26 - Modes of Liability and Individual Criminal Responsibility

Summary

This chapter is a preliminary analysis of some of the most obvious and pressing challenges that suggest that the overall approach to modes of responsibility in Article 28N of the Statute of the African Court of Justice and Human and Peoples’ rights (African Court), which deals with modes of liability lacks the clarity and specificity required to keep trials moving and provide workable platforms for the effective adjudication of the range of crimes within the jurisdiction of the Court. The author concludes that if Article 28N proceeds in its current form, the new African Court of Justice and Human Right’s (AU Court) will face difficult challenges concerning many of the proposed new modes of liability, including their application to a range of old (e.g., genocide or crimes against humanity) and new (e.g., corruption or piracy) crimes and new types of entities (e.g., legal persons) and their overall impact upon future trials.

1. Introduction

The last two decades have been a period of remarkable growth in the prospects for accountability at the international level through the establishment of an array of international criminal tribunals, including the International Criminal Court (ICC). Despite well documented (and ongoing) travails, these institutions have driven understanding, debate and codification of important aspects of the legal framework required to ensure individual criminal liability for serious violations of international humanitarian law (IHL).

For reasons and motivations that will remain a source of debate and a degree of understandable cynicism, these developments appear to have breathed life into the African Union (AU)’s own efforts towards a regional mechanism governed by its own African Court of Justice and Human Rights Statute (AU Statute).¹ On one view, the proposed AU Statute represents an attempt to improve on its predecessors, such as the Rome Statute that governs the ICC, including expanding the range of applicable crimes and modes of liability, as well as containing an unprecedented recognition of corporate liability in international criminal law.

However, obviously, expansions and modifications do not necessarily equate to genuine progression or enhanced effectiveness. As will be discussed in this Chapter, the AU’s approach to modes of liability in Article 28N of the proposed Statute, whilst being ambitious and innovative, particularly with regard to the addition of new modes of liability that provide an expanded range of ways that crimes may be committed, may not foreshadow improvement or increased efficiency in the AU’s putative adjudicative processes.

On examination, in many instances, it is questionable whether these additions will produce sufficiently specific or certain modes of liability to facilitate effective or more efficient prosecutions. Modes of liability are ‘linking principles’ used to connect accused with particular actions, criminals with other criminals, past decisions with consequences, either foreseen or unforeseen and punishment with moral desert.² As such, especially in complex cases, they must be clearly and specifically defined if they are to prove fit for purpose in the practical setting of a courtroom. As will be discussed in this Chapter, the proposed AU Statute’s approach to liabilities and their probable impact upon these linking capabilities raise many preliminary concerns. If Article 28N proceeds in its current form, the new African Court of Justice and Human Rights (AU Court) will face difficult challenges concerning many of the proposed new modes of liability, including their application to a range of old (e.g. genocide or crimes against humanity) and new (e.g. corruption or piracy) crimes and new types of entities (e.g. legal persons) and their overall impact upon future trials.

This Chapter does not purport to address each and every concern arising from the drafting of Article 28N. It is a preliminary analysis of some of the most obvious and pressing issues that suggest that the overall approach to modes of responsibility in Article 28N lacks the clarity and required to provide routes to the effective adjudication of the range of crimes and to keep trials moving. Some of these problems may have arisen due to simple drafting errors, such as, perhaps, the absence of a clear distinction between principal and accessory liability; others however appear to originate from (well intentioned) practical missteps that include the introduction of a range of new modes of liability (e.g. organizing, directing, facilitating, financing and counselling) that appear to be duplicative or overlapping, with no apparent purpose other than to provide anxious prosecutors with the reassurance that every iota of conceivable misconduct is captured within its reach.

