1  Officially, the Assembly of States Parties of the International Criminal Court (ICC) adopted a resolution on 14 December 2017 deciding ‘to activate the Court’s jurisdiction over the crime of aggression as of 17 July 2018’ (ICC-ASP/16/Res.5, 14 December 2017 (‘2017 resolution’ or ‘activation resolution’)). Adoption actually took place early the next morning; clocks had been removed and time suspended to permit last-minute manoeuvring. That the decision came down to the wire was appropriate for an issue that it had not been possible to finalize when the Court’s Statute was adopted in Rome in July 1998 (‘Rome Statute’ or ‘Statute’), and which had been the subject of intense disputation in the two decades since.

2  The Rome Statute provided, in Article 5 (1), that the crime of aggression is one of the four crimes over which the Court ‘has jurisdiction’, along with the crime of genocide, crimes against humanity, and war crimes. But the ‘exercise’ of aggression jurisdiction was postponed. Article 5 (2) provided that:

3  Article 5 (2) foresaw two future tasks, ‘defining the crime’, and ‘setting out the conditions’ for the ‘exercise’ of jurisdiction over it. ‘Defining’ was necessary because earlier efforts at making the crime concrete, notably the Charters of the Nuremberg and Tokyo Tribunals (International Military Tribunals), had been rudimentary at best. While most participants in Rome were of the opinion that the crime existed under international customary law, there was significant disagreement about its parameters. The principle of legality (Nulla poena nullum crimen sine lege) dictated that definition of its substantive, criminal law elements was fundamental to prosecuting the crime under the Statute.

4  ‘Conditions’ was itself ambiguous. It was ultimately required to carry significant freight concerning how much the procedural structure for the crime of aggression could differ from the structure the Statute applied to the other three crimes. This included, in particular, the role of the United Nations Security Council (‘Security Council’; United Nations, Security Council) in relation to this crime. It also implicated the (non-)application of the Statute to the nationals of States that are not party to the Statute, and the extent to which it would even be applicable to all States Parties to the Statute.

5  Article 5 (1) is about the subject-matter of the Court’s jurisdiction, the four crimes over which it has competence. Article 5 (2)’s words ‘exercise of jurisdiction’ relate, inter alia, to other aspects of jurisdictional competence than subject-matter jurisdiction.

6  First, they echo Article 13 of the Statute, headed precisely, ‘Exercise of jurisdiction’. Article 13 asserts that the Court may exercise its jurisdiction with respect to a crime referred to in Article 5 if, (a) a situation in which one or more such crimes appears to have been committed has been referred to the Prosecutor by a State Party (Referral by a State party: International Criminal Court (ICC)), (b) a situation in which one or more such crimes appears to have been committed is referred to the Prosecutor by the Security Council acting under Chapter VII of the Charter of the United Nations (‘UN Charter’ or ‘Charter’; United Nations Charter) (Referral by the United Nations Security Council: International Criminal Court (ICC)), or (c) the Prosecutor has exercised his or her power under Article 15 (1) of the Statute to ‘initiate investigations proprio motu on the basis of information on crimes within the jurisdiction of the Court’ (Proprio motu investigation: International Criminal Court (ICC)). These three entry points are called collectively the ‘trigger mechanism’, denoting the means by which the Court becomes seized of particular situations.

7  Second, the phrase ‘exercise of jurisdiction’ in Article 5 echoes Article 12 of the Statute, titled ‘Preconditions to the exercise of jurisdiction’. Article 12 (1) provides that a State which becomes a party to the Statute thereby accepts the jurisdiction of the Court with respect to the four crimes referred to in Article 5. Article 12 (2) provides that, in the case of referrals to the Prosecutor by a State Party, or in the case where the Prosecutor exercises his or her proprio motu power to commence proceedings, the Court may exercise its jurisdiction if one or more relevant States have accepted the jurisdiction of the Court. These are ‘(a) the State on the territory of which the conduct in question occurred or, if the crime was committed on board a vessel or aircraft, the State of registration of that vessel or aircraft’, or ‘(b) the State of which the person accused is a national’. It is striking that, in the case of a referral by the Security Council (Art 13 (b) Rome Statute), there is no requirement that the territorial State or the State of nationality of the accused be a party to the Statute. Article 13 (b) is a form of universal jurisdiction.

8  Article 12 is thus about jurisdiction over persons, jurisdiction ratione personae—in respect of whom may the Court take adjudicative power? The Court has jurisdiction only over natural persons (Art 25 (1) Rome Statute) but, in the absence of a Security Council referral, those persons must fit one of the categories delineated in Article 12. They must act within the territorial jurisdiction of a State Party, or be a national of a State Party to the Statute.

9  As we shall see below (paras 27–31, 41, 44–45), the Statute also speaks to ‘exercise of jurisdiction’ in one of its amendment provisions, Article 121 (5), which likewise refers to jurisdiction over persons.

