peacebuilding (Lynne Rienner Publishers 2004) 71; P Kahn, The Cultural Study of Law: Reconstructing Legal Scholarship (U Chicago Press 1999) 79 cited in OG Chase, Law, Culture, and Ritual: Disputing Systems in Cross-Cultural Context (New York U Press 2005) 119; DI Kertzer, Ritual, Politics, and Power (Yale U Press 1988) 132. Contra H Ruiz Fabri, ‘La justice procédurale en droit international’ (Speech, Académie de droit international, 11 July 2022). See also H Ruiz Fabri, ‘La justice procédurale en droit international’ (2023) 432 Recueil des Cours de l’Académie de Droit International 14.
[85] B Authers, H Charlesworth, M-B Dembour, E Larking, ‘Introduction to Humanity Rituals Special Issue’ (2018) 9(1) Humanity 63, 63-64.
[86] B Ertür, Spectacles and Spectres: Political Trials, Performativity and Scenes of Sovereignty (Birkbeck, U London, PhD thesis, 2015) 19; RN Bellah, ‘Durkheim and ritual’ in JC Alexander, P Smith (eds), The Cambridge Companion to Durkheim (CUP 2008) 183, 193.
[87] J Allen, ‘Theater of International Justice’ (2012) 3(1) Creighton Intl Comparative L J 131, 137.
[88] J Bentham, Scotch Reform; Considered, with Reference to the Plan, Proposed in the Late Parliament, for the Regulation of the Courts, and the Administration of Justice, in Scotland (2^nd ed, Richard, Taylor & Co 1811) vol 5, letter 2 reiterating the similar arguments Bentham made in his 1791 Panopticon. See also Soave (n 83) 269. Cf H Arendt, Eichmann in Jerusalem: A Report on the Banality of Evil (Penguin 1977) 8 warning against trials that descend into mere show.
[89] G Lynch, Performances of Injustice: The Politics of Truth, Justice and Reconciliation in Kenya (CUP 2018) 20. See also ME O’Connell, The Art of Law in the International Community (CUP 2019) 264; P Lahav, ‘Theater in the Courtroom: The Chicago Conspiracy Trial’ (2004) 16(3) L & Literature 381, 392; C Stahn, Justice as Message (OUP 2020) 252; J Bens, The Sentimental Court: The Affective Life of International Criminal Justice (CUP 2022) 24. For analysis of this approach as applied to international criminal trials see K Leader, ‘The Trial’s the Thing: Performance and Legitimacy in International Criminal Trials’ (2018) 24(2) Theoretical Criminology 241.
[90] A Allo, ‘The ‘Show’ in the “Show Trial”: Contextualizing the Politicization of the Courtroom (2010) 15 Barry L Rev 41, 55; B Kotecha, ‘The Art of Rhetoric: Perceptions of the International Criminal Court and Legalism’ (2018) 31 Leiden J Intl L 939, 943-4; J Boyd White, Heracles’ Bow: Essays on the Rhetoric and Poetics of the Law (U Wisconsin Press 1985) 28.
[91] ‘Verbatim Record 2009/32’, Accordance with international law of the unilateral declaration of independence in respect of Kosovo (10 December 2009) 47 (Crawford). See also examples in Soave (n 83) 268.
[92] Talmon (n 34) 1171.
[93] Ball (n 72) 102.
[94] See screenshot by @OriPomson (Ori Pomson) (Twitter, 17 November 2022, 8:11pm ACDT): <https://twitter.com/OriPomson/status/1593177599244914688>.
[95] It goes beyond the scope of this article to engage closely with the philosophies of speech acts and performativity, but very briefly speech act theory posits the existence of performative utterances: a form of constitutive speaking. See J L Austin, How To Do Things With Words (OUP 1975) and J R Searle, Expression and Meaning: Studies in the Theory of Speech Acts (CUP 1979). For Judith Butler, speaking performatively creates and defines categories of identity (eg in relation to gender): see Gender Trouble: Feminism and the Subversion of Identity (Routledge 1990). For a summary of Butler’s work and how it can relate to law and courts, see M Merill Umphrey, ‘Law in Drag: Trials and Legal Performativity’ (2011) 21(2) Columbia J Gender L 114.
[96] A Alvarez-Nakagawa, ‘Rituals of (dis)possession: Appropriation and Performativity in the Early Modern Law of Nations’ (2022) 10 London Rev Intl L 33, 69.
[97] OG Chase, J Thong, ‘Judging Judges: The Effect of Courtroom Ceremony on Participant Evaluation of Process Fairness-Related Factors’ (2012) 24(1) Yale J L Humanities 221, 223-24.
[98] Pnina Lahav (n 89) 392.
[99] RA Rappaport, Ecology, Meaning, and Ritual (North Atlantic Books 1979) cited in Schirch (n 84) 85.
[100] Soave (n 83) 269.
[101] JL Gibson, M Lodge, B Woodson, ‘Losing, but Accepting: Legitimacy, Positivity Theory, and the Symbols of Judicial Authority’ (2014) 48 L Society Rev 837, 838.
[102] ibid 857. See also C Tomlins, ‘Pursuing Justice, Cultivating Power: The Evolving Role of the Supreme Court in the American Polity’ (2006) 17(1) Researching L 1, 8.
[103] J Allen, ‘A Theory of Adjudication: Law as Magic’ (2008) 41 Suffolk University L Rev 773, 796-802; F Gélinas et al (eds), Foundations of Civil Justice: Toward a Value-Based Framework for Reform (Springer 2015) 17-22; Soave (n 83) 270.
[104] Authers et al (n 85) 67.
[105] Stahn (n 89) 250. On the role of symbolism in the consolidation of rules, see T Franck, The Power of Legitimacy among Nations (OUP 1990) 91-105.
[106] See R Kolb, ‘General Principles of Procedural Law’ in A Zimmermann et al (eds), The Statute of the International Court of Justice: A Commentary (3^rd edn, OUP 2019) 876-77; Quintana (n 24) 14; F Fontanelli, P Busco, ‘What We Talk About When We Talk About Procedural Fairness’ in A Sarvarian et al (eds), Procedural Fairness in International Courts and Tribunals (BIICL 2015) 17, 22.
[107] H Ruiz Fabri, J Paine, ‘The Procedural Cross-Fertilization Pull’ (Max Planck Institute Luxembourg for Procedural Law Research Paper Series, No 6, 2019) 25. For similar lists, see S Schiettekatte, ‘The Faces of Procedure in International Adjudication: Servant, Justice, and Power’ (2020) 34 Temple Intl Comparative L J 227, 239; O’Connell (n 89) 48; A D Lahav, ‘Procedural Design’ (2018) 71 Vanderbilt L Rev 821, 871; CT Kotuby, Jr, LA Sobota, General Principles of Law and International Due Process: Principles and Norms Applicable in Transnational Disputes (OUP 2017) 70 ff; H Stewart, ‘Concern and Respect in Procedural Law’ in W Waluchow, S Sciaraffa (eds), The Legacy of Ronald Dworkin (OUP 2016) 373, 375; Fontanelli, Busco, ‘What We Talk About’ (n 106) 33-35; H Thirlway, ‘Procedural Fairness in the International Court of Justice’ in A Sarvarian et al (eds), Procedural Fairness in International Courts and Tribunals (BIICL 2015) 243, 244; CF Amerasinghe, Jurisdiction of Specific International Tribunals (Martinus Nijhoff 2009) 13-14; B Cheng, General Principles of Law as Applied by International Courts and Tribunals (Grotius 1987) 290 ff. Cf MH Redish, L C Marshall, ‘Adjudicatory Independence and the Values of Procedural Due Process’ (1986) 95 Yale L J 455 who argue that the core of procedural fairness relates to judicial independence. For a summary of the elements common to all fairness theories and how they apply to international criminal trials, see Y McDermott, Fairness in International Criminal Trials (OUP 2016) 27-31 and D Luban, ‘Fairness to Rightness: Jurisdiction, Legality, and the Legitimacy of International Criminal Law’ in S Besson, J Tasioulas (eds), The Philosophy of International Law (OUP 2010) 569, 580.
[108] A Lahav, In Praise of Litigation (OUP 2017) 32.
[109] L B Solum, ‘Procedural Justice’ (2004) 78 Southern California L Rev 181, 273.
[110] Alexandra Lahav, In Praise of Litigation (n 108) 96.
[111] Certain Activities (n 45) 179. Other references to procedural fairness are merely in passing: Judge Donoghue in Question of the Delimitation of the Continental Shelf Between Nicaragua and Colombia Beyond 200 Nautical Miles from the Nicaraguan Coast (Nicaragua v Colombia) (Preliminary Objections) [2016] ICJ Rep 100, 187 identified res judicata as applicable ‘for reasons of procedural fairness’ but does not elaborate further. Judges Al-Khasawneh and Simma in Pulp Mills on the River Uruguay (Argentina v Uruguay) (Judgment) [2010] ICJ Rep 14, 114 argue that invisible experts go against ‘transparency, openness, procedural fairness’.
[112] Ruiz Fabri (n 84).
[113] V S Mani, International Adjudication: Procedural Aspects (Martinus Nijhoff 1980) 12.
[114] Ruiz Fabri, Paine (n 107) 25; Cheng (n 107) 290; Jurisdictional Immunities of the State (Germany v Italy) (Counterclaims) [2010] ICJ Rep 310, 342 (Judge Cançado Trindade).
[115] SI Strong, ‘General Principles of Procedural Law and Procedural Jus Cogens’ (2018) 122 Pennsylvania State L Rev 347, 404. See also H Thirlway, ‘Dilemma or Chimera? Admissibility of Illegally Obtained Evidence in International Adjudication’ (1984) 78 AJIL 622, 626; H Thirlway, ‘Procedural Fairness’ (n 107) 244; AD Lahav, ‘Procedural Design’ (n 107) 871; Kotuby and Sobota (n 107) 70 ff.
[116] Mani (n 113) 15. See also WB Rubenstein, ‘The Concept of Equality in Civil Procedure (2002) 23(5) Cardozo L Rev 1865, 1867.
[117] Nicaragua (Merits) (n 24) 26: ‘The provisions of the Statute and Rules of Court concerning the presentation of pleadings and evidence are designed to secure a proper administration of justice, and a fair and equal opportunity for each party to comment on its opponent’s contentions’. See also Case Relating to the Territorial Jurisdiction of the International Commission of the River Oder [1929] PCIJ Ser A No 23, 45; Border Area and Construction of a Road (Joinder) (n 111) 195-97 (Judge Cançado Trindade); CF Amerasinghe, Evidence in International Litigation (Martinus Nijhoff 2005) 148.
[118] Examination of the Situation in Accordance with Paragraph 63 of the Court’s Judgment of 20 December 1974 in the Nuclear Tests Case (New Zealand v France) (Order) [1995] ICJ Rep 288, 325. See also Allegations of Genocide under the Convention on the Prevention and Punishment of the Crime of Genocide (Ukraine v Russia) (Admissibility of the Declarations of Intervention) para 4 (Gevorgian).
[119] Statute (n 22).
[120] Statute (n 22). See Review of Judgment No 158 of the United Nations Administrative Tribunal (Advisory Opinion) [1973] ICJ Rep 166, 178.
[121] See GA Res 957(X), UN Doc A/RES/957(X) (8 November 1955) which recommended avoiding oral statements in such review proceedings, so as to avoid inequality. Another scenario arose in Consistency of Certain Danzig Legislative Decrees with the Constitution of the Free City (Advisory Opinion) [1935] PCIJ Ser A/B No 59, 41 whereby the Senate of the Free City was able to submit a written memorial, and also to have oral statements made on its behalf, while three minority parties were only allowed to send written notes and could not take part in the oral proceedings. Judge Anzilotti in his separate opinion was critical of ‘the resulting inequality’ (65).
[122] These events are recounted in Review of Judgment No 158 of the United Nations Administrative Tribunal (Advisory Opinion) [1973] ICJ Rep 166, 179. See also Judgments of the Administrative Tribunal of the ILO upon complaints made against the UNESCO (Advisory Opinion) [1956] ICJ Rep 77.
[123] Review of Judgment No 158 (n 122) 179. See also Complaints made against the UNESCO (n 122).
[124] Complaints made against the UNESCO (n 122) 86.
[125] Review of Judgment No 158 (n 122) 180. See also Complaints made against the UNESCO (n 122) 108 (Judge Winiarski) 112 (Judge Klaestad) 114-5 (Judge Zafrulla Khan) 166-7 (Judge Córdova).
[126] Review of Judgment No 158 (n 122) 180.
[127] ibid 183.
[128] ‘Written Statement of the International Fund for Agricultural Development’, Judgment No 2867 of the Administrative Tribunal of the International Labour Organization upon a Complaint Filed against the International Fund for Agricultural Development (International Court of Justice, General List No 146, 29 October 2010) 10, 25, 30. See further Kolb, The International Court of Justice (n 74) 974-75.
[129] Judgment No 2867 of the Administrative Tribunal of the International Labour Organization upon a Complaint Filed against the International Fund for Agricultural Development (Advisory Opinion) [2012] ICJ Rep 10, 60 (Judge Cançado Trindade).
[130] ibid 27-31.
[131] ibid 88 (Judge Cançado Trindade).
[132] ibid 96 (Judge Greenwood).
[133] Contra Soave (n 83) 267-68.
[134] Talmon (n 34) 1130.
[135] ibid.
[136] Vice-President Gevorgian appears cognisant of this when he states that in Ukraine v Russia the addition of 32 intervenors to the case could result in an inequality which is ‘further exacerbated in future oral proceedings’: Ukraine v Russia (Admissibility of the Declarations of Intervention) (5 June 2023) para 3 (Gevorgian).
[137] J Resnik, D Curtis, A Tait, ‘Constructing Courts: Architecture, the Ideology of Judging, and the Public Sphere’ in A Wagner, RK Sherwin (eds), Law, Culture and Visual Studies (Springer, 2014) 515, 520; NW Spaulding, ‘The Enclosure of Justice: Courthouse Architecture, Due Process, and the Dead Metaphor of Trial’ (2012) 24(1) Yale J L Humanities 311, 330.
[138] Soave (n 83) 257.
[139] D Kritsiotis, ‘The Power of International Law as Language’ (1998) 34(2) California Western L R 397, 399, 402.
[140] J Resnik, D Curtis, ‘Representing Justice: From Renaissance Iconography to Twenty-First-Century Courthouses’ (2007) 151(2) Proceedings of the American Philosophical Society 139, 175.
[141] LA Viola, D Snidal, M Zürn, ‘Sovereign (In)Equality in the Evolution of the International System’ in S Leibfried et al (eds), The Oxford Handbook of Transformations of the State (OUP 2015) 221, 226.
[142] T Kelly, ‘Two Cheers for Ritual: The UN Committee Against Torture’ (2018) 9 Humanity: An Intl J of Human Rights, Humanitarianism, and Development 93, 103: ‘The bureaucratic and ritual forms of human rights monitoring can give a material presence to otherwise disparate and intangible aspirations’.
[143] P-M Dupuy ‘Comments on chapters 4 and 5’ in M Byers, G Nolte (eds), United States Hegemony and the Foundations of International Law (CUP 2003) 176, 177. S Steininger, N Deitelhoff, ‘Against the Masters of War: The Overlooked Functions of Conflict Litigation by International Courts’ (2022) 84 L and Contemporary Problems 95, 107 observe that ‘[t]he transferal of conflicts from the battlefield to the courtroom is a strategy often pursued by the less powerful’. See also A A Fatouros, ‘International Law and the Third World’ (1964) 50 Virginia L Rev 783, 791; D Guilfoyle, ‘Litigation as Statecraft: Small States and the Law of the Sea’ (2023) British YB Intl L (advance).
[144] D Kennedy, ‘Tom Franck and the Manhattan School’ (2003) 35 Intl L Politics 397, 433.
[145] Solum (n 109) 259.
[146] See AA Higgins, ‘The Rule of Law Case Against Inconsistency and in Favour of Mandatory Civil Legal Process’ (2019) 39(4) Oxford J Legal Studies 725, 726 and discussion in Solum (n 109) 262 ff.
[147] JL Mashaw, Due Process in the Administrative State (Yale U Press, 1985) 162-3. See also Solum (n 109) 263.
[148] KF Röhl, ‘Procedural Justice: Introduction and Overview’ in KF Röhl, S Machura (eds), Procedural Justice (Routledge 1997) 1, 13.
[149] Gélinas et al (n 103) 34.
[150] T Khaitan, ‘Dignity as an Expressive Norm: Neither Vacuous Nor a Panacea’ (2012) 32(1) Oxford J Legal Studies 1, 3.
[151] Khaitan (n 150) 4.
[152] TM Franck, The Power of Legitimacy Among Nations (OUP 1990) 214 referencing I Kant, The Philosophy of Law tr W Hastie (2^nd edn, 1798). See also JT Gleeson, ‘Law, Values and the Advocate’ in JT Gleeson, RCA Higgins (eds), Constituting Law: Legal Argument and Social Values (Federation Press 2011) 4, 23-4.
[153] A Lahav, In Praise of Litigation (n 108) 32. See also OM Fiss, ‘The Allure of Individualism’ (1993) 78 Iowa L Rev 965, 978: ‘participation has a value in its own right, manifesting a public commitment to the dignity and worth of the individual’.
[154] Stahn (n 89) 312; TR Tyler, ‘Procedural Justice and the Courts’ (2007) 44 Court Rev 26, 30.
[155] EE Sward, ‘Values, Ideology, and the Evolution of the Adversary System’ (1989) 64(2) Indiana L J 301, 310.
[156] Gleeson (n 152) 24.
[157] For example, Russia refused to participate in the 2022 provisional measures oral proceedings: Allegations of Genocide under the Convention on the Prevention and Punishment of the Crime of Genocide (Ukraine v Russia) (Provisional Measures) [2022] ICJ Rep 211
[158] J Resnik, ‘The Contingency of Openness in Courts: Changing the Experiences and Logics of the Public’s Role in Court-Based ADR’ (2015) 15 Nevada L J 1631, 1668; Solum (n 109) 262-3.
[159] ibid. See also AE Lind, TR Tyler, The Social Psychology of Procedural Justice (Plenum 1988) 37.
[160] L Mulcahy, E Rowden, The Democratic Courthouse: A Modern History of Design, Due Process and Dignity (Routledge 2020) 78; cf Gélinas et al (n 103) 14-15.
[161] Chase and Thong (n 97) 236. See also Ball (n 72) 113.
[162] E Kidd White, ‘On Emotions and the Politics of Attention in Judicial Reasoning’ in A Amaya, M Del Mar (eds), Virtue, Emotion and Imagination in Law and Legal Reasoning (Hart, 2020) 101, 116.
[163] Allen, ‘Theater of International Justice’ (n 87) 141.
[164] RN Lebow, National Identities and International Relations (CUP 2016) 35. See also J d’Aspremont, ‘The Doctrine of Fundamental Rights of States and Anthropomorphic Thinking in International Law’ (2015) 4 Cambridge J Intl Comparative L 501, 504 who refers critically to ‘anthropomorphic constructions’. See further Marshall Islands v United Kingdom [2016] ICJ Rep 833, 1093 (Judge Crawford): ‘that indefinite object, the state of mind of a State’.
[165] Mani (n 113) 15; Kolb, ‘General Principles’ (n 106) 969; P Palchetti, ‘Making and Enforcing Procedural law at the International Court of Justice’ (2019) 61 QIL-Questions of International Law 5, 19; R Higgins, ‘Respecting Sovereign States and Running a Tight Courtroom’ (2001) 50 ICLQ 121.
[166] P C Jessup, A Modern Law of Nations: An Introduction (Macmillan Co 1948) 28.
[167] Treaties of friendship have a long lineage, eg: Articles of Peace, Friendship & Entrecourse, Concluded and Agreed between England and France, In a Treaty at Westminster, 3 November 1655.
[168] See G Simpson, The Sentimental Life of International Law: Literature, Language, and Longing in World Politics (OUP 2021) 153.
[169] ibid 169.
[170] G Simpson, Great Powers and Outlaw States: Unequal Sovereigns in the International Legal Order (CUP 2009) 54. See also C Cruise O’Brien, Feliks Topolski, The United Nations: Sacred Drama (Hutchinson & Co 1968) 66.
[171] Talmon (n 34) 1171.
[172] Soave (n 83) 270; Ruiz Fabri and Paine (n 107) 31.
[173] Chase and Thong (n 97) 226; D Wasserman, ‘The Procedural Turn: Social Heuristics and Neutral Values’ in KF Röhl, S Machura (eds), Procedural Justice (Routledge 1997) 37, 37.
[174] AD Lahav, ‘Rites without Rights: A Tale of Two Military Commissions’ (2012) 24(1) Yale J L Humanities 439, 439; Allen, ‘Theater of International Justice’ (n 87) 139; Soave (n 83) 270. See also J Stone Peters, Law as Performance: Theatricality, Spectatorship, and the Making of Law in Ancient, Medieval, and Early Modern Europe (OUP 2022) 298; F Baetens, ‘Transparency Across International Courts and Tribunals – Enhancing Legitimacy or Disrupting the Adjudicative Process?’ (2022) 91 Nordic J Intl L 595, 598; A Peters, ‘Towards Transparency as a Global Norm’ in A Bianchi, A Peters (eds), Transparency in International Law (CUP 2013) 534; K Leader, Trials, Truth-Telling and the Performing Body (U Sydney, PhD thesis, 2008) 94.
[175] J Crawford, J McIntyre, ‘The Independence and Impartiality of the “International Judiciary”’ in S Shetreet, C Forsyth (eds), The Culture of Judicial Independence (Brill 2012) 187, 199.
[176] Statute (n 22) art 46; Rules (n 24) art 59. See also discussion in M Shaw, Rosenne’s Law and Practice of the International Court: 1920-2015 vol III (5^th edn, Brill 2017) para 317; D Bethlehem, ‘Submissions on Points of Fact and Law: Written and Oral Pleadings Before the International Court of Justice’ in F Weiss (ed), Improving WTO Dispute Settlement Procedures: Issues and Lessons From the Practice of Other International Courts and Tribunals (Cameron May 2000) 175, 176; A von Bogdandy, I Venzke, ‘In Whose Name? An Investigation of International Courts’ Public Authority and Its Democratic Justification (2012) 23(1) Eur J Intl L 7, 26.
[177] Rose (n 73) 54; Soave (n 83) 270.
[178] Stahn (n 89) 309.
[179] JA Chapper, ‘Oral Argument and Expediting Appeals: A Compatible Combination’ (1983) 16(3) U Michigan J L Reform 517, 517; RY Jennings, ‘The International Court of Justice After Fifty Years’ (1995) 89(3) AJIL 493, 497; Soave (n 83) 270.
[180] N Grossman, ‘The normative legitimacy of international courts’ (2013-2014) 86 Temple L Rev 61, 67, 81. Grossman argues that procedural fairness alone is insufficient to sustain the legitimacy of a court. I do not suggest that fairness is the sole source of legitimacy, but it is certainly a contribution. See also F Fontanelli, P Busco, ‘The Function of Procedural Justice in International Adjudication’ (2016) L & Practice Intl Courts Tribunals 1, 4; Fontanelli and Busco, ‘What We Talk About’ (n 106) 26; Jennings, ‘The Role of the International Court of Justice’ (n 82) 14-5; Soave (n 83) 270.
[181] See J Bentham, Rationale of Judicial Evidence (1827) in The Works of Jeremy Bentham, Published under the Superintendence of his Executor, John Bowring (William Tait, 1843) 352: ‘establishment of publicity … will stand recommended as the general rule, as being, in most cases, conducive to the direct ends of justice’. See also S Forlati, The International Court of Justice – An Arbitral Tribunal or a Judicial Body? (Springer 2014) 101. Cf the Kantian principle, by which laws must be able to be publicly proclaimed in order to be morally justified: ‘All actions relating to the rights of other men are wrong, if the maxims from which they follow are inconsistent with publicity’. I Kant, Perpetual Peace: A Philosophical Essay tr Smith (3^rd edn 1795, Allen & Unwin 1917) 185. See further KR Davis, ‘Kantian “Publicity” and Political Justice’ (1991) 8 History of Philosophy Quarterly 409 and J Clinger, ‘The Kantian Publicity Principle and the Transparency Presumption in Public Affairs: When Private Citizens’ Participation Becomes Public Knowledge’ (2017) 19 Public Integrity 394.
[182] J Resnik, ‘The Contingency of Openness in Courts: Changing the Experiences and Logics of the Public’s Role in Court-Based ADR’ (2015) 15 Nevada L J 1631.
[183] Charter of the United Nations (adopted 26 June 1945, entered into force 24 October 1945) 1 UNTS 16 art 92. The Court’s public role is underscored by the fact that there is no fee for States to engage the Court; the institution is financed exclusively by the United Nations. See further Kolb, The International Court (n 74) 1002-06. See also A Pellet, ‘Strengthening the Role of the International Court of Justice as the Principal Judicial Organ of the United Nations’ (2004) 3 L & Practice Intl Courts Tribunals 159, 163.
[184] von Bogdandy and Venzke, ‘In Whose Name?’ (n 176) 17; Y Kryvoi, ‘Private or Public Adjudication? Procedure, Substance and Legitimacy’ (2021) 34 Leiden J Intl L 681, 691.
[185] Quintana (n 24) 340; S Rosenne, The Perplexities of Modern International Law (Martinus Nijhoff 2004) 85; A Wiik, Amicus Curiae before International Courts and Tribunals (Hart/NOMOS 2018) 63; Kolb, The International Court (n 74) 994.
[186] Leader, Trials, Truth-Telling and the Performing Body (n 174) 258. See also S Falk Moore, ‘Selection for Failure in a Small Social Field: Ritual, Concord, and Fraternal Strife among the Chagga Kilimanjaro, 1968-1969’ reprinted in BG Myerhoff, S Falk Moore (eds), Symbols and Politics in Communal Ideology (1975) 109, 112 cited in Chase, Law, Culture, and Ritual (n 84) 116.
[187] Luban (n 107) 579. Cf Y Shany, ‘International Courts in a Politicized World’ (24 October 2020): <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3794365>.
[188] D Shelton, ‘Form, Function, and the Powers of International Courts’ (2009) 9 Chicago J Intl L 537, 545.
[189] C Renning, ‘Subjective Procedural Justice and Civil Procedure’ in KF Röhl, S Machura (eds), Procedural Justice (Routledge 1997) 207.
[190] Gibson et al (n 101) 838.
[191] ibid 857.
[192] Stahn (n 89) 45, 206. See also J Stone Peters, ‘Legal Performance: Good and Bad Law’ (2008) 4 Law, Culture and the Humanities 179, 180. See generally E Goffman, Interaction Ritual – Essays on Face-to-Face Behavior (Routledge 1967, 2017).
[193] T Franck, ‘Fairness in the International Legal and Institutional System’ (1993) 240 Recueil des Cours de l’Académie de Droit International 22, 303-5; S Rigney, Fairness and Rights in International Criminal Procedure (Edinburgh U Press 2022) 66.
[194] F Zarbiyev, ‘Saying Credibly What the Law Is: On Marks of Authority in International Law’ (2018) 9 J Intl Dispute Settlement 291, 312. Like Zarbiyev, I am using the term ‘authority’ in a sociological sense, rather than in the jurisprudential normative sense. See also I Venzke, ‘Understanding the Authority of International Courts and Tribunals: On Delegation and Discursive Construction’ (2013) 14 Theoretical Inquiries in Law 381. For the jurisprudential concept of authority see J Raz, The Morality of Freedom (Clarendon Press 1986) 35-37; J Raz, ‘The Problem of Authority: Revisiting the Service Conception’ (2006) 90 Minnesota L Rev 1003; S Besson, ‘The Authority of International Law – Lifting the State Veil’ (2009) 31 Sydney L Rev 343.
[195] South Africa v Israel, Order of 28 March 2024 (n 11) 1 para 2 (Nolte).
[196] ibid 4 para 22 (Barak).
[197] ibid 2 para 5 (Nolte).
[198] Application of the International Convention on the Elimination of All Forms of Racial Discrimination (n 46) 404-5.
[199] The only time the Court has made use of this provision was in Question of the Delimitation of the Continental Shelf between Nicaragua and Colombia Beyond 200 Nautical Miles from the Nicaraguan Coast (Nicaragua v Colombia) (Decision on the organization of the public hearings) [2022] ICJ Rep 563. It was controversial, with Judges Tomka, Xue, Robinson, Nolte and Judge ad hoc Skotnikov issuing a Joint Declaration indicating that the procedural rights of the parties had not been ‘respected’ (at 566).
[200] Rules (n 24) art 61(2)-(4). See eg ‘Written reply of the State of Israel to the question put by Judge Nolte at the public sitting held on 17 May 2024’ Application of the Convention on the Prevention and Punishment of the Crime of Genocide in the Gaza Strip (South Africa v Israel) (Provisional Measures) (18 May 2024).
[201] Urgent Request (n 9) 1.
[202] South Africa v Israel, Order of 24 May 2024 (n 14) 6 para 20.
[203] ibid 5 para 15.
[204] ibid.
[205] Urgent Request (n 9) 2 para 9.
[206] South Africa v Israel, Order of 24 May 2024 (n 14) 9 para 28 (Sebutinde).
[207] O Flasch, ‘Equality of Arms: Due Process in South Africa v Israel’ Lawfare Blog (22 July 2024).
[208] H Genn, Judging Civil Justice (CUP 2010) 14.

