By Malcolm Wu, LLM candidate (Dist.), London School of Economics. 


The crime of aggression surfaced in the post-WWII Nuremberg and Tokyo trials from 1946-1948. 70 years later, with the 2010 Kampala amendments and the 2017 activation resolution, the crime of aggression has found itself a working definition and jurisdiction under the International Criminal Court (ICC). At the 2018 UN event titled ‘20th anniversary of the Rome Statute: the need for universality and the International Criminal Court’s jurisdiction over the crime of aggression’, several State parties have hailed this as a historical milestone. However, is that really the case?

This analysis argues that the crime of aggression as developed by the ICC is not only ineffective in contributing to international law’s attempts to prohibit the use of force, but it also entrenches preconceived notions of victor’s justice and threatens to undermine the principles developed under the law of use of force. To that extent, it is here examined how the crime of aggression’s jurisdictional scope is curtailed by State consent, and how the accommodation of genuine humanitarian intervention in the definition of the ICC’s crime of aggression undermines the law governing use of force.

Problems with the jurisdictional scope of the crime of aggression

The most pressing problem with the crime of aggression (CoA) lies in its jurisdictional scope. Despite aspirations from the African and Caribbean delegations for the CoA to be universally applicable, negotiations at the Kampala amendments resulted in the CoA only being applicable on the basis of State consent (as advocated by Western States). The result of which is that the CoA under article 15bis of the Rome Statute would only apply to State Parties which have ratified the Kampala amendments and did not opt out of its jurisdiction. Additionally, State Parties that have ratified the amendments are entitled to withdraw consent and opt out of jurisdictional regime of the CoA. For a more thorough analysis on the complex negotiations and compromise of the CoA’s jurisdictional scope, one may like to refer to Barriga and Grover’s ‘A Historic Breakthrough on the Crime of Aggression’.

Nonetheless, article 15ter might provide an alternative measure. Article 15ter grants the UNSC the authority to refer a State to the ICC for CoA, thus entailing the possibility of bringing a non-ratifying State under the ICC jurisdiction. However, the willingness for the UNSC to resolve to such a referral is questioned. One should recall the inherent problems of veto abuse among the Permanent 5 members of the UNSC that have been frequently observed in prior resolutions. The resolving of an act of aggression is yet another politically charged matter and as such, there is nothing to indicate that veto abuse would not be strife when article 15ter is in question, rendering the possibility of non-ratifying states being brought under the CoA, unlikely.

Where does this leave the efficacy of the CoA? Firstly, as of 2019, only 39 out of 123 State Parties have ratified the amendment. Countries that are not parties to the Rome Statute are excluded altogether. This leaves the vast majority of countries outside its scope, severely curtailing any potential it may have had on signalling the seriousness of aggression and deterring those from breaching the UN Charter. The prerequisite of State consent has essentially rendered the CoA a ‘toothless tiger’. Secondly, one should recall that the CoA was first introduced under questionable grounds. At the Nuremberg trials, while the P5 were trying Nazi leaders under the CoA for Germany’s annexation of Europe, the P5 conveniently decided to take a Nelsonian blind-eye to the USSR’s annexation of Poland and Finland that similarly took place during WWII. This has led to the view taken by some, such as Justice Pal in the Tokyo trials, that the CoA is nothing but ‘victor’s justice’. While the Kampala Amendments had an opportunity to rectify this preconceived notion by applying the jurisdiction of the CoA to every State party, the adoption of State consent as its jurisdictional basis has rendered this a disappointingly missed opportunity.

Problems with the definition of crime of aggression

Another problem lies in the definition of the crime of aggression. Its content as stipulated in article 8bis of the Rome Statute closely mirrors the wording of the prohibition of the use of force in article 2(4) UNC, i.e. use of force against the sovereignty; territorial integrity; political independence; or in any other manner inconsistent with the Charter of the United Nations. Additionally, article 8bis(2) enshrines a comprehensive list of acts that has been deemed to constitute aggression previously determined by UNGA Resolution 3314 (1974).