10  ‘Exercise of jurisdiction’ in Article 5 (2) Rome Statute, then, implied that attention needed to be paid, in negotiating a ‘provision’ on aggression, to the trigger mechanism, to jurisdiction ratione personae, and perhaps to amendment modalities.

11  To complete the background from Rome: Resolution F of the Final Act of the Rome Conference established a Preparatory Commission to facilitate the entry into force of the Statute. Among its tasks was to ‘prepare proposals for a provision on aggression, including the definition and Elements of Crimes of Aggression and the conditions under which the International Criminal Court shall exercise its jurisdiction with respect to this crime’ (Resolution F). The Resolution contemplated submission of the ‘provision’ to a Review Conference to be convened seven years after the Statute came into force. Aggression was, indeed, the main business of the Review Conference held in Kampala, Uganda, in 2010. ‘A provision’ in Article 5 (2) and Resolution F was an understatement of the task, as the Amendments required new articles to be added to the Statute, alterations to others, a set of Elements of Crimes (‘Elements’), and a group of Understandings (Crime of aggression, RC/Res.6; and, generally, Barriga and Kress, 2012, passim).

12  The Preparatory Commission was unable to complete the task. When the Rome Statute came into force in 2002, the Assembly of States Parties (‘ASP’) immediately created a Special Working Group on the Crime of Aggression (‘SWGCA’), open to all States, not only parties to the Statute. The SWGCA did the bulk of its drafting work in intersessional meetings at the Liechtenstein Institute for Self-Determination at Princeton University in New Jersey (Barriga, Danspeckgruber, and Wenaweser, 2009). This process produced a draft for consideration in Kampala. The SWGCA’s definitional draft, which became Article 8 bis of the Statute, was adopted verbatim in Kampala (RC/Res.6). Its draft on ‘conditions’ for exercise of jurisdiction was, however, the subject there of intense debate and amendment, entailing a split into two amending articles, Articles 15 bis and 15 ter.

13  As this Encyclopedia is concerned with procedural law, this entry is not the place for a detailed analysis of the definition. Nevertheless, a brief exposition of the substantive aspects of the crime will help to place the jurisdictional issues in context. Article 8 bis of the Statute proceeds from a drafting convention distinguishing between the ‘crime of aggression’, what a natural person commits (para 1), and the ‘act of aggression’, which is what a State does (para 2). ‘Act of aggression’ (para 2) deals with the use of armed force by a State against another State. It references the standards of the UN Charter and adds a list of concrete examples. The list tracks United Nations General Assembly (‘General Assembly’; United Nations, General Assembly) Resolution 3314 of 1974 (‘Definition of Aggression’). ‘Crime of aggression’ entails certain actions, namely planning, preparation, initiation, or execution of an ‘act of aggression’ (as defined). Those actions must be taken by a (natural) person (or persons) in a position effectively to exercise control over or to direct the political or military action of a State. Aggression is said to be a leadership crime and this control or direction element reflects that. The most controversial part of the definition of the ‘crime’ is one further limiting the scope of the term ‘act of aggression’. The act must be one that by its ‘character, gravity and scale, constitutes a manifest violation of the Charter of the United Nations’. ‘Manifest’ is, as the Elements point out, an objective qualification. It serves two purposes. Firstly, it makes clear that a minor breach of the Charter, limited firing across a border for example, is not enough for the crime—there may well be State responsibility, but the criminal threshold is higher. Secondly, it seeks to deal with the ‘grey area’ where use of force may be protected by a justification such as humanitarian intervention. Such justifications are hotly contested under general international law and there was no possibility that the issue could be resolved by a negotiation under the auspices of the ICC. It must, perforce, be left to the judges (and prosecutorial discretion) in individual cases.

14  I now turn to the ‘conditions’ for the exercise of jurisdiction.

15  Much discussion following Rome involved the second sentence of Article 5 (2) of the Statute: ‘[s]uch a provision shall be consistent with the relevant provisions of the Charter of the United Nations’. Most participants accepted that these cryptic words related, at least in part, to the relationship between the ICC and the Security Council. The language of Article 13 (b) of the Statute, relating to referrals of situations by the Security Council, made no distinction between the four offences, so it was reasonable to assume that the Security Council could make referrals concerning the crime of aggression, including cases involving nationals of States that were not party to the Rome Statute. It also appeared that there was no reason why Article 16 of the Statute, relating to deferral of investigation or prosecution at the request of the Security Council (Deferral requested by the United Nations Security Council: International Criminal Court (ICC)), should not apply to the crime of aggression. Article 16 provides that:

16  Although there was little discussion on this, it was generally accepted that the principle of complementarity applies to the crime of aggression. That is to say, under Article 17 of the Statute, a case is ‘inadmissible’ in the Court when it is being investigated or prosecuted by a State which has jurisdiction over it, unless the State is unwilling or unable genuinely to carry out the investigation or prosecution. The extent to which, under international law, a victim State or a State claiming universal jurisdiction is entitled to prosecute for the crime of aggression is debated. Suffice it to note here, the understanding that, as is the case for other crimes within the jurisdiction of the ICC, that jurisdiction is secondary to that of an appropriate State jurisdiction (Clark, 2011).