1. Introduction

The present article aims to address an issue that generally attracts little attention: the link between substantive obligations for the protection of collective interests of the international community and procedural rules that govern their implementation. This analysis has been prompted by the recent flood of intervention requests before the International Court of Justice (hereafter ‘ICJ’ or ‘the Court’) in cases relating to genocide, showing the impact that the structure of the substantive rules on genocide has on the procedural regimes applying to them.
Two preliminary remarks are in order. First, we will use the term ‘procedure’ in a very broad sense including not only rules governing judicial proceedings before international courts and tribunals but also, more generally, rules dictating the procedures to be followed for the interpretation, application and implementation of international substantive obligations for the protection of collective interests.[1]
Second, the substantive primary rules that we will consider are the obligations arising under the Genocide Convention, ie the prohibition to commit genocide and the duties to prevent, suppress and punish genocide.[2] The erga omnes character of those rules is the reflection of a particular normative structure devised for the purpose to protect a collective interest of the international community as a whole. This collective normative structure is not to be confused with ius cogens character, which concerns the place of the rules in the hierarchy of international sources. We are rather interested here in the relationship that erga omnes obligations create between the international community (or a group of States if they are erga omnes partes) and each one of its members and the implications of that relationship in terms of the procedures for their interpretation, the ways for establishing their violation, and the means for ensuring compliance with them.
It comes as no surprise that the main consequences of the erga omnes structure become apparent at the stage of the assessment of their respect or breach. This is because the legal relationship they entail is no longer confined to the narrow bilateral relationship between the duty-bearer State and the right-holder State but covers a much broader web of relationships between the former and all the other members of the international community. The settlement of the disputes that may arise in that regard would require to go beyond the traditional bilateral framework of international litigation.
The recognition of the existence of international law rules that protect collective interests inevitably raises the issue of their implementation in a horizontal society whose institutionalization depends on the will of its members. Immediately a tension is produced between the substantive protection of such interests and the procedures that should actually establish their existence and ensure their respect, those procedures being typically private and voluntary in character.[3]
The practice that will be examined below shows that, despite the reluctancy of States to accept organized, centralized or institutionalized procedures for the implementation of erga omnes obligations, their very structure inevitably entails some kind of procedural implications challenging the unilateral power of normative assessment that States have under international law.
The following analysis will be based on the positions of third States recently expressed in their requests to intervene in three genocide cases currently pending before the Court.
In Ukraine v Russia, 33 States submitted declarations of interventions under Article 63 of the ICJ Statute at the preliminary objections stage; the Court admitted all but one.[4] At the merits stage, 6 States have maintained the position expressed in their previous declarations, 8 States have updated their declarations, and 9 States have sent new declarations with two special situations: Poland requested to intervene under both Article 63 and Article 62 of the ICJ Statute (as a non-party), and Austria, the Czech Republic, Finland and Slovenia decided to submit a joint declaration of intervention in relation to the merits stage.[5]
In The Gambia v Myanmar, 2 declarations of intervention under Article 63 have been filed with the Court respectively by Maldives, on the one hand, and by a group of 6 States (Canada, Denmark, France, Germany, the Netherlands and the United Kingdom), on the other; they have both been declared admissible.[6]
In South Africa v Israel, 10 States have submitted requests of interventions under both Articles 62 and 63 of the Statute. Nicaragua requested to intervene under Article 62 (as a party), and Palestine submitted two requests: one under Article 63 and the other under Article 62 (presumably as a non-party, because the requirement of a jurisdictional link is not even mentioned, even though the declaration is extremely succinct and vague as to the object of intervention). All other requests have been introduced under Article 63. The Court has not ruled on their admissibility yet.[7]
Taken as a whole, those declarations are remarkable. Not only for their number which is unprecedented. Most notably, they offer a unique collection of positions concerning the substantive character of the rules prohibiting genocide and the procedures ideally to be followed in the establishment of their violation. It goes without saying that the three disputes are different and, as a consequence, third States have focused on different aspects of the international regime concerning genocide. But there is an important convergence on certain aspects. They draw interesting consequences from the collective structure of primary rules that challenge States’ unilateral power of appreciation. The main reason is that these rules call for collective, if not institutionalized, procedures ensuring their implementation and compliance with them. The contention prompted by this practice is that the normative structure of substantive obligations can hardly be constrained into the straight jacket of bilateralism and that sooner or later this substantive structure impacts the procedural regime that frames their implementation.
The following analysis is organized around a selection of the main arguments advanced by the third States in their requests for intervention. It is to be noted that these arguments are often closely linked one to the other so that a separation may occasionally appear artificial. This separation is nonetheless maintained as these arguments are raised individually in the requests for intervention. The most direct implications regard intervention itself, its scope and organization before the ICJ (Section 2). Beyond these implications that have a procedural character in the narrow sense, as they refer to the conduct of judicial proceedings, third States refer to procedural issues in a broader sense concerning the way in which the substantive rules are to be ascertained: the need for an objective assessment of the breach (Section 3), the preferential recourse to independent sources (Section 4), and the involvement of international institutions (Section 5). Section 6 examines this practice and concludes.

2. Implications for intervention before the Court

The mentioned declarations and requests for intervention in genocide cases before the Court have one main aspect in common. Quite obviously they are all inspired by the need to protect a collective interest of the international community as a whole. Germany’s document provides a good example of a recurring formulation: ‘it follows from this erga omnes character of the obligations enshrined in the Convention that all States parties to the Convention have an interest of their own in the proper interpretation, application and fulfilment of those obligations’.[8] The statement of Luxembourg is more concise but no less effective: ‘les États parties se sont engagés à supprimer le génocide dans le monde entier pour le bien de l’humanité dans son ensemble, et non pour protéger leurs propres intérêts’.[9] Almost all States refer to the erga omnes character of the rule prohibiting genocide and the rulings of the ICJ in their declarations or requests for intervention.[10]
One procedural consequence of this normative structure is straight-forward, that is, the generalization of the locus standi to the omnes. Every State can institute judicial proceedings for the protection of the collective interest. The question of actio popularis for erga omnes obligations has already been widely discussed, analysed and finally settled by the Court at first with respect to the Torture Convention[11] and more recently the Genocide Convention:

More interestingly, the collective interest underlying erga omnes obligations is invoked by would-be intervening States with a similar purpose, that of opening not only the principal proceedings but also intervention to all the omnes. This generally raises no objection in relation to Article 63 and the purposes of ‘interpretive intervention’. The novelty resides in the fact that the same argument is now used to claim that the omnes also possess the ‘legal interest’ required by ‘protective intervention’ under Article 62 so that they could intervene almost automatically at least as non-parties.[13] These two implications will be examined in turn.

a) Interpretive intervention

Admissibility of intervention under Article 63 is premised on the existence of a collective interest shared by the members of the multilateral convention to be interpreted by the Court. Even if this interest needs not be ascertained by the Court – because ‘is presumed by virtue of its status as a party thereto’[14] – it plays an important role.
First, the decision of the Court in Ukraine v Russia that ruled on the admissibility of all declarations of intervention, except for that of the US, clarified the character and role of the collective interest under Article 63. The lack of this interest may be ascertained by the Court and lead to the inadmissibility of the declaration of intervention: ‘the legal interest that the United States is presumed to have in the construction of the Genocide Convention, as a party to that instrument, does not exist in respect of Article IX’[15] because it had entered a reservation to that provision.
The declarations of intervention show that some States were hesitant in that respect and carefully underlined that from the collective interest they could draw a corresponding ‘direct interest’ in the prohibition of genocide. It is true that, lacking representative bodies, action in the name of the collective interest mostly occurs when there is an overlap (of the collective interest) with a private interest and that private interest is often the triggering factor for action in the name of common interests. Indeed, the condition of a legal interest that is implicit under Article 63 and explicit under Article 62 may be confusing. However, the Court made it clear that a general interest is sufficient under Article 63.
The second impact of the collective character of rules prohibiting genocide concerns the spirit of cooperation that, according to many States, should inspire incidental proceedings under Article 63. A number of States inferred from that character a ‘duty to assist the Court’.[16] Some referred, more broadly, to a duty of cooperation among States.[17] Isolated positions brought back the duty to cooperate with the Court to Article IX of the Genocide Convention.[18]
In any case, this spirit of cooperation may have various practical implications. The main one put forward by many third States was their readiness to accept ‘grouping this intervention with similar interventions from other States for future stages of the Proceedings, should the Court deem such a move useful in the interest of good and expedient administration of justice’.[19] This procedural adjustment would ensure at the same time an effective protection of the collective interest (by allowing participation to a considerable number of third States) and sound administration of justice.
It is to be noted that this willingness has already been put into practice. Substance has already had an impact on procedure. Some States decided to institute proceedings jointly, to submit joint declarations of interventions and to present joint observations during the oral phase of preliminary objections proceedings.
Canada and the Netherlands have jointly instituted proceedings against Syria for widespread use of torture on the basis of the common interest underlying the Torture Convention.[20] In Ukraine v Russia, Canada and the Netherlands have acted jointly from the beginning of the proceedings: they have submitted a joint declaration and presented their common position together at the preliminary objection stage. Joint oral statements at that stage were presented by two groups of States: Austria, Liechtenstein, Czech Republic and Slovakia, on the one hand, and Belgium, Croatia, Denmark, Estonia, Finland, Ireland, Luxembourg, Romania and Sweden, on the other hand.[21] At the merits stage, a new joint declaration of intervention has been submitted by Austria, the Czech Republic, Finland and Slovenia. In Gambia v Myanmar, a joint declaration of intervention has been submitted by Canada, the Netherlands, Denmark, France, Germany and the United Kingdom. The Court praised ‘the joint presentation of shared views [that] can advance the good administration of justice.’[22]

b) Protective intervention

The three requests of intervention under Article 62 – submitted by Poland in the Ukraine v Russia case and by Nicaragua and Palestine in the South Africa v Israel case, respectively – have in common the fact of relying on the erga omnes character of the prohibition of genocide to justify the possession of an essential requirement, that is, the legal interest susceptible of being affected by the future decision of the Court.
The reasoning is simple and based on two prongs. First, there is the assumption that the prohibition of genocide protects a collective interest.[23] Intervention under Article 62 is regarded as aimed at the protection of that interest because the interveners may present their views not only on the interpretation of abstract rules, as provided by Article 63, but may also discuss the application of those rules to the facts of the case. Second, it is maintained that if the collective interest is sufficient to justify that the omnes have locus standi in relation to the principal proceedings, that same interest must necessarily suffice to justify the existence of a qualified legal interest under Article 62.[24] Provided that the other conditions are met, intervention under Article 62 should thus be generalized to all the omnes.
The argument according to which the erga omnes character (substance) has this impact on Article 62 intervention (procedure) is not new. It has been supported by members of the Court[25] and international law scholars.[26] Views challenging that possibility have remained isolated.[27] The Court will hopefully soon clarify that issue when ruling on the admissibility of the mentioned intervention requests.
One specific aspect of the request of Nicaragua deserves to be outlined here because it implies a broader interrelation between substance and procedure. The multilateral dimension of genocide is said to entail a duty of cooperation according to which States should have recourse to all procedural means of prevention and protection. Accordingly, recourse to intervention would become the object of an obligation:

3. The need for an objective assessment of the breach

The conditions surrounding unilateral claims of genocide and action that can be taken in that regard are at the heart of Ukraine v Russia. Third States interventions in that case are particularly relevant, but views expressed in the framework of other cases will also be taken into account.
As clarified in its preliminary objections’ decision, the Court has jurisdiction to hear the part of the dispute between Ukraine and Russia that relates to the existence of ‘no credible evidence that Ukraine is responsible for committing genocide in violation of the Genocide Convention in the Donetsk and Luhansk oblasts of Ukraine.’[29] The purpose of the claim of Ukraine is obviously to preclude Russia the possibility to invoke genocide as a lawful basis for its ‘special military operation’ of February 2022. Thus, the Court must decide whether Ukraine had committed genocide before that attack.
Third States having submitted declarations of intervention largely focused on two aspects, namely, the interpretation of the definition of genocide and the way in which claims of genocide should be substantiated. One of the positions on which there is clear consensus among them is that a State cannot be accused of genocide lightly. Such a serious claim is to be made in good faith and needs strong evidence before using the commission of genocide or the risk thereof as a justification for action (especially military action) against the alleged wrongdoer. The most striking point that is explicitly made by many States is that the assessment of the commission/risk of genocide must be made objectively.
The UK considered that

The objective character of the assessment is not clarified further, but it is justified by the multilateral character of the prohibition of genocide and the cooperation at the basis of the Genocide Convention (see below). Similar statements are to be found in the declarations of Italy,[31] Belgium[32] and Luxembourg[33]. According to the latter, ‘pour qu’il y ait genocide en vertu de l’article II, il est necessaire d’établir objectivement et de bonne fois un “acte” et une “intention” de génocide’.
The declaration of Chile submitted in the South Africa v Israel case adds an interesting element: the objective assessment of the Court prevails over the unilateral assessment of Israel. It seems obvious that once the Court has established certain requirements of genocide the respondent State can no longer deny them.[34] The interesting aspect is that the third State refers to the provisional measures stage when the Court’s assessment is only made prima facie. It shows the importance accorded to the centralized establishment of the Court, at least pending the final decision.
It must be recalled that the precise contours of the ‘objective determination’ are rarely provided, and generally the body that is supposed to provide such an objective assessment is not identified. However, the statements can hardly be taken as unintended to hint at the need for some form of centralized or organized assessment, especially when coupled with other explicit positions expressed against the subjective establishment of the commission/risk of genocide.
The declaration of New Zealand refers to both aspects: ‘whether acts amount to “genocide” so as to trigger the application of Article I is not simply a matter of a party’s subjective interpretation … the Court must look to the party taking measures to prevent genocide under Article I of the Convention to prove the objective basis for its determination’.[35] Other States merely exclude that the establishment of genocide could be left to the subjective determination of one interested party.[36]
In accordance with these positions, the erga omnes character of the rule has a clear procedural implication. Just as the interest at the basis of the substantive rule is collective in character, the procedure leading to the establishment of its violation cannot be subjective. The risk of abuse is too high. The assessment can only be made objectively.

4. The preferential recourse to independent sources

In the same vein, the declarations and requests for intervention in the three cases under examination show a clear preference accorded to the use of independent sources. Again, third States do not identify the entity that should carry out the ‘objective’ assessment, but the relevant statements add another piece of the puzzle in that respect.
The explicit reference to the need to have recourse to independent sources is to be found with a first formulation in the Declaration of Lithuania:

The position in that regard found a definitive formulation in the Declaration of Sweden: ‘It therefore constitutes good practice to rely on the results of independent investigations under UN auspices before qualifying a situation as genocide and taking any further action pursuant to the Convention.’[38] This statement is present in many other declarations of intervention with an almost identical wording.[39]
The joint declaration of Canada, Denmark, Finland, the Netherlands, Germany and the UK provides a good example of the link between the erga omnes character of the substantive rule and the need to establish its violation in the most impartial way:

5. The involvement of international institutions

Declarations and requests of intervention plead in favour of a third procedural implication. The breach of erga omnes obligations not only must be established objectively and by having recourse to independent sources, but it should also be carried out collectively, if not entrusted with international institutions.
As stated by Sweden:

Most declarations focus on Articles VIII and IX of the Genocide Convention and interpret these provisions as expressions of the collective character of the prohibition of genocide. The first recommends recourse to UN political bodies; the second is the compromissory clause establishing compulsory jurisdiction of the ICJ.[42] These provisions are interpreted as evidence of this collective character because they translate the need to act first of all (if not exclusively) at the collective level and possibly by having recourse to institutions able to speak in the name of the entire international community, some with binding force.
Some States underline the duty of cooperation that is implicit in the said provisions[43] and the collective assessment that should be preferred in the case of genocide.[44] Other States, explicitly prone recourse to international institutions.
In the vast majority of cases, such recourse is not regarded as compulsory. The point is clearly made by Portugal:

The link with the collective nature of the interest protected by genocide primary rules is explicitly discussed in a number of declarations.[46] In the end, all these positions agree on an interpretation of the Genocide Convention as expressing a priority of collective enforcement over unilateral enforcement.[47] Which does not rule out the latter.
A few States make further steps towards the recognition of a duty to assess genocide by having recourse to collective/institutional means. The Czech Republic considers that:

If litigation before the ICJ is the only available, bilateral means, other types of unilateral assessments are then to be excluded, and the remaining possibility is recourse to collective/institutional means.
On a different vein Estonia considers that ‘the legality of any extraterritorial unilateral preventive measure is contingent on the prior seizing of competent United Nations organs pursuant to Article VIII’.[49] Institutional reaction becomes a pre-condition for having recourse to unilateral action.
Finally, Poland starts by considering that originally Article VIII did not provide for an obligation to consult UN organs. However, it concludes that, due to subsequent practice, a duty has emerged to ‘call upon the competent organs of the UN before they decide on unilateral action’ and that today the Genocide Convention ‘directs States to multilateral institutions to properly assess the situation in an unbiased fashion’.[50] Unfortunately, the declaration lacks precision on two aspects: the relevant subsequent practice and the agreement between the members of the Genocide Convention that it indirectly shows. These are both required under the Vienna Convention.[51] Nonetheless, this position is interesting because it relies on the idea that gradually the procedural implications of erga omnes substantive obligations may crystallize into general international law obligations.