Concerns that such a definition may encompass ‘minor breaches’ and ‘legally uncertain’ violations of the UNC led to the act of aggression being qualified by the ‘character, gravity and scale’ in its manifest violation of the UNC, as stipulated in article 8bis(1). This is accompanied with the inclusion of several interpretive ‘understandings’ of the article 8bis – that all three elements of character, gravity and scale had to be met for the act to amount to a manifest violation of the UNC and thus the CoA. Barriga and Grover argued that this was specifically intended so that good faith use of force without UNSC authorisation or ‘genuine humanitarian interventions’ may not be caught by the CoA; it has been argued by Kleb that a genuine humanitarian intervention would not meet the elements of gravity and character.

While the pro-humanitarian intervention camp, such as the UK Government, may welcome this flexibility in accommodating genuine humanitarian interventions unauthorised by the UNSC, one must recall that upon a strict positivist interpretation of the law of use of force, one would reach the ‘inescapable conclusion’ that humanitarian intervention cannot be accommodated for in the UNC and in customary international law, and is thus unlawful. Hence, by accommodating for humanitarian intervention in the definition of the CoA, what kind of signalling effect does this give to the international community about the lawfulness of humanitarian intervention? Indeed, as pointed out by both Ruys and Murphy, when the ICC adjudicates on how humanitarian intervention interacts with the thresholds of the 3 elements to qualify for the CoA, this would entail clarifying and adding to the doctrine of humanitarian intervention such as establishing principles and criteria to ascertain ‘genuine’ humanitarian interventions from non-genuine ones. In doing so, this will create and strengthen perceptions that humanitarian interventions are not unlawful contrary to its standing in the law on use of force, unduly adding impetus to its legitimacy and legality.

This would also mean that the ICC, a criminal court, will have to grapple with highly sensitive and politically charged principles of use of force that even the ICJ has only approached tangentially and with caution. Where the ICC has to ascertain whether the defendant’s alleged humanitarian intervention meets the threshold of the 3 elements so as to qualify for an act of aggression, this would entail setting precedential standards and thresholds i.e. proportionality. However, in doing so, the ICC risks setting precedential standards that would render prior politically controversial humanitarian interventions such as the NATO bombing of Kosovo as manifest acts of aggression and a violation of the UNC. Conversely, the ICC risks setting standards that are too low so that it erodes the principles developed under the law of use of force. While the ICJ may be a better forum to resolve such matters, the findings of prior politically controversial humanitarian interventions as amounting to CoA may come at the cost of displeasing relevant State Parties, which may inadvertently lead to their pulling out of the court’s jurisdiction. This may perhaps offer an explanation as to why even the ICJ has been hesitant in approaching the matter.

Hence, through the recognition and accommodation of genuine humanitarian intervention through article 8bis, the ICC risks undermining the principles developed under the law of use of force through signalling the legality and legitimacy of humanitarian intervention and shaping its doctrine. It is thus submitted that the definition of the crime of aggression should not accommodate ‘genuine humanitarian interventions’.


While hailed as a historical milestone, the weaknesses of the crime of aggression as enacted and activated by the ICC speak louder than its strengths. With its jurisdictional basis of State consent and the unlikeliness of the UNSC to resolve towards a referral, this has led to the majority of States unaffected by the crime of aggression. Any deterrent effect that the crime was intended to produce has been severely curtailed. Ironically, the workings of articles 15bis and 15ter have led to a situation which leaves the P5 unaffected by the crime but able to impose it on others. This entrenches the notion of the CoA being illegitimate, that of ‘victor’s justice,’ and may signal to the international community that the P5 are above the laws of the use of force. Additionally, the accommodation of genuine humanitarian interventions in the definition of the crime of aggression risk the ICC shaping and granting undue legality and legitimacy to a doctrine that has been widely regarded as unlawful under the international law of use of force.