17  Beyond those propositions, however, there were some fundamental disagreements within the SWGCA. In particular, the Permanent Members of the Security Council insisted that the Security Council’s power under Article 39 UN Charter to ‘determine the existence of any threat to the peace, breach of the peace or act of aggression’ was exclusive. Accordingly, claims of aggression could reach the Court only through the Security Council trigger; and there could be no act of aggression for purposes of the crime of aggression unless the Security Council so determined. Other States tended to rely on Article 24 UN Charter, which affords the Council ‘primary responsibility for the maintenance of international peace and security’. ‘Primary responsibility’ permits a role for other actors, they argued, including the ICC. They added that the General Assembly has, in practice, made several findings of aggression, that the United States, the United Kingdom, and France were co-sponsors of the Uniting for Peace Resolution (1950) (UNGA Res 377) which recognizes the Assembly’s powers, and that all five Permanent Members have voted pursuant to that resolution when it suited them. Non-permanent members of the Council tended to add that the International Court of Justice (ICJ) has also addressed issues where aggression is in play. The draft sent to Kampala included the General Assembly and the ICJ as alternative ‘filters’ to the Security Council for the crime of aggression, but these were deleted in Kampala.

18  The major achievement in respect of this Security Council part of the negotiation in the period of the SWGCA was to de-couple the ‘definition’ from the ‘conditions’. In the version of the definition and conditions that was on the table at the last meeting of the Preparatory Commission, the Security Council (or possibly the General Assembly or the ICJ) would make a definitive decision on the existence of the element of ‘act of aggression’, which was binding on the ICC (Clark, 2002, 869–72). The ICC would itself decide the other elements of the crime. This would, however, subvert the power of the Court to decide itself the existence or otherwise of all the elements of the crime. It would also make it extremely difficult to build a criminal offence around a structure where a foundational element was decided elsewhere than the Court, and potentially based on totally political considerations. In such circumstances, there would probably be unbearable weight placed on the mental element provisions of Article 30 of the Statute, the mistake provisions of Article 32 of the Statute, or on what came to be the ‘manifest’ threshold. The Security Council’s power to decide the ‘act of aggression’ element was removed in the SWGCA’s draft and remained out in Kampala. Removing the connection between the Security Council and substance opened the way for focusing on the various procedural options put before the Review Conference of giving the Security Council (or other United Nations organ) a ‘filter’ role, providing either a ‘green light’ (permission to go forward) or a ‘red light’ (denial of right to go forward) to the ICC’s proceedings. Moreover, a solid group of States insisted that the Court should be able to proceed even in the absence of action by someone else.

19  Resolution of divergent positions on the role of the Security Council was facilitated in Kampala by dividing the SWGCA’s draft Article 15 bis on ‘conditions’ into two parts, one dealing with State referrals and referrals made by the Prosecutor proprio motu, and the other dealing with Security Council referrals. These became, respectively, Articles 15 bis and 15 ter of the Statute (RC/Res.6). Article 15 ter referrals, those by the Security Council, are the more straightforward to describe and it is helpful to discuss them first.

20  Article 15 ter (1) is the basic provision authorizing the Court to exercise its jurisdiction in respect of the crime of aggression, upon a Security Council referral: ‘The Court may exercise jurisdiction over the crime of aggression in accordance with Article 13, paragraph (b), subject to the provisions of this Article’. Then follow two paragraphs which provide a set of hurdles and a time-frame for para 1 to come into play. Article 15 ter (2) says that the Court ‘may exercise jurisdiction only with respect to crimes of aggression committed one year after the ratification or acceptance of amendments by thirty States Parties’. Thirty ratifications were received by the end of June 2016. Article 15 ter (3) adds a cumulative requirement for activation, that the Court ‘shall exercise jurisdiction over the crime of aggression in accordance with this Article, subject to a decision to be taken after 1 January 2017 by the same majority of States Parties as is required for the adoption of an amendment to the Statute’. The required majority at this later date in the event of a vote would thus be two-thirds of all the States Parties at the relevant time. In fact, in accordance with the ASP’s normal practice, the decision in December 2017 was made by consensus (although a contested one, as we shall see later in paras 44–45 below). Exercise of jurisdiction is thus prospective once the two conditions are satisfied. In practice, this means that there is jurisdiction only over events occurring after 17 July 2018, the date referred to in the 2017 ASP resolution.