6. Impacts of substance on procedure before and beyond the ICJ

Taken as a whole the views expressed by third States before the ICJ in the three genocide cases under review support a number of procedural implications entailed from the particular structure of the substantive rules prohibiting genocide. They are inspired by a common thread, ie the abandonment of the unilateral approach to the implementation of international rules protecting collective interests.[52] Third States seem to realize how dangerous the unilateral assessment of their existence/compliance can be and the risk of abuse that such an assessment could lead to. The concrete risks of the lack of a centralized assessment for rules protecting community interests have been largely voiced in the past in the works of international law scholars.[53]
Third States plead for a generalization of intervention when erga omnes breaches are at stake and for an objective assessment, based on independent sources and carried out at the collective level, if not the involvement of international institutions. In other words, substantive rules created for the protection of collective interests have procedural impacts both for the conduct of proceedings before the Court and, beyond judicial settlement, for the way in which States appreciate and ensure compliance with substantive obligations.
To be sure, the analysed State practice can hardly evidence more than a ‘common thread’, maybe a ‘tendency’. Nonetheless, to our mind it is somehow revolutionary. This for two main reasons. First, it shows a radical change of attitude of States. It suffices to compare it with the traditional opposition they had with respect to any form of centralized means for ensuring the application of and compliance with erga omnes obligations to realize how significant this common call for collective action is. Second, this remarkable emphasis on the need for collective mechanisms balances the first attempts to afford a procedural protection to erga omnes obligations that relied on a logic of pure bilateralization in framing specific rules in responsibility regimes and judicial dispute settlement.

a) Abandoning unilateral assessment and entering the uncertain ground of collective assessment

The ‘common thread’ brought to the surface by these positions corresponds first to a growing consensus against the traditional power States have to assess unilaterally the existence/breach of obligations having an erga omnes character and the consequences thereof. Rules created for the protection of collective interests call not only for collective reactions, but firstly for a collective assessment of their existence/breach. The genocide cases discussed above seem to have raised States’ awareness on the risks of purely decentralized reactions to the (alleged) breach of erga omnes obligations.
This is something that had been vividly discussed and strongly opposed by States in the past. Especially at the International Law Commission (ILC), when formal proposals had been made to centralize the decision establishing compliance with or the breach of erga omnes obligations, they have systematically been rejected by States that were extremely reluctant to relinquish their unilateral power of assessment and enforcement of international rules. It suffices here to mention the position of Special Rapporteur Gaetano Arangio-Ruiz concerning ‘the indispensable role of international institutions’ in the implementation of international crimes and his proposal to create a special procedural framework for the establishment of their commission.[54] The proposal met such strong an opposition that it was finally excluded from the draft articles on State responsibility adopted on first lecture in 1996.[55] The times were not ripe for such a procedural development.
Instead, current genocide cases before the ICJ show that States seem now ready to acknowledge the need for an ‘organized’ determination[56] of the commission of this crime and that the structure of the underlying obligation is crucial in maintaining such procedural claims. Maybe in the future the same position will be taken with regard to other collective obligations. In our opinion, this generalization is made possible by the normative structure of these obligations and not the specific provisions of treaty regimes, such as Articles VIII and IX of the Genocide Convention.
Yet the 1995 proposal made by Special Rapporteur Arangio-Ruiz deserves to be recalled. It provided for a first political step: it was the task of the political organs of the United Nations to assess whether an allegation that an international crime had been committed was ‘sufficiently substantiated’. In the affirmative, there was a second, judicial step: any member State of the UN could bring the matter – ie the commission of the international crime – to the Court’s attention. In addition, it was provided for the possibility that ‘any other member State’ could join the proceedings before the Court.[57]
It is quite astonishing to see how much overlap there is between this proposal and the positions put forward by the third States in their declarations and requests of intervention examined above. It is clearly impossible to speak of the abandonment of State’s unilateral power of assessment of international law rules. Neither is it claimed that this assessment must always be centralized at the international level.
The positions of third States having intervened or requested to intervene in the three genocide cases are too few and their geographical representation is too limited in certain areas of the world (see the map below designed by the author).

Apart from Turkey, Asia remains silent, and there are only isolated voices coming from Africa (The Gambia, Libya, South Africa) and a minority of Latin American requests (Bolivia, Chile, Colombia, and Mexico, Nicaragua). However, the role played by African States is crucial (two had the initiative to institute contentious proceedings before the Court) and certain American States have been very active before the Court (eg Nicaragua and Canada). Most importantly, this collective effort to intervene before the Court in cases of common concern is striking when compared to the previous silence of third States. In the first century of the history of the two courts (PCIJ and ICJ), 1920-2020, they had received 18 requests and only 6 States actually intervened either under Article 62 or Article 63 of the Statute.[58]
As noted above, the unprecedented number of interventions in the genocide cases is largely inspired by the goal of protecting a collective interest. They represent a collective effort based in many regards on close coordination of third States in the drafting of the declarations and in requesting to intervene.[59] Third States have clearly tried to ensure, at least before the Court, a common reaction.
The views of third States are not revolutionary when it comes to the definition of the means for the recommended ‘objective’ assessment of genocide. They advocate the use of independent resources, cooperation and the involvement of international institutions. Centralization is not at the core of their positions. Recourse to existing institutions – such as the UN political bodies or the ICJ – is neither seen as compulsory nor as exclusive. At best, institutional assessment is a precondition for unilateral assessment.
This is why ‘collective’ is possibly the best word to describe the ‘common thread’ of third States’ views. They reject a purely unilateral assessment of the commission of erga omnes violations: such an assessment alone cannot justify the decision to take measures in response. On the other hand, the ‘collective’ assessment does not necessarily imply to relinquish State power of unilateral assessment. It calls for a multilateral action but falls short of creating a representative organ proceeding on behalf of the omnes. From this perspective, third States positions are very cautious.
This indirectly confirms their awareness that they are stepping on uncertain ground and that middle-ground solutions between unilateral and collective action are extremely difficult to envisage. What is revolutionary in the end is that States are trying to be creative.

b) Balancing bilateralization and centralization of the response

The underdevelopment of procedural rules relating to the protection of erga omnes obligations was, at least a decade ago, unchallenged: ‘the development of substantive principles for the protection of common interests has so far gone unmatched with the development of procedural rules.’[60] Today, procedural developments are undeniable. They can be elaborated according to two main logics[61] that are not mutually exclusive but rather complementary.
On the one hand, the logic of isolation entails that the entitlement to the protection afforded by erga omnes obligations is extended to all the members of the relevant community, ie either the international community as a whole or treaty members. This substantially leads to the bilateralization of obligations conceived for the protection of the entire collective. The lack of common organs entails the recognition that every member can enforce erga omnes obligations. This logic has typically been applied in the codification of State responsibility (Article 48 ARSIWA) and dispute settlement before the ICJ (locus standi). Some States plead for a corresponding development in intervention under Article 62. The bilateralization of community obligations is the reflection of the horizontal character of the international legal system. It has traditionally been seen as the only way to develop at the international level the protection for common interests.[62]
On the other hand, the logic of inclusion would rather call for collective action. It is precisely the path towards the development of ‘organized’ means for the establishment of the breach of erga omnes obligations. Collective and centralized reaction are possible in certain specific fields of international law, such as the regional protection of human rights. However, the creation of institutional enforcement mechanisms or compulsory jurisdiction at the universal level has been largely seen as utopistic. This is also due to the structure of the international community.
In this regard, the practice analysed above and the positions of third States seem to support the feasibility of at least a certain degree of organization. The practice of intervention itself shows that States can be willing to cooperate or actually act jointly before the Court. The positions discussed above reveal in addition the attempt at devising new forms of coordination, collective action, and recourse to existing institutional means. One is tempted to consider that at least the collective assessment of these erga omnes obligations is less utopistic today.
More generally, the positions of third States advocating objective assessment, independent resources and forms of centralized decision plead for the existence of some limit to the pure logic of isolation in the protection of collective interests. In the framework of institutional judicial settlement of international disputes, it can be perfectly acceptable that they be litigated by all the omnes and legal standing or intervention be generalized. The risk inherent in unilateral action is counterbalanced by the decision being ultimately in the hands of the Court. This was precisely the logic behind the ‘necessity argument’ invoked to plead actio popularis in the South West Africa cases: the ultimate safeguard for the sacred trust was the Court’s decision and the unilateral right of action of each League member was just its triggering mechanism.[63]
However, when action is taken outside institutionalized mechanisms – such as in the case of third-party countermeasures or other unilateral responses – the assessment of its lawful basis should not be left to the sole unilateral, subjective appreciation of every State; some form of organization is necessary to avoid abusive uses of rules meant to protect public interests. What the intervening States seem to be advocating is at least a combination of the two logics and the balancing of the need for some degree of inclusion (knowing that this may prevent any action from being taken at all) and isolation (knowing that abusive responses can defeat the very purpose of public interests’ protection). To find this balance is at the same time very important and very difficult.
The recent flood of intervention requests has shown that devising effective procedures for the protection of collective interests of the international community has become an urgent matter and possibly that States are ready to make a step in the direction of collective action, whatever this may concretely mean. This is essential to ensure that not just law-making but also law-application be open to all members of the international community. This is all the more so for States that ‘rel[y] for [their] security on the global rules-based order with the United Nations Charter and adherence to international law as its core.’[64]

* Full Professor of Public International Law, Sapienza University of Rome.
[1] For a distinction between substantive and procedural law see in general H Thirlway, ‘The significance of procedure in the judicial settlement of international disputes’ in J Gomula, S Wittich, M Stemeseder (eds), Research Handbook on International Procedural Law (Edward Elgar 2024) 21. It is commonly acknowledged that there is an overlap between substance and procedure. See in that regard S Rosenne, The Law and Practice of the International Court 1920-2005 vol III (4^th edn, Brill 2006) 1021; E Lauterpacht, ‘Principles of Procedure in International Litigation’ (2011) 345 Recueil des Cours de l’Académie de Droit International 403. On the interactions between substantive and procedural law see in general JW Salmond, Jurisprudence (4^th edn, Stevens and Haynes 1913) 437-440; L Alexander, ‘Are Procedural Rights Derivative Substantive Rights?’ (1998) 17 L & Philosophy 19; S Talmon, ‘Jus Cogens after Germany v. Italy: Substantive and Procedural Rules Distinguished’ (2012) 25 Leiden J Intl L 979.
[2] ICJ, Application of the Convention on the Prevention and Punishment of the Crime of Genocide (Bosnia and Herzegovina v Yugoslavia) (Preliminary Objections) (Judgment of11July 1996) [1996] ICJ Rep 595 para 31. The Court explicitly recognized the erga omnes character of the obligations of prevention/punishment in ICJ, Application of the Convention on the Prevention and Punishment of the Crime of Genocide (The Gambia v Myanmar) [hereafter ‘Gambia v Myanmar’] (Preliminary Objections) (Judgment of 22 July 2022) [2022] ICJ Rep 477 para 107. Some declarations of intervention are explicit in that regard; see in particular ICJ, Application of the Convention on the Prevention and Punishment of the Crime of Genocide in the Gaza Strip (South Africa v Israel) [hereafter ‘South Africa v Israel’] Declaration of Chile (12 September 2024) para 36 and Declaration of Maldives (1 October 2024) para 38.
[3] For a general analysis of the private and public components of the international legal order see R Kolb, ‘Le droit international comme corps de « droit privé » et de « droit public »’ (2021) 419 Recueil des Cours de l’Académie de Droit International 9.
[4] ICJ, Allegations of Genocide under the Convention on the Prevention and Punishment of the Crime of Genocide (Ukraine v. Russian Federation) [hereafter ‘Ukraine v Russia’] (Admissibility of the Declarations of Intervention) (Order of 5 June 2023) [2023] ICJ Rep 354.
[5] See the Court’s press release 2024/59 of 6 August 2024 <www.icj-cij.org/sites/ default/files/case-related/182/182-20240806-pre-01-00-en.pdf>. The position of the 10 States that have submitted no new communication remains unclear.
[6] ICJ, Gambia v Myanmar (Admissibility of the Declarations of Intervention) (Order of 3 July 2024).
[7] The requests of intervention relating to the South Africa v Israel case (n 1) are published on the Court’s website <www.icj-cij.org/case/192/ intervention>.
[8] ICJ Ukraine v Russia Declaration of Germany (5 September 2022) para 13.
[9] ICJ, Ukraine v Russia Declaration of Luxembourg (13 October 2022) paras 10, 32, 45.
[10] ICJ, Ukraine v Russia Declarations of Lithuania (22 July 2022) para 17; USA (7 September 2022) para 9; Sweden (9 September 2022) paras 11-12; France (13 September 2022) para 8; Romania (13 September 2022) para 21; Italy (15 September 2022) para 33; Poland (15 September 2022) para 33; Denmark (16 September 2022) para 28; Ireland (19 September 2022) para 10; Finland (21 September 2022) paras 10, 32; Estonia (22 September 2022) para 36; Spain (29 September 2022) paras 28-29; Australia (30 September 2022) para 26; Australia (2 August 2024) para 9; Portugal (7 October 2022) paras 11-12; Greece (13 October 2022) paras 14, 39; Austria (12 October 2022) para 14; Croatia (19 October 2022) paras 10, 29; Czech Republic (1 November 2022) para 11; Bulgaria (18 November 2022) paras 11, 26; Malta (24 November 2022) paras 11, 29; Norway (24 November 2022) paras 7, 21; Belgique (6 December 2022) paras 9, 43; Canada and the Netherlands (7 December 2022) paras 11-12; Slovakia (7 December 2022) para 16; Slovenia (7 December 2022) para 28; Cyprus (13 XII 2022) para 10; Liechtenstein (15 December 2022) paras 9, 23. Similar statements in the ICJ, Gambia v Myanmar case can be found in the Joint declaration of Canada, Denmark, Finland, the Netherlands, Germany and the UK (15 November 2023) para 9; and the Declaration of Maldives (15 November 2023) para 7. Similar statements have been put forward by third States in the ICJ, South Africa v Israel case: see the Declarations of Colombia (5 April 2024) paras 16-18; Libya (10 May 2024) p. 3; Mexico (24 May 2024) paras 10-15; Spain (28 June 2024) paras 14, 41; Chile (12 November 2024) paras 20, 72; Maldives (1 October 2024) para 5 and 33; and Bolivia (8 October 2024) para 25.
[11] ICJ, Questions relating to the Obligation to Prosecute or Extradite (Belgium v. Senegal) (Judgment) [2012] ICJ Rep 422 para 69 (‘The common interest in compliance with the relevant obligations under the Convention against Torture implies the entitlement of each State party to the Convention to make a claim concerning the cessation of an alleged breach by another State party. … any State party to the Convention may invoke the responsibility of another State party with a view to ascertaining the alleged failure to comply with its obligations erga omnes partes’.).
[12] ICJ, Gambia v Myanmar (n 1) para 108. For example, this position is put forward and this precedent is recalled by Maldives in their Declaration of intervention in the South Africa v Israel case (1 October 2024 para 34).
[13] On the use of these expressions see B Bonafé, ‘La participation des tiers aux procédures devant la Cour internationale de Justice’ (2023) 69 Annuaire français de droit international (forthcoming).
[14] ICJ, Ukraine v Russia (n 2) para 27.
[15] ibid para 95.
[16] ICJ, Ukraine v Russia Declaration of Latvia (26 July 2024) para 14; Declaration of Germany (5 November 2022) para 14; Italy (15 November 2022) para 16; Slovakia (7 December 2022) para 17. See also ICJ South Africa v Israel Declaration of Maldives (1 October 2024) para 6; whereas Bolivia (8 October 2024) para 27 considers that requesting to intervene is part of its ‘responsibility to condemn the crime of genocide’.
[17] ICJ, Ukraine v Russia Declaration of Colombia (5 April 2024) paras 19 and 186.
[18] ICJ, Ukraine v Russia Declaration of New Zealand (30 July 2024) para 27.
[19] ICJ, Ukraine v Russia Declaration of Sweden (9 September 2022) para 16. That suggestion is repeated in the letter accompanying the updated declaration of intervention that Sweden has submitted on 31 July 2024. For similar proposals, made with very similar formulations, see ICJ, Ukraine v Russia Declarations of Germany (5 September 2022) para 19; Poland (15 September 2024) para 12; Denmark (16 September 2022) para 14 (reiterated in the declaration of 2 August 2024 para 14); joint Austria Czechia Finland Slovenia (2 August 2024) para 11; Estonia (22 September 2022) para 21; Spain (29 September 2022) para 15; Luxembourg (13 October 2022) para 17; Greece (13 October 2022) para 19; Austria (12 October 2022) para 8; Croatia (19 October 2022) para 15; Czech Republic (1 November 2022) para 16; Bulgaria (18 November 2022) para 15 (reiterated in the letter accompanying the new declaration of intervention 2 August 2024 para 17); Malta (24 November 2022) para 16; Norway (24 November 2022) para 11; Slovakia (7 December 2022) para 19; Slovenia (7 December 2022) para 15; Cyprus (13 December 2022) para 12.
[20] ICJ, Application of the Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment (Canada and the Netherlands v Syrian Arab Republic) (Joint application) (8 June 2023) <www.icj-cij.org/case/188>.
[21] ICJ, Ukraine v Russia, CR 2023/15 of 20 September 2023.
[22] ICJ, Ukraine v Russia (n 2) para 88.
[23] ICJ, South Africa v Israel Requests of Nicaragua (23 January 2024) paras 18-19, and of Palestine (3 June 2024) paras 24-25; ICJ, Ukraine v Russia, Request of Poland (23 July 2024) paras 10-13.
[24] ICJ, Ukraine v Russia Declaration of Poland (23 July 2024) para 14. The way in which the may-be-affected condition of the legal interest requirement is demonstrated is not entirely convincing because it is conflated with the general interest requirement of Article 63 (see in particular paras 37, 39 and 42). See also ICJ, South Africa v Israel, Declaration of Nicaragua (23 January 2024) para 16; and Declaration of Palestine (3 June 2024) paras 26 and 31.
[25] G Gaja, ‘The Protection of General Interests in the International Community’ (2013) 364 Recueil des Cours de l’Académie de Droit International 119.
[26] B Bonafé, La protezione del terzo davanti alla Corte internazionale di giustizia (Editoriale scientifica 2014) 199-205.
[27] B McGarry, ‘Obligations Erga Omnes (Partes) and the Participation of Third States in Inter-State Litigation’ 22 (2023) The Law & Practice Intl Courts and Tribunals 273.
[28] ICJ, South Africa v Israel Declaration of Nicaragua (23 January 2024) para 17. See also paras 11, 15, 16.
[29] ICJ, Allegations of genocide under the Convention on the Prevention and Punishment of the Crime of Genocide (Ukraine v Russian Federation: 32 States intervening) (Preliminary objections) (Judgment of 2 February 2024) para 149.
[30] ICJ, Ukraine v Russia Declaration of the UK (5 August 2022) para 58 (emphasis added).
[31] ICJ, Ukraine v Russia Declaration of Italy (15 September 2022) para 47 (‘a State cannot invoke the “undertak[ing] to prevent” genocide in Article I of the Convention as a justification for its conduct fi it has not carried out an objective and documented assessment of the occurrence or the risk of occurrence of genocide’, emphasis added).
[32] ICJ, Ukraine v Russia Declaration of Belgium (6 December 2022) para 24 (‘une Partie contractante ne saurait invoquer l’article I de la Convention pour rendre licite un comportement qui serait autrement illégal en droit international si elle n’a pas établi, sur une base objective et à la suite d’une évaluation de bonne foi de tous les éléments de preuve substantiels provenant de sources indépendantes, qu’un génocide est en train de se produire ou qu’il existe un risque sérieux qu’un génocide se produise’, emphasis added).
[33] ICJ, Ukraine v Russia Declaration of Luxembourg (31 July 2024) para 20 (emphasis added).
[34] ICJ, South Africa v Israel Declaration of Chile (12 September 2024) paras 44-45 (‘After the issuance of the provisional measures, and considering the Court’s finding that the right of the Palestinians in Gaza to be protected from acts of genocide was plausible, Israel cannot claim that it was not aware of the existence of this risk.’).
[35] ICJ, Ukraine v Russia Declaration of New Zealand (28 July 2022) paras 32-33, emphasis added. The Declaration of Italy in the same case also contains a rejection of the possibility that genocide be established subjectively (15 September 2022) para 52.
[36] ICJ, Ukraine v Russia Declaration of Portugal (7 October 2022) para 36.
[37] ICJ, Ukraine v Russia Declaration of Lithuania (24 July 2024) para 11, emphasis added.
[38] ICJ, Ukraine v Russia Declaration of Sweden (9 September 2022) para 46, emphasis added. The same statement is included in the Declaration submitted by Sweden on 31 July 2024 para 8.
[39] See ICJ, Ukraine v Russia Declarations of Denmark (2 August 2024) para 20; Austria, Czech Republic, Finland and Slovenia (2 August 2024) para 38; Estonia (2 August 2024) para 21; Spain (2 August 2024) para 21; Luxembourg (31 July 2024) para 12; Bulgaria (2 August 2024) para 23. See also ICJ, South Africa v Israel Declarations of Colombia (5 April 2024) para 35; Spain (28 June 2024) para 30; Turkey (7 August 2024) para 39.
[40] ICJ, Ukraine v Russia Joint declaration of Canada, Denmark, Finland, the Netherlands, Germany and the UK (15 November 2023) para 76.
[41] ICJ, Ukraine v Russia Declaration of Sweden (9 September 2022) para 54 (emphasis added).
[42] The Court clarified the distinct areas of application of these two provisions in Application of the Convention on the Prevention and Punishment of the Crime of Genocide (The Gambia v Myanmar) (Preliminary Objections) (Judgment of 22 July 2022) [2022] ICJ Rep 477 paras 88-90.
[43] ICJ, Ukraine v Russia Declarations of the UK (5 August 2022) paras 56-7; France (13 September 2022) para 45; Denmark (16 September 2022) para 34; Luxembourg (13 October 2022) para 43; Canada and the Netherlands (7 December 2022) para 35.
[44] ICJ, Ukraine v Russia Declarations of New Zealand (28 July 2022) para 30; Romania (13 September 2022) para 44; Norway (24 November 2022) para 29; Canada and the Netherlands (7 December 2022) para 14.
[45] ICJ, Ukraine v Russia Declaration of Portugal (7 October 2022) para 41. See also ICJ, Ukraine v Russia Declarations of Sweden (9 September 2022) para 47; Belgium (6 December 2022) paras 27-28.
[46] See ICJ, Ukraine v Russia Declarations of Sweden (9 September 2022) para 53; Estonia (22 September 2022) para 50; Czech Republic (1 November 2022) para 29.
[47] See ICJ, Ukraine v Russia Declarations of France (13 September 2022) paras 46-47; Denmark (16 September 2022) para 34 and 42; Luxembourg (13 October 2022) para 43; Norway (24 November 2022) para 29.
[48] ICJ, Ukraine v Russia Declaration of the Czech Republic (1 November 2022) para 29, emphasis added.
[49] ICJ, Ukraine v Russia Declaration of Estonia (22 September 2022) para 51.
[50] ICJ, Ukraine v Russia Declaration of Poland (15 September 2022) para 30.
[51] Art 31(3)(b) of the Vienna Convention allows treaty interpretation in light of subsequent practice, provided that those two requirements are met. The Convention did not codify the possibility of a modification of the treaty in light of subsequent practice. See in that regard the ILC ‘Draft conclusions on subsequent agreements and subsequent practice in relation to the interpretation of treaties’ (2018) II/2 YB ILC.
[52] J Combacau, ‘Le droit international : bric-à-brac ou système ?’ (1986) 31 Archives de philosophie du droit 97-98 (‘La souveraineté des Etats impose l’établissement d’un ordre dans lequel ce sont eux qui confrontent des faits légalement pertinents aux règles desquelles ils tirent leur signification juridique : chaque Etat atteste celle-ci pour lui-même, sans plus de pouvoir d’imposer aux autres la signification qu’il attribue subjectivement à un fait que tout autre n’a le pouvoir de lui imposer la sienne ; à moins qu’une autorité extérieure se voie attribuer par eux la fonction de dire objectivement le droit à leur place’).
[53] See for various views in that regard A Cassese (ed), Realizing Utopia. The Future of International Law (OUP 2012).
[54] G Arangio-Ruiz, ‘Seventh Report on State Responsibility’ UN Doc A/CN.4/469 and Add.1 and 2 (1995) II/1 YB ILC 3.
[55] See especially the summary of the debates in the General Assembly sixth committee concerning that proposal, UN Doc A/CN.4/479/Add.1 (1 May 1997) paras 77 ff.
[56] G Arangio-Ruiz (n 48) 20 ff.
[57] ibid 30.
[58] See in general B Bonafé (n 26).
[59] See J McIntyre, ‘Consenting to Be Bound or Co-Operative Condemnation? Article 63 Interventions at the International Court of Justice’ (2022) 29 Australian Intl L J 1.
[60] A Nollkaemper, ‘International Adjudication of Global Public Goods: The Intersection of Substance and Procedure’ (2012) 23 Eur J Intl L 791.
[61] B Bonafé, ‘Adjudicative Bilateralism and Community Interests’ (2012) AJIL Unbound 164.
[62] See in particular B Simma, ‘From bilateralism to community interest in international law’ (1994) 250 Recueil des Cours de l’Académie de Droit International 217; M Benzing, ‘Community Interests in the Procedure of International Courts and Tribunals’ (2006) 5 The Law and Practice of International Courts and Tribunals 377 ff; S Villalpando, ‘The Legal Dimension of the International Community: How Community Interests Are Protected in International Law’ (2010) 21 Eur J Intl L 409-410; B Bonafé, ‘La violation d’obligations envers la communauté internationale dans son ensemble et la compétence juridictionnelle de la Cour internationale de Justice’ in E Cannizzaro (ed), The Present and Future of Jus Cogens (Sapienza Università Editrice 2015) 148 ff.
[63] ICJ South West Africa Cases (Ethiopia v South Africa; Liberia v South Africa) (Preliminary Objections) (Judgment of 21 December 1962) [1962] ICJ Rep 28-29; South West Africa (Second Phase) (Judgment of 18 July 1966) [1966] ICJ Rep 6 paras 80-98. When the Court finally accepted to generalize locus standi in relation to erga omnes obligations’ litigation, the same logic justified individual action being put at the service of the collective interest. See ICJ Questions relating to the Obligation to Prosecute or Extradite (Belgium v Senegal) (Judgment 20 July 2012) [2012] ICJ Rep 422 para 69.
[64] ICJ Ukraine v Russia Declaration of Cyprus (13 December 2022) para 11.