Nonetheless, throughout this Chapter, the authors have endeavoured to keep in mind an obvious practical reality, namely that every international tribunal must engage ‘in a ‘continuous quest’ for theories of liability that can adequately address the systemic character of international crimes’.³ Modes of participation inspired by domestic legislation and integrated into international tribunals’ statutes may not always be a perfectly good fit to address the challenges confronting international criminal law. They must continuously evolve (or be revealed) as new types of involvement or means of participation are exposed in real life trials. Bringing individuals to justice under the (sometimes restrictive) confines of international criminal statutes demands a considerable degree of judicial creativity with regard to honing their utility if misconduct is to be captured and individuals are to be allowed due process and fairly held accountable for any crimes.

As for the latter, due process demands that any judicial creativity must proceed cautiously. Modes of liability may only be interpreted in light of the objectives and principles of international criminal justice. Tribunals must ensure respect for fundamental due process considerations, such as the principles of nullum crimen sine lege, and nulla poena sine lege that are well-established principles in customary international law and apply to the various modes of liability, as well as being codified in the Rome Statute.⁴ Amongst several other prerequisites, such principles demand clarity of pleading of the modes of liability⁵ and that criminal liability should be individual, and sufficiently foreseeable and accessible at the time of the commission of the act or omission.⁶

Whatever the rights or wrongs of Article 28N, the AU Court’s challenges will be no different and, in the end, much will depend on the inventiveness and practical knowhow of the judges working to meet the multifarious demands of the trial processes. However, as will be discussed in this Chapter, given the serious ambiguities and anomalies that run through the critical terms of Article 28N, the drafters have handed these judges a herculean, if not impossible, task. In summary, Article 28N’s drafting may give rise to insuperable obstacles that stand in the way of both the practical and principled application of international law within the AU Court.

Criminal law frameworks generally rest on one of two basic models of criminal liability ‘the unitary perpetrator model’ and ‘the differential participation model’.⁷ The first Section of the Chapter examines whether the drafters of 28N intended to opt for one of these models. On the face of the pleading, Article 28N may have adopted a unitary participation model. It contains a broad mix of modes that include both principal and accessorial modes of liability. However, as will be discussed, this is far from clear. Article 28N contains a myriad of overlapping modes of liability that suggest that the drafter may have been more focused on ensuring that the provision captured every conceivable form of conduct, rather than making an active selection for one and not the other. As unlikely as it may seem, it may be that the drafter simply stumbled into the unitary participation model whilst focused upon this objective.

The second Section of the Chapter examines the various modes of liability contained in Article 28N Statute and discusses some of the interpretative issues that will arise. Article 28N introduces an array of modes of responsibility that have not been part of modern international criminal law statutes. While reproducing many of the modes of liability that are usual, Article 28N includes new forms of complicity, namely organizing, directing, facilitating, financing, counselling and ‘accessory before and after the fact’. As will be discussed, these ‘new’ and overlapping modes of liability may not in the final analysis prove necessary, let alone useful, as vehicles for practical criminal process and adjudication. Most indictments and trials at the international level suffer from overload and vagueness and the pleading of a multitude of liabilities that play no meaningful role in the proceedings, proceedings are likely to do nothing more than distract from the core issues in contention. Additionally, many trials over the last few decades have already suffered the deleterious effects of judicial attempts to assemble joint enterprise forms of liability from statutes that failed to adequately contemplate the challenges of linking remote ‘masterminds’ to those directly perpetrating the crimes. As will be discussed in the second Section, there is little to suggest that several of the liabilities that constitute ‘commission’ in the AU Statute will not lead to the same process problems and appear to have been included without a reasoned consideration of necessity or practical utility.

Finally, the third Section of the Chapter will examine Article 46C of the AU Statute and the manner in which it innovates to define a form of corporate criminal liability in international criminal law. Article 46C appears to describe a mode of liability that is close to the Australian ‘corporate culture’ model of corporate criminal liability which is a variant of the organizational liability model (and not the identification or vicarious model). However, many questions concerning its physical and mental elements remain unanswered and in need of significant judicial interpretation if it is to provide a useful mechanism for determining whether corporations are responsible for criminal conduct. One thorny but essential question concerns more generally how Article 46C will interact with Article 28N, particularly with regard to the aiding and abetting mode of liability.