21  Article 15 ter (4) adds the important principle that ‘[a] determination of an act of aggression by an organ outside the Court shall be without prejudice to the Court’s own findings under this Statute’. Including this language in the Article dealing with Security Council referrals underscores the way the negotiation developed towards making the Court master of its own decisions in respect of the elements of a particular (alleged) crime of aggression. A Security Council determination, for or against aggression, if there is one, may have some evidentiary weight, but is not determinative in the criminal proceedings.

22  One final point about Security Council referrals does not feature on the face of Article 15 ter, but bears emphasis. It will be recalled that under the combined effect of Articles 12 and 13 (b) of the Statute, the Security Council may refer a situation to the Prosecutor regardless of whether the territorial state or state of nationality is a party to the Statute. Article 15 ter (1) keeps this in place for aggression. But how does this play out for a State that is a party to the Statute but has not ratified Kampala? Note that the second sentence of Article 121 (5) dealing with amendments to Articles 5, 6, 7, and 8 of the Statute, provides that: ‘[i]n respect of a State Party which has not accepted the amendment, the Court shall not exercise its jurisdiction regarding a crime covered by the amendment when committed by that State Party’s nationals or on its territory’. This language is arguably broad enough to protect a Party to the Statute which does not accept the Kampala Amendments from a Security Council referral. The drafters of the Kampala Amendments, however, did not accept that a State Party to the Rome Statute could shield itself from a Security Council referral in a way a non-Party could not. Paragraph 2 of the Understandings adopted in Kampala reads as follows:

The wording is awkward. All States Parties have already ‘accepted the Court’s jurisdiction’ over aggression pursuant to Article 12 (1) of the Statute: ‘[a] Party to this Statute thereby accepts the jurisdiction of the Court with respect to the crimes referred to in article 5’. The issue is surely whether it has agreed to the exercise of the jurisdiction now that the appropriate ‘provision’ has been accepted. The Understanding would have been clearer if it had read: ‘irrespective of whether the State concerned has accepted the exercise of the Court’s jurisdiction in this regard’. But the import of the Understanding is apparent. The Security Council may refer to the Court situations involving State Parties to the Statute which have not accepted the Kampala Amendments. The legal status of the Understandings has been questioned (Heller, 2012), but one suspects that the Court would regard what has just been suggested as a reasonable interpretation of some clumsy drafting.

23  So much for Security Council referrals. Article 15 bis, as finally adopted, deals with the exercise of jurisdiction over the crime of aggression following State referrals and referrals by the Prosecutor proprio motu. The Court is authorized to exercise jurisdiction over the crime of aggression in accordance with Article 13 (a) and (c), subject to the other provisions of the Article (Art 15 bis (1)). Once again, there is the requirement of ratification or acceptance by 30 States Parties, the passage of a year after that, and the further vote after 1 January 2017 (Art 15 bis (2) and (3)). Then follows a crucial authorization/opt-out provision (Art 15 bis (4)) reading: ‘The Court may, in accordance with article 12, exercise jurisdiction over a crime of aggression, arising from an act of aggression committed by a State Party, unless that State Party has previously declared that it does not accept such jurisdiction by lodging a declaration with the Registrar. The withdrawal of such a declaration may be effected at any time and shall be considered by the State Party within three years.’ The careful reader will have noted the language: ‘exercise jurisdiction over a crime of aggression, arising from an act of aggression committed by a State Party’. ‘State Party’ must mean State Party to the Rome Statute. There is no suggestion here that the Court’s jurisdiction is limited to those States ratifying the amendment. A State which has not done so can, in the plain language of the amendment, protect its people from the jurisdiction by utilizing the opt-out provision. Indeed, the opt-out language, on its face, seems to be coherent only on the possibility that any State Party may want to opt out! On this interpretation, the requirement of 30 ratifications, like the seven-year delay, is to be understood as among the ‘conditions’ of activation. As we shall see (paras 44–45 below), this plain meaning is not what the 2017 activation resolution adopted.

24  Article 15 bis (5) addresses the non-State Party problem. It was of particular significance for the three Permanent Members of the Security Council who have not become party to the Rome Statute—China, the Russian Federation, and the United States—and for other non-parties who are wont to use force outside their own territories. It provides that ‘[i]n respect of a State that is not a party to this Statute, the Court shall not exercise jurisdiction over the crime of aggression when committed by that State’s nationals or on its territory’. In the negotiations leading up to Kampala, there was widespread support for accepting that where an aggression occurs against a State Party to the Statute, the Article 12 precondition of ratification by the State of territoriality should be sufficient for the Court’s jurisdiction. Article 12 requires that either the State of territoriality or the State of nationality be a party. An aggression, so the argument goes, can, as a matter of territoriality, take place both in the State where the aggression is plotted, and in the place where it is executed (the ‘victim State’). This is in accordance with the normal rules on ‘effects’ or ‘objective territorial’ jurisdiction and is the case with genocide, crimes against humanity, and war crimes. Thus, a citizen of a non-State Party who commits genocide, war crimes, or crimes against humanity on the territory (or having effect on the territory) of a State Party is subject to ICC jurisdiction. The present provision is aimed at upsetting the implication of jurisdiction on the basis of events occurring on the territorial/victim State, specifically in respect of aggression. There is no jurisdiction over aggression in such cases, in the absence of a Security Council referral.