1. Introductory remarks

Malicious cyber operations[1] by States and non-state actors are on the rise. According to a recent European Union Agency for Cybersecurity (ENISA) threat landscape report, between July 2022 and June 2023 there was ‘a significant increase in both the variety and quantity of cyberattacks and their consequences’.[2] The report identifies ransomware, malware, social engineering, data threats, denial of service, internet threats, information manipulation and interference, supply chain attacks as the most important threats.[3] According to the report, there was 31,32% increase in ransomware attacks since 2022; 21,4% increase in distributed denial-of-service (DDoS) attacks, and 20,9% increase in data attacks. Targets of such attacks have been the public sector, individuals, the health sector, digital infrastructure, digital service providers, education, media, defence, transport, as well as financial services.[4]
Over the years, cyber attacks have become professional, technically advanced, and are often part of other cyber or physical attacks.[5] There has also been a diversification of Advanced Persistent Threats (APTs) which include non-state actors or States working independently or in collaboration with each other.[6]
The harm cyber attacks can cause to States, individuals, society, economy, or institutions is significant. To give an example, according to the World Economic Forum, the global cost of cybercrime is forecast to jump to US$23,84 trillion by 2027, up from US$8,44 trillion in 2022.[7]
Beyond this, cyber attacks can have pernicious effects on international relations by fracturing inter-state relations, causing suspicion and mistrust, perpetuating disputes or triggering conflicts. Cyber attacks can also undermine the international rule of law.
For this reason, the United Nations Secretary-General proposed the establishment of an independent multilateral accountability mechanism for malicious use of cyberspace by States to reduce incentives for such conduct; enhance compliance with agreed norms and principles of responsible State behaviour; and strengthen criminal justice capacity to investigate, prosecute and adjudicate cyberactivity by terrorist actors against such infrastructure.[8]
One issue that is critical to achieving these aims is attribution. Attribution refers to the process of identifying the source of a malicious cyber attack and is critical for adopting legal, technical, and other measures to prevent and suppress cyber attacks, deter further attacks[9] and hold the authors of cyber attacks and their sponsors responsible. By doing so, attribution may induce compliance with international law which applies to cyberspace and regulates cyber conduct.[10] It may also induce compliance with the norms of responsible State behaviour in cyberspace.[11] The term ‘norm’ is used in the United Nations Group of Governmental Experts (UN GGE) and Open-Ended Working Group (OEWG) reports as an umbrella term to include capacity-building norms, confidence-building norms, non-binding norms but also binding norms. Non-binding norms reflect expectations of proper State behaviour[12] whereas binding norms include international law rules but also norms grafted on specific international law rules such as those on the prevention of transboundary harm or due diligence.[13] The aim of these four categories of norms is to prevent threats and maintain international peace and security.
Although the importance of attribution is undisputed, attribution in cyberspace is quite complex due to the technical dimension of cyber attacks but also due to anonymity, the potential for impersonation, the multi-layer nature of cyber attacks, the ability to hijack devices or use private and secure networks to commit attacks. This means that attributing cyber attacks is more demanding than attributing physical attacks and may take more time to collect and assess information before attributing.
The remainder of this article discusses in Section 2 the types of cyber attribution whereas in Section 3 it critiques proposals for the establishment of an attribution agency to deal with the problem of attribution.[14] It contends that such proposals underestimate the legal, political, and technical challenges posed by cyber attribution and misjudge States’ interests and preferences.
Before proceeding, two clarifications are in order. The article focuses on inter-state attribution that is attribution of malicious cyber operations by one State to another State and not attribution performed by private security companies unless of course it is done on behalf of States in which case they become State attributions. Second, legal attribution refers to international law attribution where a State attributes a malicious cyber operation by using international law criteria.

2. Types of cyber attack attribution

In previous publications I argued that cyber attribution has a technical, political, and legal dimension.[15] Technical attribution is about identifying the source of a malicious operation by investigating the tactics, techniques, and procedures of the attack on the basis of technical evidence such as Internet Protocol (IP) addresses, websites, activity logs, malware, and so on.[16] Political attribution is about ascribing responsibility for a malicious cyber operation to a State, a State entity or an entity sponsored by a State by assessing technical evidence, intelligence information, and by making broader political judgments.[17] Legal attribution is about the identification of a State as the author of a cyber attack on the basis of legal criteria of attribution, particularly those found in the law of State responsibility.[18] Whereas technical attribution informs all other forms of attribution, political or legal attribution may apply their own criteria which do not always coincide.
Technical attribution is performed by private security companies[19] but also States, often in cooperation with the private sector. However, States are not always keen to publish their evidence or attribution methodology. Political attribution is mainly performed by States either individually or collectively.[20] Following such attributions, States may take measures against the culprit State, organisation or individual. States however stop short of performing legal attribution because of the legal consequences attribution or misattribution may trigger and because they use attribution to pursue multiple aims.[21] Even if they use the language of international law, albeit very infrequently, they do not pursue the matter further.[22]
That having been said, because attribution, as was said, is important and is a precondition for other actions, attribution claims are always contested and by contesting attribution, any related political or legal claim or responses are also contested.
To give an example, Albania[23] but also other States[24] attributed the 2022 attacks on Albanian public services[25] to Iran. According to Albania, ‘a thorough in-depth investigation conducted in cooperation with specialized partner agencies on cyberterrorism has now confirmed, beyond any doubt, that the cyberattack was a State-sponsored aggression carried out by four groups, orchestrated and sponsored by the Islamic Republic of Iran’.[26] Albania went on to sever diplomatic relations and expel Iranian diplomats. Iran rejected ‘any kind of unwarranted attribution’ characterising the accusations ‘unsubstantiated’ and ‘unfounded’ and added ‘given the nature and technical characteristics of cyberspace and the challenges of attribution in the Information and Communication Technology (ICT) environment, Iran warns of the negative consequences of falsified and forged attribution to the States’.[27]
It transpires then that attribution can add another layer to an existing conflict or become itself the object of conflict. Thus, attribution can cause disorder instead of maintaining a modicum of order.

3. Proposals for a cyber attribution agency and their challenges

In order to prevent attribution conflicts but also recognising the importance of attribution, certain proposals have been put forward for the establishment of agencies to perform attribution. These proposals envisage intergovernmental, private, or multistakeholder agencies.[28] The aim behind the proposals is to introduce a politically neutral, transparent and as far as possible objective process of attribution to assist States’ decision-making capacity, particularly technologically underdeveloped States. I am quite sceptical about the ability of such mechanisms to rise to the challenge, and I will explain the reasons for my scepticism in addition to those mentioned by François Delerue in his article.[29]
First, States view attribution as a sovereign prerogative.[30] They attribute when it serves their interests and at a time of their choice.[31] They also perform attribution themselves or collaborate with those States or agencies they wish in order to make attributions. States also chose whether to make public attributions.[32] In other words, their attribution practices indicate that States want to maintain their decision-making freedom and discretion whether, when and how to attribute.
Second, attribution requires evidence, but States are reluctant to disclose the evidence supporting their attributions.[33] According to the 2015 UN GGE Report, accusations of wrongful acts should be substantiated[34] but States maintain that they are not under a legal obligation to disclose their evidence[35] and their practice confirms this view. States do not also reveal their attribution methodology or technology or their intelligence sources. Doing so will compromise their ability to attribute and will strengthen their rivals who can use this information to improve their cyber attack capabilities and their capacity to remain undetected.
Third, the process of attribution is invasive in that it involves searches on State networks. Because of the interconnectivity of networks and cyber infrastructure, such searches may inadvertently access a State’s sensitive information or critical networks. This can also be done deliberately if those conducting investigations abuse their position. If there is any such intrusion, accusations and recriminations will follow, and as a consequence such an attribution agency will expand rather than contain any underlying dispute. Disputes can also arise with third States if, due to the interconnected nature of cyberspace, accidental or bad faith searches are carried out in their networks.
Fourth, because of the invasive nature of the investigations, the consent of the States involved will be necessary either on an ad hoc basis or a priori when they become members of such an agency. However, States can calibrate their consent, for example by excluding certain networks. States may also obstruct investigations even if they give their full consent. This may affect the work of the agency and the reliability and credibility of its investigations and findings. Moreover, due to the interconnected nature of cyberspace, the consent of third States may be needed. However, which third States are involved may not be clear from the beginning, neither searches into third States’ infrastructure can become immediately perceptible.
Fifth, in order to maintain the transparency of its process, such an agency may need to publish certain information, for example information about vulnerabilities, resources, etc, which may disadvantage the States concerned. It may also publish information about the agency’s attribution methodology and standards which can be used by States or non-state actors to evade attribution. Such an agency may also publish technical information about any detected malware which can be re-designed and reused by malicious actors.
Sixth, the services of such attribution agencies may be useful to less technologically developed States rather than developed ones. As François Delerue argues, this is a worthy reason for having such an agency.[36] There are many problems with such a view, however. Technologically developed States may not become parties to such an agency or, even if they become parties, they may not require its services because they can perform attribution themselves or together with their trusted partners. Attribution will thus become two-speed or multi-speed, contrary to the aims behind the establishment of such an agency. Even if technologically developed States use the services of such an agency, they may be able to obstruct investigations or hide information. Also, depending on the status and powers of such an agency, it may be reluctant or unable to investigate more powerful States. Finally, in order to provide good quality attribution services to technologically underdeveloped States, such an agency will require advanced resources which can only be provided by technologically advanced States. Lacking such resources, it will not be able to provide good quality services even to technologically weak States but may not also be able to investigate technologically advanced cyber attacks.
Seventh, in order to make credible attributions such agencies should apply a common and well-defined standard of proof and a transparent methodology.[37] The standard of proof refers to the degree of conviction the agency will need to have when making an attribution determination but also with the sufficiency of the evidence supporting the determination whereas the methodology refers to the type of evidence it will take into consideration and the process of their analysis. However, neither technical nor political or legal attribution as practiced today have established any common standard of proof and, moreover, the standards of proof may differ between the three types of attribution. As a matter of fact, standards peregrinate between convincing, high probability, sufficient evidence, and incontrovertible evidence.[38] Neither does international law provide any common standard of proof but the standard changes depending on the nature of the case.[39] It can be said that such an agency can set its own standard of proof and thus streamline attributions. There are however certain problems in this regard which may be proved counterproductive. States, but other organisations including private companies, will continue to do attributions using their own (undisclosed) standard of proof. As a result, there will be competing claims based on different standards, and the attribution findings of such an agency will be challenged. If that is to happen, an adjudicatory body could arbitrate, but this is a rather remote possibility. Even if there is such an adjudicatory mechanism, for example a court, it may apply different standards than the attribution agency.
Eighth, if such an agency is established, it will most probably conduct technical attribution. Political attribution is too close to State interests and too reliant on intelligence information that such an agency will not be able to acquire. If it does perform technical attribution, it will need technical resources and know-how. These can be acquired by the private sector or States and indeed technologically advanced States. The question that arises is whether States or private companies will be willing to share their knowledge and resources with such an agency which will compete with them. Moreover, even if technical attribution in principle appears to be more neutral, the reverse engineering involved will reveal the tools, techniques or networks used for the attack which may involve many territories or jurisdictions. This may expand the scale of accusations and recriminations as well as the scale of any subsequent political or legal actions. For example, if the malware has traveled through the territories of two or more States, what prevents the victim State from accusing them for complicity in the cyber attack and respond in kind? Technical attribution can thus become the vector of more disputes.
Nineth, if such an agency is to perform legal attribution, a major challenge it will face is that the attribution criteria may be proved inadequate for cyber attribution but also their content is often disputed. François Delerue mentioned the legal debates surrounding the ‘effective’ and ‘overall’ control attribution standards but there are also debates regarding ‘direction’, ‘instructions’, and also the definition of organs and agencies.[40] The application of the attribution criteria also depends on facts but as explained above, such an agency may face difficulties in collecting evidence and thus its findings may be disputed. It can be said that such an agency may be able to streamline and clarify the attribution criteria, but this can give rise to a number of systemic challenges. One such challenge concerns the place of the referent agency’s interpretations of the law in the international law-making process. To explain, the question is whether its interpretations will contribute to the formation of customary international law in general or be à la carte to cyber attribution.[41] Another question is what will happen if its interpretations are disputed by States, courts, or international organisations. Still another and more fundamental question is whether such an agency and in particular a private or an agency which is not purely intergovernmental can participate in the making of international law.[42]
Tenth, the above criticisms about the role and place of such agencies in the international law process also apply to those who view the existence of an attribution agency as beneficial to business because it can deconflict the relations between business and its clients. To explain, since cyber technology and infrastructure are mainly controlled by the private sector and States are their clients, attribution by a private company (for example Microsoft) of a malicious cyber attack perpetrated by a State to another State when both States are its clients will sour relations between the company and States. An international attribution agency will in principle prevent such conflicts by removing companies from performing attribution. I do not however believe that conflicts will be avoided. Such an attribution agency may request and indeed rely on information provided by said company to make its determinations. States may also request said company to investigate before or after an attribution determination has been made. Conflicts will thus simmer or be elevated to inter-state conflicts by using the company to support or deny claims.
Eleventh, there is the question of the accountability of such an agency. Its findings, whether binding or non-binding, may be relied upon by States to inform their actions. If they are proven wrong, would that agency bear some of the responsibility for wrongful actions? Will it be deemed to be complicit in the wrongful act?
François Delerue proposes the creation of an attribution mechanism on the footprints of the Permanent Court of Arbitration (PCA) without however arbitration functions. Its function will be to develop rules and standards for attribution but also capacity building. This is a very interesting proposal which deserves serious consideration. Having a rostrum of legal, technical, and political experts available to States would of course be beneficial. The problem however is that for reasons already explained and due to the sensitivity of attribution investigations, States may be reluctant to use these experts. Although according to François Delerue’s proposal, this agency will provide its services at the pre-dispute stage, it may be catalytic in the formation of a dispute in which case nothing prevents any of the parties from using it or make reference to its findings at the stage of arbitration or dispute resolution which is what the PCA does.
Regarding the development of attribution rules and standards by such an agency, this would be an important contribution if they reflect the needs and characteristics of cyber attribution. The development of such rules can also streamline attributions but there are some problems in this regard. First, the attribution practices of States remain obscure and, as was said, States do not provide information on the evidence they use, the standards of proof, and their methodology. Moreover, technical, political, and legal attribution feed into each other. Any rules produced by such an agency may thus appear from thin air and remain academic. Most critically though is the question raised earlier about the role and place of such agency in the law-making process.
Even if such rules and standards were to develop, whether they can be used by States to settle any ensuing dispute is a different question. The PCA rules and standards for example can be used when States decide to use any of the existing dispute settlement mechanisms to settle their disputes and with their consent. In order to use the rules and standards developed by that agency in a similar way, there needs to be a dispute, but the threshold for a legal dispute to arise is quite high.[43] Moreover, States are reluctant to pronounce that they are in a cyber dispute with another State in regards to attribution or other cyber matters because they use attribution as a political tool to exert pressure on other States but also because the legal framework applying to cyberspace is still unsettled. Finally, such rules and standards can be used only with the consent of the parties to the dispute but since membership to that agency will be limited, consent may not be secured.

4. Conclusions

In the preceding sections, I explained why I remain sceptical about the creation and effectiveness of an international cyber attribution agency. If however the legal framework that applies to cyberspace densifies and matures and States start using existing dispute settlement mechanisms to resolve their cyber differences, [44] an attribution agency can have an important role to play. It can offer a procedural framework to be used by States if they otherwise disagree on what procedures to follow; provide a rostrum of experts for States to use; and provide administrative and perhaps financial support. To the extent that its services will be flexible and rely on State consent, States can use it with confidence. In the meantime and until that moment arrives, some order can be forged if States centralise their attribution mechanisms and lay down their modes of operation, in particular their mandate, membership, methodology, resources, and review processes. States can also do the same at a bilateral or multilateral level.[45] Even if attributions remain multi-vocal, the attribution process will become more stable. Also, if States wish to settle any cyber attribution dispute peacefully, they can establish a fact-finding commission[46] or engage with any other dispute settlement method on a bilateral and consensual basis. In short, international law and relations can still offer solutions to the cyber attribution problem.