25  Article 15 bis (6), (7), and (8) resolve the various Security Council ‘red light’ and ‘green light’ options concerning State and proprio motu referrals that had been considered intensively but inconclusively before Kampala. Where the Prosecutor concludes that there is a reasonable basis to proceed with an investigation in respect of a crime of aggression, proprio motu or following a State referral, he or she is required to first ascertain whether the Security Council has made a determination of an act of aggression committed by the State concerned. The Prosecutor is to notify the Secretary-General of the United Nations of the situation before the Court, including any relevant information and documents. If, in fact, the Security Council has made a determination of an act of aggression, the Prosecutor may proceed with the investigation. Then comes the crunch: what if the Security Council has not acted, and does not now act? The consensus in Kampala represented a strong resolution of the matter. The relevant language reads:

The ‘filter’ in the case of silence from the Security Council is the Pre-Trial Division, that is to say, a majority of all six members of that Division sitting together en banc. If the Security Council wishes to enter the fray without making a substantive decision, it may do so by putting up its Article 16 stop-light. But notice that, consistent with the existing Rome compromise contained in Article 16, a dissenting member of the Permanent Five members of the Security Council cannot stop the process simply by exercising a veto. It is only where the five are agreed (and obtain the other necessary votes) that proceedings may be halted.

26  A fundamental issue remained: what is the jurisdictional situation of a State Party to the Statute which has not ratified the Kampala amendments? Can prosecutions be brought against its nationals for actions taken on the territory of another State Party, or for actions taking place on its territory, on the basis of a State referral or the Prosecutor acting proprio motu? Does Article 15 bis (4) mean what it says?

27  The seeds of this problem arose from the compromise in Rome when aggression was included in the Statute, but subject to further action, an ‘amendment’ of some sort. Article 5, the inclusion clause of the Statute, has already been discussed. Article 5 (2) referenced Article 121, on amendments, which provided, in relevant part:

28  Rome’s preparatory work on how Articles 5 and 121 fit together on aggression is only partially illuminating. The ultimate formulation was foreshadowed in a proposal from the NonAligned Countries (Amendments Submitted by the Movement of Non-Aligned Countries to the Bureau Proposal, UN Doc A/CONF.183/C.1/L.75), introduced late in the Rome Conference, on 14 July 1998. It read:

29  Note that the proposal would suspend ‘exercise’ of ‘jurisdiction’ until a ‘definition’ had been ‘adopted’. It did not use the word ‘amendment’, and the words ‘in accordance with the Statute’ in the third sentence are, at best, vague.

30  Meanwhile, the Coordinator of the negotiations on the Final Clauses had forwarded his final document to the Committee of the Whole on 11 July (Report of Coordinator on Preamble and Final Clauses, 11 July 1998). It had not been possible to resolve all the issues and some of the material was bracketed. Thus, what became Article 121 (4) was essentially in its final shape, except that the Coordinator had bracketed five-sixths and seven-eighths as options for bringing an amendment into force. The Bureau of the Committee of the Whole apparently settled for seven-eighths at the last minute, a minority position supported by some of the major powers. What is now Article 121 (5) was controversial and emerged in the Coordinator’s final document in a manner that indicated deadlock:

[Any amendment to article 5 shall enter into force for those States Parties which have accepted the amendment one year after the deposit of their instruments of ratification or acceptance [, unless the Assembly or the Conference has decided that the amendment shall come into force for all States Parties once it has been accepted by [5/6] [7/8] of them.]]

The brackets around the whole paragraph recognized that many participants, perhaps most, did not want to distinguish between different kinds of amendments. Article 121 (4), on that view, would apply to all except those ‘technical’ amendments covered by the simplified procedure in Article 122 (2). The internal brackets within draft Article 121 (5) reflected a fallback position of leaving it to the Assembly to decide whether or not a particular amendment would apply to all States Parties. The Coordinator’s draft had been negotiated vigorously against the backdrop of the debate raging throughout the Diplomatic Conference on the inclusion or exclusion of aggression and of nuclear weapons and other weapons of mass destruction in the Statute (nuclear and other weapons of mass destruction were not ultimately included). Those who wanted them in, even if later, sought more flexible amendment procedures; those opposed wanted to close the door as firmly as possible. No one negotiating the Final Clauses could know at that point, less than a week before the end of the Diplomatic Conference, what would happen with these items. Since the Non-Aligned compromise on aggression had not yet been presented, it is not surprising that the group working on the Final Clauses did not consider it.