* Professor of International Law, University of Sheffield (Nicholas.Tsagourias@sheffield.ac.uk). I would like to thank Professor Emanuele Cimiotta and the anonymous reviewer for their valuable comments and suggestions as well as Professor François Delerue for sharing his ideas with me. The usual disclaimer applies.
[1] Malicious cyber operations and cyber attacks are used in this article interchangeably.
[2] European Union Agency for Cybersecurity, ‘Threat Landscape 2023’ (19 October 2023) 4 ˂https://www.enisa.europa.eu/publications/enisa-threat-landscape-2023˃.
[3] ibid section 1.1.
[4] ibid section 1.4.
[5] For example, Russia launched cyber attacks on the Ukrainian energy sector prior but also during the war. See European Parliament, ‘Russia’s war on Ukraine: Timeline of cyber-attacks’ (June 2022) ˂https://www.europarl.europa.eu/RegData/etudes/BRIE/2022/733549/EPRS_BRI(2022)733549_EN.pdf˃; European Parliament, ‘The role of cyber in the Russian war against Ukraine: Its impact and the consequences for the future of armed conflict’ (September 2023) ˂https://www.europarl.europa.eu/RegData/etudes/BRIE/2023/702594/EXPO_BRI(2023)702594_EN.pdf˃.
[6] States may use proxies which offer them ‘plausible deniability’. See T Maurer, Cyber Mercenaries: The State, Hackers, and Power (CUP 2018); FJ Egloff, Semi-State Actors in Cybersecurity (OUP 2021). For a critique of plausible deniability see JK Canfil, ‘The Illogic of Plausible Deniability: Why Proxy Conflict in Cyberspace May No Longer Pay’ (2022) 8 Journal of Cybersecurity 1-16.
[7] E Charlton, ‘2023 was a big year for cybercrime – here’s how we can make our systems safer’ World Economic Forum (10 January 2024) ˂https://www.weforum.org/agenda/2024/01/cybersecurity-cybercrime-system-safety/˃.
[8] United Nations, ‘Our Common Agenda Policy Brief 9: A New Agenda for Peace’ (July 2023) 27 ˂https://www.un.org/sites/un2.un.org/files/our-common-agenda-policy-brief-new-agenda-for-peace-en.pdf˃.
[9] JS Nye, ‘Deterrence and Dissuasion in Cyberspace’ (2016) 41 Intl Security 44-71, 46; S Baliga, E Bueno de Mesquita, A Wolitzky, ‘Deterrence with imperfect attribution’ (2020) 114 Am Polit Sci Rev 1155-1178; N Kostyuk, ‘Deterrence in the cyber realm: public versus private cyber capacity’ (2021) 65 Intl Stud Quart 2021, 1151-1162.
[10] UNGA, ‘Report of the Group of Governmental Experts on Developments in the Field of Information and Telecommunications in the Context of International Security’ UN Doc A/70/174 (22 July 2015) section VI; UNGA, ‘Report of the Group of Governmental Experts on Advancing Responsible State Behaviour in Cyberspace in the Context of International Security’ UN Doc A/76/135 (14 July 2021) s IV; UNGA, ‘Final Substantive Report of the Open-Ended Working Group on Developments in the Field of Information and Telecommunications in the Context of International Security’ UN Doc A/AC.290/2021/CRP.2 (10 March 2021) paras 34-37. Regarding States, see for example Italy, ‘Italian Position Paper on “International Law and Cyberspace”’ (2021) 3.
[11] UN Doc A/70/174 (n 10) section III; UN Doc A/76/135 (n 10) section III; UN Doc A/AC.290/2021/CRP.2 (n 10) paras 24-29.
[12] M Finnemore, ‘Cybersecurity and the concept of norms’ (30 November 2017) Carnegie Endowment for International Peace 1-2 ˂https://carnegie-production-assets.s3.amazonaws.com/static/files/Finnemore_web_final.pdf˃.
[13] UN Doc A/76/135 (n 10) section III, norm 13(c).
[14] See F Delerue, ‘Reflections on the Opportunity of an International Attribution and Accountability Mechanism for Cyber Operations’ in this Zoom in, supra 5-21.
[15] See N Tsagourias, ‘Cyber Attacks, Self-Defence and the Problem of Attribution’ (2012) 17 Journal of Conflict and Security Law 229-244; N Tsagourias, M Farrell, ‘Cyber Attribution: Technical and Legal Approaches and Challenges’ (2020) 31 European J Intl L 941-967. For States’ view, see Italy (n 10) 5: ‘Italy is aware that attribution entails technical, legal and political considerations’. According to Costa Rica, ‘[l]egal attribution must be distinguished from the processes of technical and political attribution. Technical attribution comprises a factual investigation into the source of a cyber operation. This often requires technical expertise and is fraught with challenges given cyberspace’s decentralized nature and the widespread use of spoofing techniques and ‘false flags’. Political attribution is the discretionary decision of a State to single out a certain entity, whether a State or a non-State actor, as the author of a certain cyber operation. While international law neither imposes a specific evidentiary threshold for legal attribution nor requires the publication of any evidence for this purpose, States should consider all relevant information when legally attributing cyber operations to another State, publicly or not’, Costa Rica, ‘Costa Rica’s Position on the Application of International Law in Cyberspace’ (21 July 2023) 3-4, para 12. According to Ireland, ‘[t]here is a distinction to be drawn between legal attribution and political attribution. The former is a strictly legal exercise grounded in international law and the rules of state responsibility, while the latter is likely to be informed by political and technical assessments, often heavily based on intelligence reports. It is important to maintain clarity between the different frameworks in which legal attribution and political attribution are to be considered’. Ireland, ‘Position Paper on the Application of International Law in Cyberspace’ (6 July 2023) 5-6, para 24.
[16] See Defense Advanced Research Projects Agency, ‘Enhanced Attribution’ ˂https://www.darpa.mil/program/enhanced-attribution˃; Office of the Director of National Intelligence, ‘A Guide to Cyber Attribution’ (14 September 2018), ˂https://www.dni.gov/files/CTIIC/documents/ODNI_A_Guide_to_Cyber_Attribution.pdf˃; Tsagourias, Farrell (n 15) 947-951; A Kazakova, I Kwiatkowski, J Ryng, K Chan, ‘Unpacking Technical attribution and challenges for ensuring stability in cyberspace’ (20^th June 2022) Security List ˂https://securelist.com/unpacking-technical-attribution/106791/˃.
[17] KE Eichensehr, ‘The Law and Politics of Cyberattack Attribution’ (2020) 67 UCLA Law Review 520-598.
[18] International Law Commission, ‘Report of the International Law Commission on the Work of Its Fifty-third Session’ (23 April-1 June and 2 July-10 August 2001) UN Doc A/56/10 (2001) II/2 YB ILC.
[19] Mandiant, ‘APT1 Exposing One of China’s Cyber Espionage Units’ (30 December 2021) ˂https://www.mandiant.com/resources/reports/apt1-exposing-one-chinas-cyber-espionage-units˃; Microsoft, ‘Special Report: Ukraine, An Overview of Russia’s Cyber Attack Activity in Ukraine’ (27 April 2022) ˂https://query.prod.cms.rt.microsoft.com/cms/api/am/binary/RE4Vwwd˃. Private security companies may also use broader information if a State or a state-sponsored actor is involved. See B Lenaerts-Bergmans, ‘Naming Adversaries and Why It Matters to Your Security Team’ (31 May 2022) Crowdstrike ˂https://www.crowdstrike.com/blog/naming-adversaries-and-why-it-matters-to-security-teams/˃; K Vanderlee, ‘DebUNCing Attribution: How Mandiant Tracks Uncategorized Threat Actors’ (17 December 2020) Mandiant ˂https://cloud.google.com/blog/topics/threat-intelligence/how-mandiant-tracks-uncategorized-threat-actors/˃.
[20] For example, the WannaCry Malware attack in 2017 which was attributed to North Korea. See Australia, Minister for Foreign Affairs, ‘Press Release’ (20 December 2017) ˂https://www.dfat.gov.au/sites/default/files/australia-attributes-wannacry-ransomware-to-north-korea.pdf˃; United States, The White House, ‘Press Briefing on the Attribution of the WannaCry Malware Attack to North Korea’ (19 December 2017) ˂https://trumpwhitehouse.archives.gov/briefings-statements/press-briefing-on-the-attribution-of-the-wannacry-malware-attack-to-north-korea-121917/˃; New Zealand, National Cyber Security Centre, ‘New Zealand concerned at North Korean Cyber Activity’ (20 December 2017) ˂https://www.ncsc.govt.nz/newsroom/new-zealand-concerned-at-north-korean-cyber-activity/˃; Japan, Ministry of Foreign Affairs, ‘The U.S. Statement on North Korea’s Cyberattacks (Statement by Press Secretary Norio Maruyama)’ (20 December 2017) ˂https://www.mofa.go.jp/press/release/press4e_001850.html˃; United Kingdom, Foreign and Commonwealth Office, ‘Foreign Office Minister condemns North Korean actor for WannaCry attacks’ (19 December 2017) ˂https://www.gov.uk/government/news/foreign-office-minister-condemns-north-korean-actor-for-wannacry-attacks˃. See also Microsoft’s support of this coordinated response, B Smith, ‘Microsoft and Facebook disrupt ZINC malware attack to protect customers and the internet from ongoing cyberthreats’ (19 December 2017) Microsoft ˂https://blogs.microsoft.com/on-the-issues/2017/12/19/microsoft-facebook-disrupt-zinc-malware-attack-protect-customers-internet-ongoing-cyberthreats/˃. For the NotPetya cyber-attack which was attributed to Russia see United States, The White House, ‘Statement from the Press Secretary’ (15 February 2018) ˂https://trumpwhitehouse.archives.gov/briefings-statements/statement-press-secretary-25/˃; United Kingdom, Foreign and Commonwealth Office, ‘Foreign Office Minister condemns Russia for NotPetya Attacks’ (15 February 2018) ˂https://www.gov.uk/government/news/foreign-office-minister-condemns-russia-for-notpetya-attacks˃; United Kingdom, National Cyber Security Centre, ‘Russian military “almost certainly” responsible for destructive 2017 cyber-attack’ (14 February 2018) WiredGov ˂https://www.wired-gov.net/wg/news.nsf/articles/Russian+military+almost+certainly+responsible+for+destructive+2017+cyber+attack+16022018091500?open˃. Private security companies can also attribute cyber attacks to States (see for example references included in n 19). The difference however with political attribution as described here is that, even if they use geopolitical factors and cyber threat intelligence information (otherwise called raw intelligence collection), they do not use the same intelligence information as states do and their attribution to a state is mainly based on technical attribution. Cyber threat intelligence has been defined as ‘data that is collected, processed, and analyzed to understand a threat actor’s motives, targets, and attack behaviors’. See K Baker, ‘What is Cyber Threat Intelligence?’ (23 March 2023) Crowdstrike ˂https://www.crowdstrike.com/cybersecurity-101/threat-intelligence/˃. Crowdstrike also uses open-source intelligence (OSINT) such as the dark web, and social media.
[21] M Finnemore, DB Hollis, ‘Beyond Naming and Shaming: Accusations and International Law in Cybersecurity’ (2020) 31 European J Intl L 969-1003.
[22] According to the European Repository of Cyber Incidents (EuRepoC), out of 2,901 attributions, only 1,54% were legal (as from 18 June 2024). See EuRepoC, ‘Attribution Tracker’ ˂https://eurepoc.eu/attribution-tracker/˃.
[23] Albania, ‘Videomessage of Prime Minister Edi Rama’ (7 September 2022) ˂https://www.kryeministria.al/en/newsroom/videomesazh-i-kryeministrit-edi-rama/˃.
[24] United States, The White House, ‘Statement by NSC Spokesperson Adrienne Watson on Iran’s Cyberattack against Albania’ (7 September 2022) ˂https://www.whitehouse.gov/briefing-room/statements-releases/2022/09/07/statement-by-nsc-spokesperson-adrienne-watson-on-irans-cyberattack-against-albania/˃; United Kingdom, Foreign, Commonwealth & Development Office, ‘UK condemns Iran for reckless cyber attack against Albania’ (7 September 2022) ˂https://www.gov.uk/government/news/uk-condemns-iran-for-reckless-cyber-attack-against-albania˃.
[25] United States, Cybersecurity and Infrastructure Security Agency, ‘Iranian State Actors Conduct Cyber Operations Against the Government of Albania’ (23 September 2022) ˂https://www.cisa.gov/news-events/cybersecurity-advisories/aa22-264a˃; Microsoft Threat Intelligence, ‘Microsoft Investigates Iranian Attacks against the Albanian Government’ (8 September 2022) Microsoft Security ˂https://www.microsoft.com/en-us/security/blog/2022/09/08/microsoft-investigates-iranian-attacks-against-the-albanian-government/˃.
[26] ‘Letter dated 7 September 2022 from the Permanent Representative of Albania to the United Nations addressed to the Secretary-General and the President of the Security Council’ UN Doc A/76/943-S/2022/677 (9 September 2022).
[27] ‘Letter dated 10 September 2022 from the Permanent Representative of the Islamic Republic of Iran to the United Nations addressed to the Secretary-General and the President of the Security Council’ UN Doc S/2022/685 (12 September 2022).
[28] See F Delerue, ‘Reflections on the Opportunity of an International Attribution and Accountability Mechanism for Cyber Operations’ in this Zoom in, supra 5-21 at 11-13.
[29] ibid. For a sympathetic view, see Y Shany, MN Schmitt, ‘An international Attribution Mechanism for Hostile Cyber Operations’ (2020) 96 Intl Law Studies 196-222.
[30] According to Czechia, ‘[a] decision to attribute a cyber activity is a national prerogative and remains at the discretion of a sovereign State’. See Ministry of Foreign Affairs of the Czech Republic, ‘Czech Republic – Position paper on the application of international law in cyberspace’ (27 February 2024) 4. See also France, Ministry of Defense, ‘International Law Applied to Operations in Cyberspace’ (9 September 2019) 11; Sweden, ‘Position Paper on the Application of International Law in Cyberspace’ (July 2022) 5; Italy (n 10) 5.
[31] According to Germany, ‘[a]lso, it is within the political discretion of a State to decide on the timing of a public act of attribution’. Federal Government of Germany, ‘On the Application of International Law in Cyberspace – Position Paper’ (March 2021) 10-12.
[32] FJ Egloff, M Smeets, ‘Publicly attributing cyber attacks: a framework’ (2023) 46 Journal of Strategic Studies 502-533; H Lee, ‘Public attribution in the US government: implications for diplomacy and norms in cyberspace’ (2023) 6 Policy Design and Practice 198-216.
[33] Y Ronen, ‘Some Evidentiary Dimensions of Attributing Unlawful Cyber Operations to States’ (2020) Hebrew University of Jerusalem Legal Studies Research Paper Series No 20-11; Eichensehr, ‘The Law and Politics of Cyberattack Attribution’ (n 16).
[34] See United Nations General Assembly, ‘Group of Governmental Experts on Developments in the Field of Information and Telecommunications in the Context of International Security’ UN Doc A/70/174 (22 July 2015) para 28(f).
[35] See BJ Egan, ‘International Law and Stability in Cyberspace’ (2017) 35 Berkeley Journal of Intl Law 169-180, 177; J Wright, ‘Speech at Chatham House Royal Institute for International Affairs: Cyber and International Law in the 21st Century’ (23 May 2018) United Kingdom Government ˂https://www.gov.uk/government/speeches/cyber-and-international-law-in-the-21st-century˃. According to Czechia, ‘States are not obliged to disclose evidence in order to publicly attribute a cyber activity. The disclosure of evidence may be relevant when a legal proceeding is initiated (eg, before the International Court of Justice)’ (n 30) para 58. According to France, ‘international law does not require States to provide the evidence on which the public attribution of a cyberattack is based, though such information helps to legitimise the validity of such attribution’ (n 30) 10-11. Germany ‘agrees that there is no general obligation under international law as it currently stands to publicize a decision on attribution and to provide or to submit for public scrutiny detailed evidence on which an attribution is based. This generally applies also if response measures are taken. Any such publication in a particular case is generally based on political considerations and does not create legal obligations for the State under international law’ (n 31) 10-12. See also Italy (n 10) 5.
[36] See F Delerue, ‘Reflections on the Opportunity of an International Attribution and Accountability Mechanism for Cyber Operations’ in this Zoom in, supra 5-21.
[37] According to Brazil, ‘[t]he technical difficulties in tracing cyber operations and in determining its authorship may lead to additional challenges in attributing an internationally wrongful act to a State. However, these added difficulties must not serve as a justification to lower the bar for determinations on attribution, which must be substantiated’ General Assembly, ‘Official compendium of voluntary national contributions on the subject of how international law applies to the use of information and communications technologies by States’ UN Doc A/76/136 (13 July 2021) 21.
[38] According to Czechia, ‘[a]lthough States are not required to be absolutely certain to attribute a wrongful act, but rather to gather and reasonably assess the information available, a sufficient degree of certainty on attribution of a wrongful act to a particular State needs to be reached’ (n 30) 14 para 57.
[39] See Military and Paramilitary Activities in and against Nicaragua (Nicaragua v United States of America) (Merits: Judgment) [1986] ICJ Rep 14 para 110; Application of the Convention on the Prevention and Punishment of the Crime of Genocide (Bosnia and Herzegovina v Serbia and Montenegro) (Judgment) [2007] ICJ Rep 43 para 209.
[40] Tsagourias, Farrell (n 15); K Mačák, ‘Decoding Article 8 of the International Law Commission’s Articles on State Responsibility: Attribution of Cyber Operations by Non-State Actors’ (2016) 21 Journal of Conflict and Security Law 405-428.
[41] African Union: ‘[s]ubject to the emergence of specific rules of attribution, the African Union affirms that the customary rules on State responsibility, as reflected in the ILC Draft Articles on the Responsibility of States for Internationally Wrongful Acts provide the applicable rules of the attribution to States of conduct undertaken through ICTs in cyberspace’. African Union Peace and Security Council, ‘Common African Position on the Application of International Law to the Use of Information and Communication Technologies in Cyberspace’ (29 January 2024) 10.
[42] N Tsagourias, ‘Digitalization and its Systemic Impact on the Use of Force Regime: Legal Uncertainty and the Replacement of International Law’ (2023) 24 German Law Journal 494-507.
[43] Mavrommatis Palestine Concessions (Judgment, Objection to the Jurisdiction of the Court) [1924] PCIJ Series A no 2, 11; Territorial and Maritime Dispute between Nicaragua and Honduras in the Caribbean Sea (Nicaragua v Honduras) [2007] ICJ Rep 659 para 130. It should be noted however that outside a legal context, the concept of dispute may be broader. See arts 2(3) and 33 of the United Nations Charter.
[44] N Tsagourias, R Buchan, D Franchini (eds), The Peaceful Settlement of Inter-state Cyber Disputes (Hart 2024) (forthcoming).
[45] See F Delerue, ‘Reflections on the Opportunity of an International Attribution and Accountability Mechanism for Cyber Operations’ in this Zoom in, supra 5-21 at 14.
[46] N Tsagourias, F Middleton, ‘Fact-Finding and Cyber Attribution’ in R Buchan, D Franchini, N Tsagourias (eds), The Changing Character of International Dispute Settlement: Challenges and Prospects (CUP 2023) 439-466.

1. Introduction

The Secretary-General of the United Nations recently recommended ‘[e]stablish[ing] an independent multilateral accountability mechanism for malicious use of cyberspace by States to reduce incentives for such conduct’.[1] This is the latest illustration of the repeated questioning and advocacy for the establishment of an international mechanism on cyber operations and international security to address the issues of attribution, responsibility of wrongdoing States and more broadly accountability of different actors, in particular with respect to norms of responsible behaviour.[2] Various proposals have been put forward over the years,[3] but the basic idea remains the same: the creation of an international mechanism – either multilateral, multi-stakeholder, or exclusively involving non-state actors – aimed at conducting investigations into specific cyber incidents – focusing only on alleged state-sponsored cases or taking a broader approach – and identifying the involved actors. Some proposals suggest that the mechanism, once it has identified the wrongdoing actors, should be able to refer the cases to the United Nations Security Council (UNSC) or the International Court of Justice (ICJ), or should be able to impose consequences directly on the wrongdoing actors.
I had the honour to be invited together with Nicholas Tsagourias by Questions of International Law to participate in this Zoom-in dialogue on the question of an international mechanism on the attribution of cyber operations. In this dialogue, he focused on the arguments against such a mechanism, while I took the opposite view. This is an interesting exercise for me, as I am generally skeptical on the naïve and general idea that we should create a ‘cyber international court’; however, I am generally optimistic on the possibility to advance further the implementation of international law in cyberspace, including through specific multilateral and multi-stakeholder processes. For this reason, I have focused my reflections on the possible forms of such a mechanism and what it could achieve.
There are different factors that can explain the interest in an international mechanism on cyber operations and international security. Unlike other forms of state-sponsored unlawful or unfriendly acts such as the violation of another State’s airspace by a military aircraft, cyber operations take place every day and in every corner of the world, including between allies and close partners. In cyberspace, activities that violate the rights of other States are very common, they take different forms and have a wide range of consequences. Moreover, cyber operations conducted by one State against another can have significant consequences for the economy of the targeted State and other States. I argue that the speed of development of cyber capabilities, the diversity of forms and effects of cyber operations, and their commonality and perception of risk are the factors that foster the specific objective of advancing attribution and accountability in cyberspace. Together with a clearer understanding of the interpretation of the rules of international law in cyberspace,[4] the development of an international attribution mechanism is sometimes considered necessary to achieve such an objective.
However, despite this specific objective, identifying the perpetrating or sponsoring State and the actors acting on its behalf is sometimes difficult; this observation could explain the particular interest in capacity-building and information sharing for the purpose of attribution. In addition, it is usually difficult for the victim to hold the wrongdoing State accountable, to obtain the cessation of cyber operations and the reparation of their consequence. This last observation could be the reason for the particular interest in collective and coordinated action, such as coordinated campaigns of public attribution of cyber operations and the introduction of dedicated sanction mechanisms.
This article aims to assess the merits of a potential international attribution and accountability mechanism and to offer some reflections on the possible forms and functions of such a mechanism. After this Introduction, section 2 analyses existing mechanisms and cooperations relevant to the attribution of cyber operations. Section 3 analyses the various proposals that have been made for the creation of such a mechanism. Sections 4 and 5 reflect on the possible objectives and main features of a possible international mechanism for attribution and accountability.

2Relevant existing mechanisms and institutions

Cyber operations do not take place in a vacuum, and the existing rules of international law are applicable to them.[5] Indeed, there is nothing pre-
venting the existing rules from applying to these activities.[6] Moreover, there is no specificity attached to cyber operations that would prevent the UNSC, the ICJ, or other international mechanisms from dealing with such threats to international peace and security. There is also no reason for the cyber-related damage to be excluded from the Register of Damage Caused by the Aggression of the Russian Federation against Ukraine established by the Council of Europe.[7] There are, however, other reasons that may explain why international institutions may not be able to deal with certain situations, such as the use of veto power by permanent members of the UNSC, which we will not discuss in this article. With the goal of advancing the attribution and accountability for cyber operations in mind, it is important to identify the specific mechanisms and practices in place.
Regarding the technical dimension of attribution, cyber incident investigations are generally conducted at the victim organization level, in particular by Computer Security Incident Response Teams (CSIRTs), and by national cybersecurity and law enforcement agencies. However, it is important to note that there is important cooperation between law enforcement agencies, either bilaterally or through Interpol or Europol. In addition, the Forum of Incident Response and Security Teams (FIRST) has been facilitating cooperation between these actors since 1990. At the European level, the European Union Agency for Cybersecurity (ENISA) coordinates cooperation among the information security authorities of the EU Member States. Institutions such as Interpol, Europol, ENISA or FIRST do not publicly attribute cyber operations to States, but they contribute to the exchange of information between different actors involved in the attribution of cyber operations.
The legal and political dimension of attribution generally falls within the competence of States. Some States, in particular the Five Eyes,[8] regularly publicly attribute cyber operations to other States and actors, including through coordinated and joint attribution campaigns.[9] The practice of public attribution by the United States, for example, takes generally one or more of four forms: ‘(1) criminal indictments; (2) economic sanctions; (3) technical alerts; and (4) official statements or press releases’.[10] Despite some abstract references to ‘international law’ and the ‘international rules-based order’, international law is generally absent from this practice,[11] and no explicit reference to the rules violated is made in the declarations of the attributing States.[12] Moreover, in case of coordinated attribution campaigns, a great diversity in the content of the declarations can be observed. This last observation emphasizes the need for greater cooperation, which is one of the key elements of the proposals developed in this article. Interestingly, it can also be observed that non-state actors – for example, non-governmental organizations (NGOs) such as Citizen Lab[13] and private companies such as Mandiant[14] and Microsoft[15] – also publicly attribute cyber operations to States and other actors.[16] States have also developed specific sanction regimes in relation to cyber operations. The European Union and its member States, for instance, have developed a collective diplomatic response to malicious cyber activities and sanction mechanism called the EU Cyber Diplomatic Toolbox.[17]
There are two other examples that are relevant to this article. Firstly, the Tallinn Mechanism for the coordination of civilian cyber assistance to Ukraine launched by ten States[18] in December 2023.[19] The Tallinn Mechanism focuses on crisis management in relation to the cyber operations targeting Ukraine and may thus be related, to some extent, to the investigation and attribution of cyber operations even if it is not its objective. Secondly, the CyberPeace Institute, an NGO created in 2019 and funded by the William & Flora Hewlett Foundation, Mastercard, Microsoft, and the Ford Foundation.[20] The CyberPeace Institute primarily analyses cyber operations, from the technical perspective and in assessing whether international law or norms of responsible behaviour have been violated. It also supports other NGOs and actors. One of their initiatives is a database that collects information on cyber operations related to the war in Ukraine, and provides an assessment of their attribution and whether they have violated international law and norms of responsible behaviour.