31  No further discussion of the details of the negotiations around Articles 5 and 121 appears on the public record. The crucial last-minute decisions by the Bureau of the Committee of the Whole on these provisions were not accompanied by any explanation. Nor were they passed on by the Drafting Committee, certainly not formally; nor were they referred back to the Coordinator of the Final Clauses. What is plain, is that Article 5 (2) and Article 121 (4) emerged in their final form; in Article 121 (5), the first part of the Coordinator’s draft was retained, but the second part of it (giving the ASP an option to make an amendment applicable erga omnes partes) was removed. Moreover, and this was a bolt from the blue, a second sentence was added to Article 121 (5). It read: ‘In respect of a State Party which has not accepted the amendment, the Court shall not exercise its jurisdiction regarding a crime covered by the amendment when committed by that State Party’s nationals or on its territory.’ This sentence was not negotiated among the vast majority of States participating in the Conference. It appears to have been negotiated, without transparency, by the Chair of the Committee of the Whole and senior British, French, and perhaps United States delegates. Most participants did not see it until the final morning of the Conference, 17 July, and had no chance to debate its meaning or to alter it. It was, however, to take on considerable significance during the work of the SWGCA, in Kampala, and in the final efforts at activation in 2017. Notably, it arguably contemplated the possibility that criminal proscriptions adopted by amendment (including aggression?) might not necessarily be applicable to all parties to the Statute.

32  Indeed, the application of the amendment provisions referenced in Article 5 to the task of negotiating a provision on aggression, namely Articles 121 and 123, was far from straightforward. Article 121 dealt with amendments in general and the role of the Assembly of States Parties in the amendment process. Article 123 contemplated that a Review Conference to be convened seven years after the entry into force of the Statute to ‘consider any amendments’ to the Statute. It would function the same way as the Assembly under Article 121.

33  There was significant, but inconclusive, discussion of how Article 121 applied to the crime of aggression in the SWGCA, particularly in its 2005 Informal Inter-Sessional meeting (Barriga, Danspeckgruber, and Wenaweser, 2009, 167–69). By now it was clear that there were at least three possibilities for interpreting the relevant provisions. At the back of the debate was the nagging political question of whether the crime could be applicable to all parties to the Statute or whether, and how, it would be possible for some of them to avoid being caught in the net. Could aggression be made subject to a special ‘consent-based’ regime, different from that applicable to the other crimes?

34  The first, all-inclusive, possibility was that once the parties were agreed upon the relevant ‘provision’, it would be effective without more, pursuant to Article 121 (3). Bear in mind Article 5 (2)’s words ‘is adopted’. It is apparent that the word ‘adoption’, as used in Article 121 (3), is used, at least in part, as it is typically used in modern treaty practice, to speak of agreement on a text, which is then sent to capitals for a decision on whether to ratify. But Article 5 (2) says nothing about ratification or acceptance; it merely refers to adoption. Could adoption be all that was required? This is the literal meaning of the two articles read together. And Article 12 asserts that States Parties have already accepted jurisdiction over all four crimes in Article 5. Only a definition and conditions are now required. On the other hand, should we conclude that the secretive drafters of Article 5 (2) were using the word ‘adoption’, albeit carelessly, to include something more than agreement on a text, namely whatever is required by Article 121 (4) or (5), as the case may be? Some participants argued functionally that this must be so, since there are special problems about how a State faced with a significant decision like the present one can cope with the necessary changes in domestic law or other Parliamentary approval without going through the ratification process. There was an answer to this. States could negotiate a delay of a year or two in the application of the provision on aggression, in order to pave the way for domestic action. The reference to ‘conditions’ in Article 5 (2) made that possible. At all events, Article 121 (3)’s adoption as the sole solution gained only minimal traction (none of the States claiming a ‘parliament problem’ made use of the seven year delay after Kampala to solve the problem).

35  The next possibility was that Article 121 (3)’s ‘adoption’ had to be accompanied by the procedure of Article 121 (4), which provides that an amendment comes into force for all parties one year after instruments of ratification or acceptance have been deposited by seven-eighths of them.

36  At least two categories of States found ways to argue for the inapplicability of Article 121 (4) and thus in favour of Article 121 (5) combined with Article 121 (3).

37  One group was utterly opposed to being required by majority vote to subject themselves to the crime, even if the majority was a large one of seven-eighths. For them, Article 121 (5) was the only way to effect amendments to the details of the definitions of the crimes. They were not persuaded by the sui generis argument that, while Article 121 (5) made sense in relation to the addition of new crimes to those already in the Statute, adding terrorism and drug offences, for example, aggression was already ‘in’ the Statute. It was the details that now had to be decided. Nor would they give the time of day to the argument some made that the defining exercise was not an ‘amendment’ to Article 5; it was a fulfilment of what it facilitated, a completion. It might be an ‘amendment’ to the Statute as a whole, but it was not an amendment to Article 5 as such. (It was certainly not an amendment to Articles 6, 7, or 8.)