3 Navigating the different proposals

The idea of an international mechanism on cyber operations and international security is not new and is a recurring theme in the literature and in policy debates. In this section, I present the main proposals and highlight their main merits and shortcomings.
In 2014, the publication resulting from a NATO Advanced Research Workshop on Confidence-Building Measures in Cyberspace convened by the Atlantic Council and Swedish National Defense College proposed, among other confidence-building measures, the idea of creating a ‘[t]he Multilateral Cyber Attribution and Adjudication Council (MCAAC)’.[21] This proposal combines the two dimensions mentioned in the introduction, namely the attribution of cyber operations, on the one hand, and the questions of responsibility and accountability, on the other hand. The authors further the idea that the MCAAC could determine the ‘damage to be paid to the plaintiff by the defendant as compensation’.[22] The MCAAC would have no enforcement dimension and would refer the cases, alongside its recommendations and evidence, to the UNSC, the ICJ, or regional security bodies. On that point, the authors noted that ‘[t]hus, the MCAAC could become a specialized advisory agency for cyber violations of international law that formalizes and institutionalizes cyber attribution and adjudication’.[23]
A few years later, alongside to the idea of a Digital Geneva Convention,[24] Microsoft also proposed the creation of an ‘attribution organization’.[25] Microsoft’s proposal focused on both on attribution and the assessment of compliance with the rules of international law and non-binding norms of responsible behaviour in cyberspace.[26] In addition, Microsoft sponsored a research project conducted by the RAND Corporation to ‘assess the potential merits and challenges of an organization for performing cyber attribution, and to explore construction of such an organization’.[27] The result was a proposal for a ‘Global Cyber Attribution Consortium’ that would have a formal selection process for cases to be investigated, collect and assess evidence, develop specific standards and methodologies, and attribute and communicate with the relevant parties and through the publication of reports. Similar to the MCAAC, the Global Cyber Attribution Consortium could refer the case to the ICJ or the UNSC for further action.[28] Both the Microsoft and RAND proposals recommend that the mechanism be established as a non-governmental organization, with very limited participation by States, and composed of independent technical, policy, and legal experts. However, both mechanisms would generally focus on severe cyber operations sponsored by States.
In 2018, Serge Droz and Daniel Stauffacher wrote a report for the ICT4Peace Foundation, in which they argue in favour of a ‘Global Cyber Attribution Network’.[29] Although this proposal is based on similar considerations, it differs from the previous ones in that it does not aim to create a new institution, but to develop a network of actors involved in attribution – State representatives, private companies, civil society and academia – for the purpose of peer-review of attributions.[30] It should be emphasized, however, that while the report also mentions international law and policy matters, the proposed network focuses primarily on the technical dimension of attribution.
More recently, in an article published in International Law Studies in 2020, Michael N Schmitt and Yuval Shany commented on the earlier proposals and set out their own perspective.[31] In general, their proposal aims to support States with more limited capacities, in particular by providing them with independent access to better intelligence and investigation capacities, as well as to further accountability, especially in conjunction with existing sanction regimes.[32]
To summarize, most of the proposals have some common features. The main objective of the proposed mechanisms is generally the investigation of state-sponsored cyber operations, notably to collect and assess evidence. In other words, they focus primarily on the technical dimension of attribution. The ability of such a mechanism to conduct investigations and collect evidence is central in most proposals. As regards to the legal form and functioning of the mechanism, most proposals favour a mechanism that would be independent from States, for instance taking the form of a non-governmental organization with limited or no involvement of States. These proposals have not been implemented, apart from the creation of the CyberPeace Institute which is partly based on the Microsoft and RAND proposals.

4 The possible objectives of an international attribution and accountability mechanism

In the previous two sections, I have presented the various proposals for an international attribution and accountability mechanism as well as the existing relevant mechanisms and institutions. For me, however, the central question is what the objectives of such a mechanism could or should be. For an article that argues in favour of an international mechanism, the reader may find this section surprisingly critical. However, the aim is indeed to take a critical stance in order to exclude issues for which an international mechanism would not be appropriate and to identify those for which it would be relevant. Nicholas Tsagourias raises a number of criticisms against a possible international attribution mechanism, which are commented in this section.[33] I share most of his criticisms, but we generally disagree on the way forward. In this section, we will break down the different dimensions of attribution, looking at the technical, legal, and political dimensions in turn.
Identifying the perpetrators of a cyber operation involves investigating and collecting evidence.[34] I agree with Nicholas Tsagourias that the way cyberspace works makes it unlikely that States would accept an international investigation. It would imply that the victim State, and possibly also the accused State, would allow external investigators access to their computer networks and thus gain knowledge about their functioning and security. Based on this observation, the question arises as to whether it makes sense to create an international mechanism or whether it would be better to strengthen existing cooperations. I think the second option makes more sense. The first response to cyber incidents and crisis management are usually handled at the level of the victim organization and coordinated at the domestic level. As already mentioned, there are some forms of international cooperation and coordination in place, in particular through FIRST. Most proposals for an international mechanism focus on the issue of the cyber incident investigation and evidence collection. These proposals notably draw from the verification mechanisms in place within international organizations such as the International Atomic Energy Agency (IAEA) and the Organisation for the Prohibition of Chemical Weapons (OPCW), as well as existing examples of fact-finding missions. In general, verification mechanisms and investigations result from the existence of a specific treaty, an agreement between the parties on a specific dispute, or at the request of the UNSC on a specific situation. In addition, these investigations rely usually on reports and investigations by States and other actors and assess the evidence and information provided, which may differ from the objective contemplated with an international attribution mechanism. Building on these observations, I believe that an international mechanism conducting investigation is not an appropriate option. Conversely, I believe that cooperation between first responders, including private cybersecurity companies, law enforcement agencies, and other actors involved in the technical dimension of attribution as well as the review of evidence could be significantly improved by the creation of an international mechanism.[35]
The standard of evidence has become an important issue in the context of cyber operations,[36] as some States have developed a practice of public attribution. In this context, the United Nations Group of Governmental Experts (UNGGE) noted in its report that ‘the accusations of organizing and implementing wrongful acts brought against States should be substantiated’.[37] The standard of evidence to be used depends on the institution in question, in particular the dispute settlement mechanisms involved. However, I believe that an international mechanism operating at a pre-dispute phase could also provide some guidance on the definition of evidence, particularly in terms of setting certain standards on the collection and presentation of evidence. I assert that this can be an important contribution of an international mechanism.
The international legal dimension of attribution aims to determine whether a cyber operation can be attributed to a State. The Articles on Responsibility of States for Internationally Wrongful Acts adopted by the International Law Commission (ILC) in 2001 codify the norms of customary international law on State responsibility, including on the question of attribution of conduct.[38] I have analysed these rules and their application to cyber operations in detail elsewhere[39] and have concluded that the different types of situations arising from the conduct or sponsorship of cyber operations by a State are covered by the different norms of international law. I assert that the difficulty to attribute does not result from a non-liquet,[40] meaning that international law would not be able to provide a definitive legal characterization and answer to a specific situation, but from factors that are not related to the content of the norms. It is other factors, such as the availability of evidence, that generally make attribution more difficult.
One of the contentious issues is the threshold of control required to attribute the behaviour of a non-state actor to a State, either as a de facto organ or a non-state actor controlled and directed by the State. On this issue, the ICJ, the International Criminal Tribunal for the former Yugoslavia (ICTY) and the ILC have adopted different positions, which I will not detail here.[41] I observed that cyberspace provides at the same time an easy way for States to incentivize non-state actors, either individuals or groups, to act against another State, and to provide them with the necessary tools to act. This might be particularly true for a large-scale campaign of distributed denial-of-service (DDoS) attacks, such as the one against Estonia in 2007.[42] In such a situation, the level of control necessary for the State to incentivize the non-state actors is low and thus might be below the different thresholds identified by the ICJ, the ICTY, and the ILC. However, lowering the threshold of control may lead to over-attribution or misattribution to States, and thus be counterproductive. Building on these observations, I have argued that this may be one of the issues on which the rules of international law may evolve in the future due to the shift to cyberspace.[43] However, this challenge could not really be addressed by the creation of an international mechanism for cyber attribution per se, mainly because it is a general issue of international law that concerns the rules of international law themselves, and not only their interpretation and application in relation to cyber operations.
The political dimension of attribution is a sovereign prerogative of States, on this point I agree with Nicholas Tsagourias. I believe that the political dimension of attribution is unlikely to be transferred to an international mechanism, and my proposal focuses on assisting States in developing their practice. As previously emphasized, international law is generally absent from the current practice of public attribution. In other words, States accuse other States of conducting cyber operations, but in their statements they do not specify the legal basis of the attribution, the rules that have been violated, or the legal basis of an eventual response. Similar observations can be made on State practice outside of cyberspace. In cyberspace, however, this observation is to be put in perspective with the practice of interpretative statements on the application of international law, the call for the substantiation of attribution, and for the development of accountability. In this context, an international mechanism could support the political dimension of attribution by providing the necessary support for States to coordinate their practice, to exchange good practice and information, for instance through the development of standards of evidence collection.
Accountability is an important issue for international law in general, and even more so in relation to cyber operations. The accountability crisis is a broader issue that is not limited to cyber activities.[44] Existing mechanisms and institutions, such as the UNSC and the ICJ, are often criticized for their failure in ensuring compliance with international law and the management of crises. The creation of a specific mechanism dealing with cyber operations might thus have a limited interest, as the issue is not coming from the shift to cyberspace, but rather from the current state of international law.
In the case of cyber operations in particular, it can generally be observed that the responsible State often escapes consequences. Some States have developed a practice of collective and coordinated attribution campaigns, with the objective to increase the pressure on the wrongdoing State.[45] The same rationale, together with other considerations, explains the current push by some States for collective countermeasures in cyberspace.[46] By acting collectively, States expect to impose more consequences on the wrongdoing State and thus to limit its ability to evade accountability. This is also the reason for the development of coordinated sanction regimes. I assert that the same logic could justify the creation of an international mechanism, not to conduct political attribution but to support States in developing cooperation and synergies in this respect. Another advantage would be to assist in the contestation of public attributions. Not in being itself the forum for such a contestation, which can be done unilaterally by the concerned State or within existing fora, but by providing specific guidance on the various elements of attribution.

5 A possible international attribution and accountability mechanism for state-sponsored cyber operations

The aim of this section is to describe the main features of a possible international attribution and accountability mechanism for state-sponsored cyber operations. Nicholas Tsagourias is of the view that an international mechanism, should it be established, will focus on technical attribution. I do not share this view and have explained already why I believe that the technical dimension of attribution could be further coordinated without being transferred to an international mechanism. The proposal I am developing focuses on the issue of evidence and accountability. However, it is important to point out that my proposal is not a dispute settlement mechanism and would most likely operate in the pre-dispute phase.
Building on the observations made in the previous sections, I believe that an international mechanism could be developed to assist States in the investigation and attribution of cyber operations. This mechanism would be relevant in the pre-dispute phase, when the victim State conducts the investigation, collects and assesses the evidence, and evaluates the possible next steps. With regard to existing organizations, I believe that the proposed mechanism could, to a limited extent, play a similar role to the Permanent Court of Arbitration outside its adjudication functions. Indeed, the objective is not to create a dispute settlement mechanism. The main function of the mechanism would be to assist States in developing their approach to and practice of attribution and accountability, by providing them with possible standards for the collection and presentation of evidence that States could voluntarily follow, as well as lists of experts.
Such an international mechanism could be created as an international organization that is also open to other actors, for instance first responders such as CSIRTs, non-governmental organizations, and private cybersecurity companies. There are already international organizations with members that are non-state actors, for instance the World Tourism Organization and the International Telecommunication Union. However, the question will be what role they should play in the governance of the organization and the decision-making processes. This is an important difference from previous proposals, which advocated for a non-governmental organization with a limited role for States. Since the aim is to focus on the behaviour of States and to uphold international law, it is crucial to have States on board if we want them to accept the mechanism and to follow its recommendations.
A key objective of the mechanism would be to establish standards of evidence and promote compliance with international law and norms of responsible behaviour. The mechanism could assess existing evidentiary standards for attribution of cyber operations used in different contexts and propose rules and standards specifically tailored for the international context. The proposed standards would not be legally binding, and it would be up to the States to follow and acknowledge them in their practice. Furthermore, these standards could serve as a basis for fact-finding missions, as well as dispute settlement mechanisms at a later stage. On this point, Nicholas Tsagourias believes that my proposal would be relevant when an actual dispute occurs, whereas I believe that it would be relevant at a pre-dispute stage to help States characterize the situation and decide whether a dispute might exist and what the next steps might be. States and other actors could also be invited to report regularly on how they apply these standards in their practice. This reporting and its assessment could become an important dimension of the work of such a mechanism, with the aim of contributing to the further development and improvement of the standards.
A second important dimension of such a mechanism would be to contribute to capacity-building pursuing two directions. Firstly, in maintaining different lists of technical, legal and policy experts at the disposal of States and other actors. As highlighted in most proposals, States have different capacities, and States with lesser capacities may have more difficulties to attribute. However, rather than proposing a centralized mechanism that may have difficulty responding to very different situations, I suggest a mechanism that helps to connect the relevant experts with the States requesting assistance. In addition, more advanced States could financially support other States to cover the costs of engaging these experts. Secondly, this mechanism could organize training and other exercises aimed at developing the capacity of States and other actors, particularly on the basis of the expert lists and reporting activities. There are already various States and international organizations conducting such activities, and the question would be how synergies could be developed.
One might ask why States and other actors should be interested in such a mechanism. As noted previously, the UNGGE called for the substantiation of attribution[47] while the question of accountability is at the centre of the current debate on State behaviours in cyberspace. Such a mechanism would provide a voluntary basis to advance these questions, in particular in supporting the development of State practice on attribution, including on contestation in case of misattribution. In this context, such a mechanism could also contribute to the development of the framework of responsible State behaviour in cyberspace which is at the heart of the work of the United Nations Open-Ended Working Group and the future Programme of Action.

6 Conclusions

The international attribution and accountability mechanism proposed in this article differs from previous proposals in that it does not seek to investigate or attribute cyber operations, but rather to support States that engage in such a practice. It is also important to note that it is not a dispute settlement mechanism. The objective is for States to benefit from appropriate assistance, in the form of expertise, evidentiary standards, best practices, and capacity-building.

* Assistant Professor of International Law, IE University (francois.delerue@ie.edu). I am thankful to Professor Nicholas Tsagourias for engaging with me in this discussion, as well as to Professor Emanuele Cimiotta and the anonymous reviewer for their valuable suggestions and comments.
[1] United Nations, ‘Our Common Agenda Policy Brief 9: A New Agenda for Peace’ (July 2023) 27 ˂https://www.un.org/sites/un2.un.org/files/our-common-agenda-policy-brief-new-agenda-for-peace-en.pdf˃.
[2] On norms of responsible behaviour in cyberspace, see M Lehto, ‘The Rise of Cyber Norms’ in R Buchan, N Tsagourias (eds), Research Handbook on International Law and Cyberspace (Elgar 2021) 32-45; B Hogeveen, ‘The UN Cyber Norms: How Do They Guide the Responsible Development and Use of Offensive Cyber Capabilities?’ (2022) 7 The Cyber Defense Review 123, 123-142.
[3] The main proposals are presented in Section 3. See, also, the analysis in B Kuerbis, F Badiei, K Grindal, M Mueller, ‘Understanding Transnational Cyber Attribution: Moving from “Whodunit” to Who Did It’ in MD Cavelty, A Wenger (eds), Cyber Security Politics: Socio-Technological Transformations and Political Fragmentation (Routledge 2022) 230-232.
[4] To date (July 2024), thirty-two States and the African Union have publicly issued detailed statements on their approach to the rules and principles of international law applicable to cyberspace. See, generally, the updated list of national positions on the website of the ‘International Cyber Law in Practice: Interactive Toolkit’ project ˂https://cyberlaw.ccdcoe.org/wiki/List_of_articles#National_positions˃.
[5] On the application of international law to cyber operations, see generally, HH Dinniss, Cyber Warfare and the Laws of War (CUP 2012); G Kerschischnig, Cyberthreats and International Law (Eleven International Publishing 2012); M Roscini, Cyber Operations and the Use of Force in International Law (OUP 2014); Y Radziwill, Cyber-Attacks and the Exploitable Imperfection of International Law (Brill/Martinus Nijhoff Publishers 2015); MN Schmitt, L Vihul (eds), The Tallinn Manual 2.0 on the International Law Applicable to Cyber Operations (2nd edn, CUP 2017); H Lahmann, Unilateral Remedies to Cyber Operations: Self-Defence, Countermeasures, Necessity, and the Question of Attribution (CUP 2020); D Akande, A Coco, TS Diaz, ‘Drawing the Cyber Baseline: The Applicability of Existing International Law to the Governance of Information and Communication Technologies’ (2022) 99 Intl Law Studies 4, 4-36; F Delerue, A Géry, A-T Norodom (eds), Digital Challenges for International Law (ILA 2023). Members of the ILA steering committee as of 2023: E Benvenisti, N Bhuta, D Hollis, Z Huang, N Ifeanyi-Ajufo, E Ivanov, J Kulesza, C Strydom, J Tridgell, R Young.
[6] F Delerue, Cyber Operations and International Law (CUP 2020) 1-50.
[7] Council of Europe, ‘Resolution CM/Res (2023)3 establishing the Enlarged Partial Agreement on the Register of Damage Caused by the Aggression of the Russian Federation against Ukraine’ (12 May 2023) adopted by the Committee of Ministers at the 1466th meeting of the Ministers’ Deputies.
[8] Australia, Canada, New Zealand, United Kingdom, and the United States.
[9] A Paulus, C Rupp, ‘Official Public Political Attribution of Cyber Operations: State of Play and Policy Options’ (12 October 2023) Stiftung Neue Verantwortung ˂https://www.interface-eu.org/publications/official-public-political-attribution-of-cyber-operations˃; FJ Egloff, M Smeets, ‘Publicly Attributing Cyber Attacks: A Framework’ (2023) 46 Journal of Strategic Studies 502, 502-533; Kuerbis and others (n 3) 221-230; M Finnemore, DB Hollis, ‘Beyond Naming and Shaming: Accusations and International Law in Cybersecurity’ (2020) 31 European J Intl L 969, 969-1003; T Rid, B Buchanan, ‘Attributing Cyber Attacks’ (2015) 38 Journal of Strategic Studies 4, 4-37; D Broeders, E Busser, P Pawlak, ‘Three Tales of Attribution in Cyberspace: Criminal Law, International Law and Policy Debates’ (April 2020) The Hague Program for Cyber Norms ˂https://www.thehaguecybernorms.nl/research-and-publication-posts/three-tales-of-attribution-in-cyberspace-criminal-law-international-law-and-policy-debates˃.
[10] KE Eichensehr, ‘The Law and Politics of Cyberattack Attribution’ (2020) 67 UCLA Law Review 520-598, 532.
[11] Finnemore and Hollis (n 9) 997-1000.
[12] F Delerue, ‘Is International Law Fading Away in State Practice on Cyber Operations?’ in F Cristiano, B van den Berg (eds), Hybridity, Conflict, and the Global Politics of Cybersecurity (Rowman & Littlefield 2023) 31-52.
[13] M Kenyon, ‘Citizen Lab Response to the UN Working Group on the Use of Mercenaries’ (18 February 2021) Citizen Lab, University of Toronto <https://citizenlab.ca/2021/02/citizen-lab-response-to-the-u-n-working-group-on-the-use-of-mercenaries/>.
[14] Mandiant, ‘APT1 Exposing One of China’s Cyber Espionage Units’ (30 December 2021) ˂https://www.mandiant.com/resources/reports/apt1-exposing-one-chinas-cyber-espionage-units˃
[15] Microsoft Threat Intelligence, ‘Microsoft Investigates Iranian Attacks against the Albanian Government’ (8 September 2022) Microsoft Security ˂https://www.microsoft.com/en-us/security/blog/2022/09/08/microsoft-investigates-iranian-attacks-against-the-albanian-government/˃.
[16] Eichensehr (n 10) 547-551.
[17] European Union, ‘Council Conclusions on a Framework for a Joint EU Diplomatic Response to Malicious Cyber Activities (“Cyber Diplomacy Toolbox”)’ doc 10474/17 (19 June 2017); European Union, ‘Council Conclusions on exploring the potential of the Joint Cyber Unit initiative – complementing the EU Coordinated Response to Large-Scale Cybersecurity Incidents and Crises’ doc 13048/21 (19 October 2021); European Union, ‘Council conclusions on the development of the European Union’s cyber posture’ doc 9364/22 (23 May 2022).
[18] Canada, Denmark, Estonia, France, Germany, the Netherlands, Poland, Sweden, the United Kingdom and the United States.
[19] Estonia, Ministry of Foreign Affairs, ‘Tallinn Mechanism, Mission Statement’ (30 May 2023) ˂https://vm.ee/sites/default/files/documents/2023-12/Mission%20Statement%20-%20Scope.pdf˃.
[20] CyberPeace Institute, ˂https://cyberpeaceinstitute.org/who-we-are/˃.
[21] J Healey, JC Mallery, KT Jordan, NV Youd, ‘Confidence-Building Measures in Cyberspace: A Multistakeholder Approach for Stability and Security’ Atlantic Council of the United States/Swedish National Defense College (November 2014) ˂https://www.files.ethz.ch/isn/185487/Confidence-Building_Measures_in_Cyberspace.pdf˃ 10-12.
[22] ibid 11.
[23] ibid.
[24] B Smith, ‘The Need for a Digital Geneva Convention’ (14 February 2017) Microsoft <https://blogs.microsoft.com/on-the-issues/2017/02/14/need-digital-geneva-convention/>.
[25] Microsoft, ‘An Attribution Organization to Strengthen Trust Online: Policy Paper’ (2017) <https://www.microsoft.com/en-us/cybersecurity/content-hub/an-attribution-organization-to-strengthen-trust-online>.
[26] S Charney and others, ‘From Articulation to Implementation: Enabling Progress on Cybersecurity Norms’ (2016) Microsoft 9-12 <https://www.microsoft.com/en-us/cybersecurity/content-hub/enabling-progress-on-cybersecurity-norms>.
[27] JS Davis and others, ‘Stateless Attribution: Toward International Accountability in Cyberspace’ (2017) RAND 6 <https://www.rand.org/pubs/research_reports/RR2081.html>.
[28] ibid 35-41.
[29] S Droz, D Stauffacher, ‘Trust and Attribution in Cyberspace: A Proposal for an Independent Network of Organisations Engaging in Attribution Peer-Review’ (2018) ICT4Peace Foundation ˂https://ict4peace.org/wp-content/uploads/2019/07/ICT4Peace-2019-Trust-and-Attribution-in-Cyberspace.pdf˃.
[30] ibid 8; this proposal builds on a proposal made by Ron Deibert at the RightsCon conference for a global network of researchers from universities on attribution of state-sponsored cyber operations, see H Solomon, ‘RightsCon Report: Universities Should Form Cyber Attribution Network’ (18 May 2018) IT World Canada News <https://www.itworldcanada.com/article/rightscon-report-universities-should-form-cyber-attribution-network/405399>. See, also, on the possible role of academia FJ Egloff, ‘Contested Public Attributions of Cyber Incidents and the Role of Academia’ (2020) 41 Contemporary Security Policy 55, 55-81.
[31] MN Schmitt, Y Shany, ‘An International Attribution Mechanism for Hostile Cyber Operations?’ (2020) 96 Intl Law Studies 196, 196-222.
[32] ibid 219-221.
[33] See N Tsagourias, ‘Cyber Attribution Agencies: A Sceptical View’ in this Zoom in, infra 23-38.
[34] N Tsagourias, M Farrell, ‘Cyber Attribution: Technical and Legal Approaches and Challenges’ (2020) 31 European J Intl L 941, 945, 955-959.
[35] For a similar view, see Droz and Stauffacher (n 29).
[36] M Roscini, ‘Digital Evidence as a Means of Proof before the International Court of Justice’ (2016) 21 Journal of Conflict & Security Law 541, 541-554.
[37] UNGA, ‘Report of the Group of Governmental Experts on Developments in the Field of Information and Telecommunications in the Context of International Security’ UN Doc A/70/174 (22 July 2015) 28(f).
[38] International Law Commission, ‘Report of the International Law Commission on the Work of Its Fifty-third Session’ (23 April-1 June and 2 July-10 August 2001) UN Doc A/56/10 (2001) II/2 YB ILC.
[39] Delerue (n 6) 111-191. See also Tsagourias and Farrell (n 34) 951-955; K Mačák, ‘Decoding Article 8 of the International Law Commission’s Articles on State Responsibility: Attribution of Cyber Operations by Non-State Actors’ (2016) 21 Journal of Conflict & Security Law 405, 405-428.
[40] A non liquet would not result from the lack of evidence or the impossibility to identify the perpetrator but from the impossibility to apply or find an answer in the rules and principles of international law on the attribution of behaviours to States. On non liquet, see generally D Bodansky, ‘Non Liquet’ in R Wolfrum (ed), Max Planck Encyclopaedia of Public International Law (OUP 2008).
[41] See, generally, O De Frouville, ‘Attribution of Conduct to the State: Private Individuals’ in J Crawford, A Pellet, S Olleson (eds), The Law of International Responsibility (OUP 2010) 257-280; P Palchetti, L’organo di fatto dello Stato nell’illecito internazionale (Giuffrè 2007); A Cassese, ‘The Nicaragua and Tadić Tests Revisited in Light of the ICJ Judgment on Genocide in Bosnia’ (2007) 18 European J Intl L 649-668. See also Delerue (n 6) 118-122, 130-144.
[42] Delerue (n 6) 146-149.
[43] ibid 185-186. See also the discussion in Tsagourias and Farrell (n 34) 961-965.
[44] See, generally, S Besson (ed), Theories of International Responsibility Law (CUP 2022) 7-8.
[45] Finnemore and Hollis (n 9) 1001.
[46] See, generally, M Jackson, FI Paddeu, ‘The Countermeasures of Others: When Can States Collaborate in the Taking of Countermeasures?’ (2024) 118 American J Intl L 231, 259-267; T Dias, ‘Countermeasures in International Law and Their Role in Cyberspace’ (23 May 2024) Chatham House 33-55 ˂https://www.chathamhouse.org/2024/05/countermeasures-international-law-and-their-role-cyberspace˃.
[47] UN Doc A/70/174 (n 37) 28(f).

1. The endless obscurity of universal jurisdiction

The past few years have revealed the growing excitement, interest and appreciation for universal criminal jurisdiction (henceforth, universal jurisdiction). The wide use of this notion in the media,[1] observations by States,[2] reports from several organisations,[3] discussions in the United Nations General Assembly,[4] as well as in academic sources,[5] among others, give the impression that universal jurisdiction is one of the most powerful tools in the fight against impunity for heinous crimes.[6]

To name a few examples of the contexts in which criminal cases have been brought, which have relied on the principle of universal jurisdiction, these include Afghanistan, Iraq, Liberia, Syria, Palestine, Rwanda, Ukraine and Venezuela.[7] Moreover, the increase of trials which are (at least allegedly) based on the principle of universal jurisdiction in, for instance, Argentina, Belgium, France, Germany, the Netherlands and Switzerland have contributed to the constantly increasing expectations regarding what universal jurisdiction can bring to the future of (international) criminal justice.[8] These trials have filled the hearts of many victims with high hopes, in particular in countries where there is limited (or even no) chance to seek justice in the State where the crimes are perpetrated. With the highest number of domestic cases based on universal jurisdiction, perhaps Syria is the most illustrative case as to show how universal jurisdiction can play a decisive role for victims hoping for justice when the access to it is completely blocked domestically.[9]

On the one hand, the thrilling attitude towards universal jurisdiction is justified, given that the very idea of applying national laws over extraterritorial conduct or even prosecuting a foreigner who committed a crime abroad against another foreigner, is surely revolutionary. This is because the ‘traditional’ heads of jurisdiction which include first and foremost territoriality, active personality and protective principle and (less ‘traditional’/more controversial) passive personality, all align with the positivist boundaries of State sovereignty. Differently, universal jurisdiction goes beyond ‘the traditional’ understanding of sovereignty and challenges the limitations of the State authority.[10] On the other hand, the efforts to bring the concept closer to a large audience have led to many simplifications which may leave the reader with the wrong impression about what universal jurisdiction truly is, leading to false expectations regarding how many cases of universal jurisdiction we have documented in the recent years. Additionally, the doctrinal sources do not always examine the term with sufficient scrutiny.[11]

Overall, despite the great interest in universal jurisdiction, and multiple previous academic efforts to clarify all the nuances regarding the concept, one may still sense a general unfamiliarity with the subtle distinctions on the content of the notion.[12] This paper offers a straightforward journey through the basics of universal jurisdiction, with the objective of providing clarity and to nourish reflections about the use of this term, which it is hoped can therefore have an impact on future analyses. The article begins by exploring the definition of universal jurisdiction, and continues by discussing the ongoing confusion between conditional and absolute universal jurisdiction. It also addresses the relationship between the treaty-based obligation of aut dedere aut judicare and universal jurisdiction.

2. Nuanced definition of universal jurisdiction

One of the most common ways to describe universal jurisdiction is to say that it is the possibility of a State to adopt criminal laws based on ‘universality’, meaning – at least in this context – that they would apply to conduct by a non-national which was committed abroad against another non-national, even when no other link exists between that State and the offender.[13] The prevalent doctrine considers the moment of the commission of the crime to be the relevant moment at which to assess the absence of the link, however, there is no universal consensus on this question.[14] Hence, it is impossible to judge objectively which jurisdictional basis has been applied by a State in question.[15]

Although it is often not stressed sufficiently, the abovementioned definition leaves no doubt that universal jurisdiction is per se a prescriptive (legislative) form of jurisdiction.[16] In this sense, the concept shares an essential feature with the ‘traditional’ heads of jurisdiction: it refers to the authority of the State to enact domestic laws (establishment of jurisdiction).[17] Yet, one thing which is distinctive for universal jurisdiction over other heads of jurisdiction is that it is always marked by the extraterritorial nature. Although they are both extraterritorial in nature, the two terms (universal and extraterritorial jurisdiction) should not be used interchangeably precisely because the former (universal jurisdiction) is a subcategory of the latter (extraterritorial jurisdiction).[18]

Still, a reference to extraterritorial jurisdiction as a general term to describe situations where the crime took place outside the territory of a State where the case is litigated is not per se incorrect. In fact, it may serve the purpose of emphasising the complexity of such cases (for example, due to evidence related constraints, political sensitivities, knowledge of the elements of international crimes and how to prove them by the domestic prosecutors and judiciary). Additionally, ‘extraterritorial’ as an umbrella term for universal jurisdiction is a safe choice of terminology whenever it is not yet clear which jurisdictional basis can and will be applied in the case in question. What is truly important is that the terminology is never incidental but rather a result of well-thought-out process in each individual case.[19]

Universal (prescriptive) jurisdiction tends to come with the corresponding form of the adjudicatory form of universal jurisdiction, which is about the possibility of criminal courts to address cases based on universality (exercise of jurisdiction).[20] But, can States even exercise criminal jurisdiction extraterritorially (be it on the basis of universality or other heads of jurisdiction)? The Lotus case, where the International Court of Justice (henceforth, ICJ) was asked to specify whether any principles of international law prevented Turkey from instituting criminal proceedings against the defendants, offers a response to this question.[21]

The majority in Lotus asserted that the State cannot exercise jurisdiction outside its territories unless there is a specific international law provision or customary international law which allows it to do so.[22] At the same time, most judges thought that the establishment and exercise of the extraterritorial jurisdiction in criminal matters by States in their territory is permissible, unless an international law rule encompasses an explicit prohibition to that end.[23] Conversely, the dissenting judges in Lotus were aligned with the more modern approach which goes in favour of the territoriality of criminal jurisdiction.[24] Exception are those situations where international law explicitly allows for extraterritorial jurisdiction as a head of jurisdiction (one example being international treaties).

It is also possible to ‘dissociate’ prescriptive and adjudicative jurisdiction. Such an option is best manifested by the possibility of domestic courts being able to exercise jurisdiction over a certain crime, based on international law as the direct source of reference, and without having the prescriptive jurisdiction in domestic legislation.[25] However, exercising jurisdiction extraterritorially based on the direct reliance on international law – especially customary law – is rare.[26] In fact, as concerns universal jurisdiction specifically, there is no known case where the State would exercise it based on customary international law without having the basis in domestic law.[27]

Next, the French case Javor et al testifies about the challenge of relying on international law directly for the exercise of extraterritorial jurisdiction even when the basis is a treaty.[28] The case was about five Bosniacs, refugees in France, who brought a civil complaint before the French criminal courts against unknown persons for various crimes, including grave breaches of the Geneva Conventions of 1949, committed against them in Bosnia and Herzegovina during the conflict of 1992.[29] At first instance, the investigating judge asserted that the Geneva Conventions of 1949 and the Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment (Torture Convention) are directly applicable in the French domestic system.[30] Yet, the Court of Cassation eventually decided that both treaties were inapplicable.[31] The Torture Convention was inapplicable because the defendants were not in the territory of France whereas this is required by the concerned treaty, and the Geneva Conventions of 1949 were perceived as too general to create rules on extraterritorial jurisdiction in criminal matters which require a certain level of preciseness and hence, cannot be directly applied in a domestic criminal system.[32] Another example, which was also adjudicated on by the French courts, is Re Munyeshyaka. This case was about a Hutu priest who had participated in the genocide in Rwanda and later fled to the French territory where he was found. In Re Munyeshyaka, the French judiciary decided that international law, including the Geneva Conventions, do not impose universal jurisdiction and hence, they cannot serve as a basis for the domestic courts to exercise it.[33]

According to Stern, Re Munyeshyaka serves as a proof that the French courts were unready to accept that customary law can be the basis for universal jurisdiction.[34] As previously mentioned, such an approach is pretty widespread among domestic judiciaries as most domestic courts would require the incorporation of universal jurisdiction as a head of jurisdiction into domestic legislation.[35] Still, generally speaking, the application of international treaties as well as customary law, including when it comes to the heads of jurisdiction, before criminal courts might be possible depending on the rules of the domestic system concerned.[36]

3. The presence of the defendant and the confusion between absolute and conditional universal jurisdiction

States tend to provide conditions for universal jurisdiction in their domestic laws, and (at least probable) presence of the defendant on the territory of a State is a frequent requirement.[37] One could say that a rationale of the requirement of the presence of the defendant is the interpretation of the ‘specific link’ between the State, which extends its criminal laws over the conduct of the concerned offender, which therefore justifies State interest in the given matter. Additionally, States tend to proceed with investigations or prosecutions based on universal jurisdiction when the defendant is present on their territory.[38] Both the frequency of the ‘condition of the presence’ in the legislation on universal jurisdiction and the tendency of States to proceed with cases on universal jurisdiction when the defendant is in their territory have contributed to the occasional belief that the presence of the perpetrator is always required for a State to either apply laws over the criminal conduct, or/and to begin an investigation or prosecution based on universal jurisdiction.[39] But, is this really the case?

To be able to answer the above question, the distinction between the absolute and conditional form of universal jurisdiction comes into play. One may be tempted to simplify by saying that absolute universal jurisdiction is the type of universal jurisdiction which requires no presence of the perpetrator on the State territory. This is nonetheless not sufficiently precise. In fact, the best way to begin is with a question on whether a State can even adopt a framework to regulate the behaviour of foreigners against foreigners, when the defendants have no link with the State, they have never entered a State territory after perpetrating the crime, and there is no prospect that they will enter voluntarily. Such a broad type of jurisdiction could be described as absolute or pure universal jurisdiction,[40] sometimes (wrongly) also called ‘jurisdiction in absentia’.[41] Given that absolute universal jurisdiction is per se of a prescriptive nature, it would usually come with the corresponding adjudicatory jurisdiction, implying that a State can initiate an investigation and possibly also a prosecution without the presence of the offender on the territory.

The exercise of absolute universal jurisdiction may be problematic from the perspective of the principle of legality. As one of the foundational principles of criminal law, the principle of legality calls for the accessibility and foreseeability of the prohibition of the criminal conduct and the penalties attached to it at the moment of the offence’s commission,[42] and the exercise of absolute universal jurisdiction would imply that a State may initiate an investigation or even a prosecution even when the perpetrator is not on a State territory. Therefore, applying the domestic definitions of crimes and the attached penalties to the criminal conduct committed abroad and when the State has no link to the perpetrator may clash with the principle of legality. For instance, this would happen when the law which a concerned State applies defines the alleged crime in a broader, less ‘perpetrator-friendly’ manner than the law of the State in which the perpetrator committed the crime or in the country of their nationality. Moreover, a risk to violate legality exists if the pertinent penalty for the alleged offence is harsher in the State which establishes and exercises pure universal jurisdiction compared to the pertinent law in States with other heads of jurisdiction (for example, with territorial or active personality heads of jurisdiction).[43]

Next, when it comes to conditional universal jurisdiction, it is about the application of national legislation to the conduct carried out abroad by a foreigner against a foreigner if the perpetrator is on the territory of a State.[44] The presence of the defendant on the forum State territory is in this sense necessary for the extension of the laws to the criminal conduct of the perpetrator (forum deprehensionis).[45] Similar to absolute universal jurisdiction, conditional universal jurisdictional usually comes with the corresponding form of the adjudicative jurisdiction, implying that the exercise is dependent on the presence of the perpetrator. There is also another possible scenario, which is that the presence of the defendant is not, as such, required in the title of universal jurisdiction, yet, it is mandatory for its exercise. In other words, one should distinguish between two situations: first, the question of forum deprehensionis, which is the requirement of the presence in the title of the jurisdiction, meaning that the domestic laws apply to the criminal conduct of the perpetrator who is present in the State and second, when the presence is required only for the adjudicatory jurisdiction.

Differently from the exercise of the absolute universal jurisdiction, the adjudicative form of conditional universal jurisdiction is less problematic from the perspective of the principle of legality because the presence of the subject in the territory can be interpreted as their voluntary subjection to the relevant domestic laws.[46] This is perhaps the reason why such conditional universal jurisdiction is most common in practice even in those countries which have the pure universal jurisdiction as a specific jurisdictional title in the domestic legislation, the best example being Germany.[47] At the same time, the lack of practice on the exercise of absolute universal jurisdiction does not imply that it is prohibited. Following the majority in the Lotus case, the question should actually be posed differently, namely, whether any prohibitive rule bars universal jurisdiction in the absence of a suspect on the State territory. In this regard, most scholars refer to the ICJ Arrest Warrant of 11 April 2000 case as instructive.[48] Especially the dissenting opinions clarified that the exercise of universal jurisdiction is a permissive form of jurisdiction, provided that certain requirements are met (most notably, the respect for the international law of immunities and only concerning most heinous crimes).[49] Yet, the dissenting judges in Lotus left some doubt regarding the form of universal jurisdiction, with the exception of piracy, for which they considered a possibility to assert and exercise absolute universal jurisdiction.[50]

One way to draw conclusions on the status of the absolute and conditional form of universal jurisdiction is also by referring to the gravity of certain crimes which affect the interest of international society as a whole. The notion of universal jurisdiction is often linked to crimes which are perceived to be of a particular gravity, and legitimate interest in the prosecutions based on universal jurisdiction is grounded in the seriousness of these crimes, which affect not only individual members, but society as a whole.[51] Nevertheless, even for crimes such as piracy, genocide, war crimes, crimes against humanity and (perhaps) torture, for which there seems to be a general agreement on the permissible legislative and adjudicative universal jurisdiction under customary international law,[52] there is no certainty regarding whether the permissible form of universal jurisdiction is the absolute or conditional form.[53]

4. The relationship between universal jurisdiction and aut dedere aut judicare

The relationship between universal jurisdiction and the specific treaty provision on the duty to prosecute or extradite (aut dedere aut judicare) enshrined in treaties such as the Torture Convention and the Geneva Conventions of 1949 deserves specific mention.[54] This is because the duty to prosecute or extradite can, under certain conditions, coincide with the conditional universal jurisdiction (as a prescriptive title of jurisdiction). Still, since this is not always the case, the two terms (aut dedere aut judicare and universal jurisdiction) cannot be used interchangeably.

Specifically, whereas aut dedere aut judicare imposes a duty to exercise jurisdiction, it may also be seen as imposing a duty of prescriptive jurisdiction in the sense that, for States to be able to meet their duty to prosecute or extradite, they have to adjust their criminal laws so as to include all heads of jurisdiction (including universal jurisdiction) in the national legislation to be able to fulfil such a duty.[55] Hence, whenever aut dedere aut judicare acts as a basis for establishing criminal jurisdiction, it overlaps with the conditional form of universal jurisdiction.

To illustrate the above with reference to the Torture Convention, this treaty crystallises the scope and content of the duty to prosecute or extradite by saying that States parties are required to submit the relevant case to the national authorities for the purpose of the prosecution or, alternatively, to extradite the suspect (Article 7).[56] At the same time, the heads of jurisdiction mentioned in the Torture Convention as a duty to establish criminal jurisdiction are based on principles of territoriality, active and passive personality and universality (regardless of any other link and head of jurisdiction) if the perpetrator is found on their territory (forum deprehensionis) (Article 5). One way to interpret the relationship between these two provisions, namely Article 7 and 5 of the Torture Convention, is to say that dedere aut judicare in Article 7 imposes the titles of jurisdiction, including the universal, spelled out in Article 5. In this sense, for the purpose of the Torture Convention, aut dedere aut judicare would overlap with (conditional) universal jurisdiction as a prescriptive title.

Another example of aut dedere aut judicare provisions are the Geneva Conventions of 1949.[57] Specifically, the obligation is formulated as the duty to search for persons who have allegedly committed grave breaches, bring them before their own courts or to hand them over to another State party for trial if that State has made a prima facie case, regardless of the nature of the link between the grave breach and the State.[58] Interestingly, the Geneva Conventions make no reference to the presence of the suspect on the territory, thus the question arises whether States parties should also search for and bring before their own courts perpetrators outside their territory.[59] Doctrine and also State practice speak against such a conclusion, and most countries have made prosecution conditional on the temporary or permanent presence of the alleged offender on the territory of the State party.[60]

Precisely because the obligation to search for and submit for prosecution is not limited to any specific jurisdictional basis, some scholars,[61] judicial opinions,[62] the International Committee of the Red Cross[63] and the majority of States[64] have concluded that the Geneva Conventions of 1949 provide for mandatory universal jurisdiction, arguing that the only way to implement the duty to search for or extradite is through including universality in the domestic legislation. More specifically, the choice of words at the beginning of the second paragraph on penal sanctions ‘each high contracting party shall be under the obligation to search (…) and shall bring such persons, regardless of their nationality, before its own courts (…)’, which establishes aut dedere aut judicare, is decisive for many States and commentators to believe that ‘universality’ is implied in the Geneva Conventions.[65] Others, however, have opposed the view, suggesting that aut dedere aut judicare does not as such include an obligation to assert extraterritorial jurisdiction because it is conceptually distinct.[66]

Independently of the view one adopts, it is evident that universal jurisdiction and aut dedere aut judicare also cannot be used interchangeably in the context of the Geneva Conventions. This is, as previously stated, because aut dedere aut judicare is about the exercise of jurisdiction, which might imply the corresponding duty of establishing prescriptive jurisdiction for this obligation to be met. At the same time, the aut dedere aut judicare principle is broader in nature than just referring to universality, as it encompasses an obligation for States to prosecute under any applicable jurisdictional basis. It also exists independently from the titles of jurisdiction. This makes the concepts of universal jurisdiction and aut dedere aut judicare interrelated but inherently distinct.[67]

5. Concluding remarks

With the hope of clarifying the persistent confusion regarding universal jurisdiction, this paper has highlighted that first, universal jurisdiction is per se a prescriptive form of universal jurisdiction, which is usually accompanied by the corresponding adjudicative form. Second, one should distinguish between the different roles which the presence of the offender may play in absolute and conditional forms of universal jurisdiction. The exercise of absolute universal jurisdiction (which would imply no requirement of presence of the defendant on the territory of a State exercising jurisdiction) is the one which raises the most controversy. Finally, by using the examples of the Torture Convention and the Geneva Conventions of 1949, this paper has portrayed the difference between universal jurisdiction and aut dedere aut judicare, and how they interact whenever the latter is imposed by an international treaty. Whereas aut dedere aut judicare may, whenever it is enshrined in the laws as a head of jurisdiction, sometimes overlap with a conditional form of universal jurisdiction, these two terms are not synonymous.

The progress on universal jurisdiction can only be appreciated if there is an understanding about what the notion really means. A desirable step in the future would be to conduct a study into how many cases, which have been labelled as ‘instances of universal jurisdiction’ would correspond to any of the described forms of universal jurisdiction, and why. This would allow us to get closer to the reality regarding the state of the art on universal jurisdiction.