38  The other group supporting Article 121 (5) included those who wanted the aggression amendments to proceed but despaired that the seven-eighths could ever be achieved. They could live with looser ratification requirements in return for exercise of jurisdiction by the Security Council and at least some parties.

39  The differences between Article 121 (4) and Article 121 (5), in the absence of further hurdles in the conditions, were stark. Under Article 121 (4), no-one was bound until seven-eighths agreed; then everyone was. Under Article 121 (5), each State could be bound upon ratification but it would never be possible for aggression to be applicable erga omnes to non-ratifiers.

40  To many, Article 15 bis (4) of the Kampala Amendments seemed quite straightforward on its face as a compromise that would result (following the thirty ratifications and the subsequent decision of the ASP) in all States Parties being bound unless those wishing to do so specifically opted out. It will be recalled that it provides that

State Party here must mean ‘party to the Rome Statute’. The assumption here, especially given the reference to Article 12, is arguably that all States Parties are now bound to accept the exercise of jurisdiction unless they opt out. After all, all States Parties have accepted the ‘jurisdiction’ of the Court over all four of the Court’s crimes pursuant to Article 12 (1). Now the ‘exercise’ of that jurisdiction has been freed up. But there is an escape hatch, namely the opportunity to opt out. The general rule in Article 120 is that no reservations may be made to the Statute. Nevertheless, the so-called ‘Transitional Provision’, Article 124, does provide for an exceptional opt-out that is effectively a limited right to make a reservation. A State, on becoming party to the Statute, may declare that, for a period of seven years after the entry into force of the Statute for the State concerned, it does not accept the jurisdiction of the court over the war crimes contained in Article 8 of the Statute ‘when the crime is alleged to have been committed by its nationals or on its territory’. Article 124, applicable on its face only to war crimes and not to genocide or crimes against humanity, was added in the last stages of the Rome Conference, apparently by much the same group that gave us the final version of Article 121 (5). The similarity in language is noticeable. It also provided some sort of model for Article 15 bis (4). Paragraph 1 of the adopting resolution in Kampala (Resolution RC/Res.6) perhaps explains something further about the opt out provision when it ‘notes that any State Party may lodge a declaration referred to in article 15 bis prior to ratification or acceptance’. ‘Previously’ in Article 15 bis (4) must mean ‘prior to the commission of the relevant act of aggression’, so reading Article 15 bis (4) along with the explanation in the Resolution, it would seem that a State Party could opt out even without ratifying, in advance of ratifying, or after ratifying but before committing the relevant ‘act’. The proposition that all States Parties are bound, subject to opting out, is consistent not only with the plain meaning of Article 15 bis (4), but also with the preparatory work of the provision and the compromise adopted in Kampala, as explained by the main drafters of the compromise (Report on the facilitation, Annex 2, Paper submitted by Liechtenstein). And this was a consent regime—consent was already given in Article 12 but could now be withdrawn by opting out.

41  The waters were, however, muddied by language in the first part of paragraph 1 of the Kampala adopting resolution. It read that the Review Conference ‘[d]ecides to adopt, in accordance with article 5, paragraph 2 of the Rome Statute the amendments to the Statute contained in annex I of the present resolution, which are subject to ratification and shall enter into force in accordance with article 121, paragraph 5’. Could this mean that, whatever is to be made of Article 15 bis (4), including the opt-out provision, nobody was ‘in’ until individual ratification? Did the reference to Article 121 (5) include a reference to the second sentence which had found its way into paragraph 5 in the final hours in Rome? Or did it, in the particular circumstances of a crime half-in and half-out of the Statute, distinguish between ‘entry into force’ and ‘exercise of jurisdiction’ in a manner that meant that the second sentence was irrelevant in the present context? Certainly some had thought so in the SWGCA where proposals had been introduced supporting (as alternatives) both possibilities for the second sentence. But these had been deleted in Kampala once the opt-out compromise was agreed upon, suggesting that the second sentence of Article 121 (5) was simply irrelevant.

42  At its 2016 meeting the ASP decided to

An able Austrian diplomat, Ms Nadia Kalb, was appointed facilitator and organized several well-attended meetings between March and November 2017. The facilitator’s report (ICC Assembly of States Parties, ‘Report on the facilitation’, ICC-ASP/16/24) made it clear that, while there was widespread support for activation, and an agreement that the Kampala amendments should not be renegotiated, ‘[a] divergence of views continued to exist only with regard to the Court’s exercise of jurisdiction over crimes of aggression committed by nationals or on the territory of States Parties which have not ratified the amendments’ (Report on facilitation para 31). It is important to note that there was no attempt to re-open the definitional provisions or the solution of the Security Council role.