* PhD Candidate and Teaching Assistant, Geneva Graduate Institute. The author is grateful to Agustina Becerra Vázquez, Chiara Gabriele, Victoria Priori, and anonymous reviewers for their constructive comments on the previous versions of this article.

[1] L Morris, ‘Why Germany Is Becoming a Go-to Destination for Trials on the World’s Crimes’ Washington Post (6 March 2021) <www.washingtonpost.com/world/europe/germany-war-crimes-justice/2021/03/05/b45372f4-7b78-11eb-8c5e-32e47b42b51b_story.html>; F Franchini, ‘Universal Jurisdiction Gains Ground in Switzerland’ SWI swissinfo.ch (21 July 2023) <www.swissinfo.ch/eng/society/universal-jurisdiction-gains-ground-in-switzerland/48669650>.

[2] ‘The Scope and Application of the Principle of Universal Jurisdiction. Information Concerning the Position of Switzerland Submitted under Paragraph 3 of General Assembly Resolution 75/142 of 15 December 2020’ <www.un.org/en/ga/sixth/75/universal_ jurisdiction/switzerland_e.pdf>; ‘Observations by Belgium on the Scope and Application of the Principle of Universal Jurisdiction’ <www.un.org/en/ga/sixth/65/ScopeAppUniJuri_ StatesComments/Belgium_E.pdf>; ‘United States Submission. Information and Observations on the Scope and Application of the Universal Jurisdiction’ <www.un.org/en/ga/sixth/65/ ScopeAppUniJuri_StatesComments/United%20States.pdf>.

[3] ‘A Conceptual Framework for Dealing with the Past’ (swisspeace) Essential 2/2016 10; S Finnin, ‘Breaking Down Barriers: Access to Justice in Europe for Victims of International Crimes’ (International Federation for Human Rights, European Center for Constitutional and Human Rights, REDRESS 2020); ‘Universal Jurisdiction over War Crimes’ (Advisory Service on IHL, International Committee of the Red Cross 2021).

[4] United Nations General Assembly (UNGA) ‘The scope and application of the principle of universal jurisdiction. Report of the Secretary-General’ UN Doc A/73/123 (3 July 2018); UNGA ‘The scope and application of the principle of universal jurisdiction. The Report of the Secretary-General’ UN Doc A/74/144 (11 July 2019); UNGA ‘The scope and application of the principle of universal jurisdiction. Report of the Secretary-General’ UN Doc A/76/203 (21 July 2021).

[5] G Abi-Saab, ‘The Proper Role of Universal Jurisdiction’ (2003) 1 J Intl Crim Justice 596; AJ Colangelo, ‘The Legal Limits of Universal Jurisdiction’ (2006) 47 Virginia J Intl L 149; M Langer, M Eason, ‘The Quiet Expansion of Universal Jurisdiction’ (2019) 30 Eur J Intl L 779; M Tiernan, A Al-Zien, ‘Domestic Modes of Liability in Universal Jurisdiction Cases: The Case of Eyad Al-Gharib in Koblenz, Germany’ Rethinking SLIC (11 April 2022) <https://rethinkingslic.org/blog/ criminal-law/150-domestic-modes-of-liability-in-universal-jurisdiction-cases-the-case-of-eyad-al-gharib-in-koblenz-germany>; YM Dutton, ‘Prosecuting Atrocities Committed in Ukraine: A New Era for Universal Jurisdiction?’ (2023) 55 Case Western Reserve J Intl L 391; RL Phillips, ‘2nd Time’s the Charm: France’s Cour de Cassation Broadens Universal Jurisdiction Law’ Just Security (24 May 2023) <www.justsecurity.org/86689/2nd-times-the-charm-frances-cour-de-cassation-broadens-universal-jurisdiction-law/>; J Rémond Tiedrez, ‘France’s Highest Court Confirms Universal Jurisdiction’ EJIL: Talk! (1 June 2023) <www.ejiltalk.org/france-is-back-on-the-universal-jurisdiction-track/>.

[6] Since the risk of impunity is particularly high when serious crimes are committed by State officials or others who are not likely to face prosecution domestically, universal jurisdiction is particularly (but not exclusively) useful in these scenarios. However, universal jurisdiction can be efficient in the fight against impunity in such cases only if one accepts the non-applicability of (at least functional) immunities for international crimes before domestic courts. P Gaeta, ‘Measures for Strengthening the Effectiveness of International Legal Principles Concerning Universal and International Jurisdiction’ in F Haldemann, T Unger (eds), The United Nations Principles to Combat Impunity: A Commentary (OUP 2018) 226.

[7] For an overview of universal jurisdiction cases, see the reports in ‘Universal Jurisdiction Annual Review – UJAR’ TRIAL International <https://trialinternational.org/ resources/ universal-jurisdiction-tools/universal-jurisdiction-annual-review-ujar/>; ‘Universal Jurisdiction Interactive Map’ TRIAL International <https://ujim.trialinternational.org/>. See also ‘Overview of National Jurisprudence’ (Genocide Network Secretariat and Eurojust 2022).

[8] See eg ‘Universal Jurisdiction Interactive Map’ (n 7); ‘Universal Jurisdiction in the European Union. Country Studies’ (REDRESS).

[9] As of 29 May 2024, most of the cases based on universal jurisdiction are regarding crimes perpetrated in Syria. ‘Universal Jurisdiction Interactive Map’ (n 7). Another database, which documents Syrian cases on universal jurisdiction, is ‘MAIN UJ Database’ Syria Justice and Accountability Centre <https://experience.arcgis.com/experience/d33632f4c36146d2b62ec3727 ed8ded6>. See also J Triscone, ‘Universal Jurisdiction, the Only Hope for Prosecuting International Crimes Committed in Syria?’ TRIAL International (6 September 2021) <https://trialinternational.org/latest-post/universal-jurisdiction-the-only-hope-for-prosecuting-international-crimes-committed-in-syria/>. Finally, the podcast Syrian Trials includes references by the victims to universal jurisdiction as their ‘only hope’ for criminal justice. ‘The Syria Trials’ 75 Podcasts <https://75podcasts.org/series/1/>.

[10] For an in depth analysis, see D Hovell, ‘The Authority of Universal Jurisdiction’ (2018) 29 Eur J Intl L 427. To explain why extending criminal laws to extraterritorial conduct challenges sovereignty, Pedretti refers to the potential conflict of jurisdictions, which may arise between the territorial and other States with different heads of jurisdiction. R Pedretti (ed), Immunity of Heads of State and State Officials for International Crimes (Brill/Nijhoff 2014) 344.

[11] For instance, Dutton defines universal jurisdiction as permission to ‘prosecute’ crimes, failing to distinguish between legislative and adjudicative forms of jurisdiction. Dutton (n 5) 5-6.

[12] See eg ‘Tribunals. Universal Jurisdiction’ JusticeInfo.net <www.justiceinfo.net/en/tribunals/ universal-jurisdiction>. The title ‘Trying the people wherever they are, whatever their nationality’ seems to wrongly suggest that universal jurisdiction is about the exercise of jurisdiction. It also gives an impression that for universal jurisdiction, the defendant does not have to be present at trial, whereas the presence at trial depends on the rules of the criminal procedure of the prosecuting State (the latter is an issue totally separate from universal jurisdiction). The following sentence is also an example of a phrase which fails to distinguish between universal jurisdiction as a prescriptive form of jurisdiction and its corresponding form of adjudicative jurisdiction: ‘Universal jurisdiction allows for the investigation and prosecution of serious crimes regardless of where the crimes were committed or the nationality of victims and perpetrators’. W-M Nosakhare, ‘Some Hope in the Struggle for Justice in Syria’ Human Rights Watch’ (15 March 2022) <www.hrw.org/news/2022/ 03/15/some-hope-struggle-justice-syria>.

[13] See in this sense Council of the European Union, ‘The AU-EU Expert Report on the Principle of Universal Jurisdiction’ (2009) 8672/1/09 REV1 7 para 8; J Geneuss, ‘Fostering a Better Understanding of Universal Jurisdiction: A Comment on the AU-EU Expert Report on the Principle of Universal Jurisdiction’ (2009) 7 J Intl Crim Justice 945, 948; Gaeta, ‘Measures for Strengthening the Effectiveness’ (n 6) 230.

[14] R O’Keefe, ‘The Grave Breaches Regime and Universal Jurisdiction’ (2009) 7 J Intl Crim Justice 811, 812.

[15] See on this Gaeta, ‘Measures for Strengthening the Effectiveness’ (n 6) 230-231.

[16] R O’Keefe, ‘Universal Jurisdiction: Clarifying the Basic Concept’ (2004) 2 J Intl Crim Justice 735, 745.

[17] Pedretti (n 10) 340.

[18] See in this sense H Varney, K Zduńczyk, ‘Advancing Global Accountability: The Role of Universal Jurisdiction in Prosecuting International Crimes’ (International Center for Transitional Justice 2020) 8.

[19] A simple way to improve the clarity of the following sentence ‘(…) Syria, victims have turned towards other countries – such as Germany, Sweden, France and The Netherlands – to investigate cases based on what is known as extraterritorial or universal jurisdiction’ would be to reformulate its last part as follows ‘extraterritorial, including universal jurisdiction’. The sentence appears in ‘Q&A on the Dabbagh Case: French Judges Order the Trial of Three Senior Syrian Officials before the Paris Criminal Court’ (International Federation for Human Rights, April 2024) <www.fidh.org/en/issues/ international-justice/universal-jurisdiction/q-a-on-the-dabbagh-case-french-judges-order-the-trial-of-three-senior>.

[20] Kreß highlights that adjudicative jurisdiction covers the activity of both the domestic courts and prosecutorial authorities. Claus Kreß, ‘Universal Jurisdiction over International Crimes and the Institut De Droit International’ (2006) 4 J Intl Crim Justice 561, 564, fn 16. In reality, the meaning of ‘exercise’ and ‘adjudicative jurisdiction’ may vary in domestic criminal systems. For instance, in some systems, ‘exercise’ would include also the investigation phase. However, Gaeta suggests that the extension of a criminal law to cover extraterritorial acts characterised by the absence of link between the State and the concerned act includes the possibility to carry out investigations over the alleged commission of a crime abroad. P Gaeta, ‘The Need Reasonably to Expand National Criminal Jurisdiction over International Crimes’ in A Cassese (ed), Realizing Utopia (OUP 2012) 602.

[21] Lotus (France v Turkey) (Judgment) [1929] PCIJ Rep Series A No 10.

[22] ibid Part III, 18-19.

[23] ibid Part III, 19.

[24] ibid. See eg the Dissenting Opinion by M Loder, 35; Dissenting Opinion by M Weiss, 44-45, 49; Dissenting Opinion by Lord Finlay, 52; Dissenting Opinion by M Nyholm, 60-61; Dissenting Opinion by M Altamira, 95-96.

[25] ‘Universal Jurisdiction over War Crimes’ (n 3) 2.

[26] Just as an example, Germany does not allow for non-written law because this would be considered a violation of the principle of legality. However, by way of contrast, some Anglo-American countries accept it. R Cryer and others, An Introduction to International Criminal Law and Procedure (2nd edn, CUP 2010) 74.

[27] Varney, Zduńczyk (n 18) 11.

[28] For the key facts of the case, see Javor et al v X International Crimes Database Project <www.internationalcrimesdatabase.org/Case/1117/Javor-et-al-v-X/>. Ferdinandusse mentions that Javor et al apparently departed from the previous jurisprudence in France which had considered the Geneva Conventions of 1949 to be self-executing. WN Ferdinandusse, ‘Direct Application of International Criminal Law in National Courts’ PhD Thesis (U of Amsterdam 2005) 76 fn 406.

[29] Even though the case was initiated by a civil complaint, the decision concerned criminal jurisdiction.

[30] France, Tribunal de Grande Instance de Paris, Juge destruction, In re Javor, Ordonnance, 6 May 1994 as cited in Ferdinandusse (n 28) 76, fn 404. The case is available at <https://competenceuniverselle.wordpress.com/wp-content/uploads/2011/07/jugement-tgi-6-mai-1994-javor.pdf>.

[31] France, Court of Cassation, In re Javor (26 March 1996) Bull, crim., no 132 as cited in ibid 76, fn 405. The case is available at <https://competenceuniverselle.wordpress.com/wp-content/uploads/2011/07/cass-26-mars-1996-javor1.pdf>.

[32] ‘(…) que ces dispositions revêtent un caractère trop général pour créer directement des règles de compétence extraterritoriales en matière pénale, lesquelles doivent nécessairement être rédigées de manière détaillée et précise (…)’. ibid.

[33] For the facts and an analysis of this case, see B Stern, ‘In Re Javor. 1996 Bull. Crim., No. 132, at 379 In Re Munyeshyaka. 1998 Bull. Crim., No. 2, at 3’ (1999) 93 AJIL 525, 527-529; I Bantekas, S Nash (eds), International Criminal Law (3rd edn, Routledge-Cavendish 2007) 90-91.

[34] Stern (n 33) 529.

[35] See in this sense Bantekas, Nash (n 33) 88.

[36] For a detailed study on the different legislation and practices regarding the direct application of international treaties and international customary law before domestic courts, see Ferdinandusse (n 28).

[37] See eg the Penal Code of Norway, LOV-2005-05-20-28 Section 5. For an overview of the different national universal jurisdiction frameworks, see ‘Universal Jurisdiction Law and Practice Briefing Papers’ Open Society Justice Initiative and TRIAL International <https://trialinternational.org/universal-jurisdiction-tools/universal-jurisdiction-law-and-practice-briefing-papers/>.

[38] With regard to the beginning of investigations regarding the crimes committed in Syria, see W Kaleck, P Kroker, ‘Syrian Torture Investigations in Germany and Beyond’ (2018) 16 J Intl Crim Justice 165, 173. See also ‘Overview of National Jurisprudence’ (n 7). The jurisdictional basis for the cases mentioned in the latter report is not always clear, but it appears the active personality principle or/and the presence of the alleged perpetrator on the State territory is in many cases the decisive element.

[39] According to Geneuss, however, the State practice is too scarce to conclude that the presence of the offender is one of the requirements for the exercise of universal jurisdiction under customary international law. Geneuss (n 13) 955.

[41] See eg O’Keefe, ‘Universal Jurisdiction’ (n 16) 749-752; Langer, Eason (n 5) 816. Since universal jurisdiction is a prescriptive form of jurisdiction, the reference to in absentia is superficial. This is because in absentia actually refers to the adjudicative or enforcement form of universal jurisdiction. J Crawford, ‘Jurisdictional Competence’ in James Crawford (ed), Brownlie’s principles of public international law (9th edn, OUP 2019) 453. Also, the term universal jurisdiction in absentia creates a risk of confusion with the question of trials in absentia, which are manifestations of States’ own enforcement jurisdiction and are in some States not allowed as they are perceived as a violation of the defendants’ rights. A Korynevych, O Chubinidze, ‘Legal Regulation of Universal Jurisdiction in National Legislation: A Comparative Aspect’ (2019) Actual Problems of Intl Relations 45, 48.

[42] The principle of legality also applies to the modes of liability and circumstances precluding wrongfulness. Claus Kreß, ‘Nulla Poena Nullum Crimen Sine Lege’ in Max Planck Encyclopedia of Public International Law (OUP 2010) 5 para 20.

[43] In this regard, some scholars suggested that the chosen provision should be the one which is the most lenient for the perpetrator. Yet, they have also acknowledged the limitation of this idea, which is the unlikely situation that the judges would actually apply the law of a different country. Gaeta, ‘The Need Reasonably to Expand’ (n 20) 604. A more realistic solution could be to consider all potentially applicable laws to decide on the definition and penalty, which however, still have to fall within the frame provided for the concerned offence in the domestic legislation of the forum State.

[44] A Cassese, P Gaeta (eds), Cassese’s International Criminal Law (3rd edn, OUP 2013) 278.

[45] Forum deprehensionis is the basis in most international treaties. At the same time, it is less common to have a requirement of presence for the whole investigative stage because this would mean that a formal investigation could probably never begin – before enough evidence had been collected, the defendant would already have left. Geneuss (n 13) 956. One country where the presence of the defendant is however required throughout the investigation is the Netherlands. ‘Universal Jurisdiction Law and Practice in the Netherlands’ (Open Society Justice Initiative and TRIAL International 2019) 11.

[46] Gaeta, ‘The Need Reasonably to Expand’ (n 20) 602.

[47] Code of Crimes against International Law of 26 June 2002 (Federal Law Gazette I, 2254), as last amended by Article 1 of the Act of 22 December 2016 (Federal Law Gazette I, 3150) Section I.

[48] Arrest Warrant of 11 April 2000 (Democratic Republic of the Congo v Belgium) (Merits) [2002] ICJ Rep 2000. The majority in this case did not contest the general possibility of trying the defendant based on universal jurisdiction. Yet, they did not address the specific detail, including whether the permission entails the exercise of universal jurisdiction in its absolute or conditional form.

[49] ibid. See eg Joint Separate Opinion of Judges Higgins, Kooijmans and Buergenthal, 80-81 paras 59-60. Cassese stresses that there is gradually increasing acceptance of absolute universal jurisdiction as ‘admissible under international law’. A Cassese, ‘When May Senior State Officials Be Tried for International Crimes? Some Comments on the Congo v. Belgium Case’ (2002) 13 Eur J Intl L 853, 862.

[50] Piracy is at least from the Arrest Warrant of 11 April 2000 case onward the most widely endorsed example of the crime for which there is the permissive universal jurisdiction under custom. See in this regard Arrest Warrant (n 48), specifically Separate Opinion of President Guillaume, 37-38 para 5. Differently, Cassese argues that even piracy is a manifestation of conditional universal jurisdiction. Cassese, ‘When May Senior State Officials Be Tried’ (n 49) 857-858.

[51] F Jessberger, ‘Universal Jurisdiction’ in A Cassese (ed), The Oxford Companion to International Criminal Justice (OUP 2009) 556; Varney, Zduńczyk (n 18) 8. In some States, domestic legislation nonetheless allows for the exercise of universal jurisdiction over ordinary crimes such as kidnapping, rape, and murder. ‘Universal Jurisdiction: The Challenges for Police and Prosecuting Authorities’ (Amnesty International 2007) 3.

[52] Varney, Zduńczyk (n 18) 11.

[53] One concern is that the representativeness of the principle is generally questionable, since the majority of trials on the abovementioned crimes have taken place in the Western sphere. Hovell (n 10) 434.

[54] Along the same lines as the Torture Convention, see eg the International Convention on the Protection of All Persons from Enforced Disappearance; The Ljubljana-The Hague Convention on International Cooperation in the Investigation and Prosecution of the Crime of Genocide, Crimes against Humanity, War Crimes and other International Crimes; and ILC Draft articles on Prevention and Punishment of Crimes against Humanity.

[55] See in this sense Questions Relating to the Obligation to Prosecute or Extradite (Belgium v Senegal) (Merits) [2012] ICJ Rep 2012, 33, para 74; R van Steenberghe, ‘The Obligation to Extradite or Prosecute: Clarifying Its Nature’ (2011) 9 J Intl Crim Justice 1089, 1104-1105.

[56] Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment (adopted 10 December 1984, entered into force 26 June 1987) 1465 UNTS 85. The wording ‘submit to the authorities’ suggests that domestic systems have prosecutorial discretion, as long as the latter exists in the concerned system, in taking the decision on whether they will prosecute. Gaeta, ‘Measures for Strengthening the Effectiveness’ (n 6) 233. On prosecutorial discretion in the Torture Convention, see also R Schmidt, ‘Aut Dedere Aut Judicare’ in M Nowak, M Birk, G Monina (eds), The United Nations Convention against Torture and its Optional Protocol: A Commentary (2nd edn, OUP 2019) 278-280.

[57] Geneva Convention for the Amelioration of the Condition of the Wounded and Sick in Armed Forces in the Field (adopted 12 August 1949, entered into force 21 October 1950) 75 UNTS 31 (Geneva Convention I) art 49. See also the corresponding articles in Geneva Conventions II (art 50), III (art 129) and IV (art 146).

[58] The obligation to bring persons before their own courts is the obligation to prosecute which arguably cannot be subject to prosecutorial discretion if there is sufficient evidence, unless the perpetrator is extradited to another party. Gaeta, ‘Measures for Strengthening the Effectiveness of International Legal Principles Concerning Universal and International Jurisdiction’ (n 6) 234.

[59] For one view, see Pedretti (n 10) 379-380. In the Javor et al case, the court asserted that the obligation to search cannot apply beyond the French border because of the jurisdictional requirement of the presence of suspected persons on the French territory. See (n 31).

[60] Cassese, Gaeta (n 44) 287; ‘Convention (I) for the Amelioration of the Condition of the Wounded and Sick in Armed Forces in the Field, Geneva, 12 August 1949. Commentary of 1949. Article 49 – Penal Sanctions’ ICRC IHL Databases 2866 <https://ihl-databases.icrc.org/ en/ihl-treaties/gci-1949/article-49/commentary/2016#110>.

[61] O’Keefe, ‘The Grave Breaches Regime and Universal Jurisdiction’ (n 14) 817; Pedretti (n 10) 347. The Geneva Conventions nevertheless do not explicitly require that States parties have to adopt any specific head of jurisdiction to bring the alleged perpetrators before domestic courts. Gaeta, ‘The Need Reasonably to Expand’ (n 20) 600, fn 9.

[62] Arrest Warrant (n 48), Dissenting Opinion by Van den Wyngaert, 173-174 para 59; Dissenting Opinion by Bula-Bula, 122 para 65.

[63] ‘Convention (I) for the Amelioration of the Condition of the Wounded and Sick in Armed Forces in the Field. Geneva, 12 August 1949. Commentary of 1949. Article 49 – Penal Sanctions’ (n 61) 2863-2867.

[64] O’Keefe, ‘The Grave Breaches Regime and Universal Jurisdiction’ (n 14) 814.

[65] M Cherif Bassiouni, ‘Universal Jurisdiction for International Crimes: Historical Perspectives and Contemporary Practice’ (2001) 42 Virginia J Intl L 81, 117-118; P Gaeta, ‘War Crimes and Other International “Core” Crimes’ in A Clapham, P Gaeta (eds), The Oxford Handbook of International Law in Armed Conflict (OUP 2014) 739 fn 9.

[66] van Steenberghe (n 55) 1105. As for Garrod, he admits that universality was used in the diplomatic conference during the negotiations of the Geneva Conventions but in the context of the jurisdiction to search and prosecute regardless of the nationality, including those who do not belong to the enemy. M Garrod, ‘Unravelling the Confused Relationship Between Treaty Obligations to Extradite or Prosecute and “Universal Jurisdiction” in the Light of the Habré Case’ (2018) 59 Harvard Human Rights J 125, 185.

[67] Gaeta, ‘Measures for Strengthening the Effectiveness’ (n 6) 232.