43  Three documents containing ‘elements of an activation decision’ presented in the course of the facilitation echoed the differences. A proposal presented by France and the United Kingdom reiterated both sentences of Article 121 (5) and insisted that the Amendments applied only to the nationals or on the territory of States that had accepted them (Report on facilitation, Annex III A). The State of Palestine introduced an element ‘recalling’ that Article 15 bis (4) meant that the Amendment applies to all States Parties unless they opt out (Report on Facilitation, Annex III B). Switzerland, assuming the same interpretation, pushed for a ‘simple’ activation resolution, in which the ASP would merely decide to activate the jurisdiction (Report on Facilitation, Annex III C).

44  As the crucial 2017 ASP session began in New York, over 30 delegations, coordinated by Switzerland, supported a ‘simple activation’ (Switzerland et al, ‘Letter to All Permanent Representatives of States Parties to the Rome Statute’). An effort was also under way to find a bridge between them and the Parties in favour of what was now known as the ‘restrictive position’. Such an approach had first been suggested during the facilitation by Professors Kress (Cologne) and Akande (Oxford). The bridge would protect the position of those in favour of the restrictive position. A statement to that effect to the Registrar would be treated as a declaration under Article 15 bis (4), in the event that the Court decided for the more permissive position. Brazil, Portugal, and New Zealand presented to the ASP slightly more complicated versions of this, which would permit States to be treated as having exercised an opt-out option without having to say so explicitly. By the last day of the session, it was clear that the British and French ambassadors, who made repeated references to the second sentence of Article 121 (5), were intransigent in respect of bridge-building. It was the second sentence of Article 121 (5) or nothing. Their position was totally transparent, unlike the end-game in Rome. No face-saver was possible. The facilitator acknowledged the lack of common ground and passed on the task of making one last effort to the two vice-presidents of the session who, in the absence of the president, chaired the final meetings.

45  The vice-presidents’ proposal, offered as a take-it-or-leave-it package and ultimately adopted, contained four operative paragraphs. The first contained the decision to activate. The second adopted the British/French position whereby the Assembly:

An aggressor State can protect its citizens from the Court simply by not ratifying, and, if it does ratify, it can, for good measure, opt out in the future. The third paragraph gave a glimmer of hope to those who believed (correctly in the author’s view) the second paragraph to be a mis-reading of the Statute and the Kampala compromise. In it, the Assembly ‘[r]eaffirms paragraph 1 of article 40 and paragraph 1 of article 119 of the Rome Statute in relation to the judicial independence of the judges of the Court’. Article 40 of the Statute asserts the basic independence of the judges. Article 119 (1) of the Statute insists that ‘[a]ny dispute concerning the judicial functions of the Court shall be settled by the decision of the Court’. ‘Judicial functions’ must include any issue as to jurisdiction ratione personae, so it is just conceivable that, relying on paragraph 3, the Court might adopt the more open position and exercise its competence over a citizen of a non-ratifying State Party. Britain and France made an effort to have the effect of this language softened by moving it to the preamble of the resolution, but ultimately joined the consensus. The fourth paragraph of the activation resolution encourages all States Parties which have not yet done so to ratify or accept the Kampala Amendments (ICC-ASP/16/Res.5, para 4).

46  The 2017 resolution renders the opt-out something of a dead letter. State practice concerning it is thin, one instrument before and one since the 2017 resolution. Kenya deposited a Declaration of Non-Acceptance on 30 November 2015 and Guatemala on 16 January 2018. Kenya’s discursive instrument seems (bizarrely) to dispute the validity of the Kampala Amendments under the Kenyan Constitution and international law. Guatemala’s begins by asserting that it will not accept the jurisdiction ‘until such time as the Guatemala State has completed its internal process of ratification of the Kampala amendments’, which seems like a re-assertion of paragraph 2 of the activation resolution. It then seems to turn to the opposite tack with these words: ‘and, pursuant to article 15bis (4) of the Statute, hereby submits the present declaration of non-acceptance of the jurisdiction of the International Criminal Court in respect of the crime of aggression’. Perhaps made out of an abundance of caution, this statement, however, looks like an acceptance that Article 15 bis (4) is governing and that an opt-out is being made to protect its position pending completion of its domestic processes.

47  The end-result was a disappointment to some of those who had already ratified the Kampala Amendments on the understanding that they thereby gained some protection from those acting on behalf of other States Parties to the Statute which had not opted out. Their remaining comfort is that since 17 July 2018 it is open to the Security Council to refer situations concerning all States and that the criminal prohibition of aggression regime applies in relations between the 37 States which have by mid-November 2018 ratified or accepted the Amendments.