A. The Distinction between Principal and Accessory Modes of Liability in the Proposed AU Statute

1. International Criminal Law’s Approach

Criminal law processes at the domestic or international level generally opt for one of two approaches when ascribing liability for action against individuals, either the ‘unitary perpetrator model’ or the ‘differential participation model’.⁸ According to the unitary perpetrator model, every person who contributes to the crime is considered a perpetrator regardless of the nature of his or her participation.⁹ This ‘expansive’ notion of perpetratorship is based on the premise that a plurality of persons implies a plurality of offences.¹⁰ Whoever ‘contributes any cause to the commission of a crime, regardless of how close or distant the cause from the final result, they must be considered as (co-) author of the crime’¹¹ and ‘[q]uestions of causation, mens rea, justification and excuse arise independently from the participant’s own act and a conviction is of the crime proper’.¹² The unitary model obviates any need for distinguishing between participants in wrongdoing: ‘[t]here are no accomplices; all are principals’.¹³ Under this theory, the actual contribution of the individual is significant only with regard to sentencing.¹⁴

In contrast, the ‘differential participation model’ distinguishes between perpetrators and accessories. In basic terms, perpetrators are those at the centre of the crime, while accessories assist in, or prompt its commission and act, for instance, as solicitors, instigators, or aiders and abettors.¹⁵ This model is based on the assumption that participation in a crime can be so different in weight and proximity that each person should be treated differently according to their involvement.¹⁶ This model is not only relevant for assessing an appropriate sentence, but also for clarifying the individual criminal liability for acts or omissions.¹⁷

Jackson identifies three stages of differentiation in the participation in wrongdoing: (1) the doctrinal differentiation that distinguishes amongst participants in wrongdoing at the level of legal doctrine. At this level, the law recognizes the category of accomplices with certain doctrinal requirements of conduct and fault; (2) the differentiation in the attribution of responsibility which distinguishes among participants in wrongdoing at the stage of conviction or responsibility. The attribution of responsibility is not linked to the wrong of the principal but to the accessories’ own contribution to that wrong; and (3) the differentiation in the consequences of responsibility which distinguishes among participants at the sentencing or remedial stage of the system.¹⁸ In stage (3), variance in role is expressed at the sentencing level.¹⁹ According to Jackson, the principles of culpability and fair labelling require differentiation amongst participants at each of these three stages.²⁰

At the sentencing stage, this model allows ‘for both sentencing guidelines according to the various modes of participation and also a unitary range of sentencing’.²¹ In the latter case, each contribution to the crime is considered on its own:²² ‘by either upgrading perpetrators or downgrading accessories’.²³ Courts thus determine the penalty according to the mode and degree of participation.²⁴

Many international criminal law commentators consider that distinguishing between principal perpetrators and accomplices is an important asset to international criminal law, especially in the identification of the masterminds behind the crimes.²⁵ This support was echoed most recently during the academic debate arising from Stewart’s argument that accomplice liability in international criminal law should be reduced to a single notion of perpetration.²⁶

Stewart observes (correctly) that complicity has not been sufficiently dealt with ‘in the scholarly revolt against international modes of liability’.²⁷ Instead the debate has mainly focused on joint criminal enterprise, command responsibility, perpetration and co-perpetration.²⁸ Consequently, complicity has escaped some of the criticism that has befallen other modes of liability.²⁹ Stewart argues that complicity conflicts with both: (1) the existence of congruence between the mental element of the crime and the mental element required of the accomplice and (2) the need for a causal connection between the accomplice’s acts and the harm contemplated in the crime.³⁰ In the end, he concludes, that the differentiated approach violates the principles of culpability³¹ and fair labelling.³²

Stewart suggests ‘the source of complicity’s departures from basic principles […] stems from international criminal law’s emulation of objectionable domestic criminal doctrine’.³³ He further argues, ‘complicity’s most objectionable characteristics are inherited from domestic exemplars that some scholars denounce as a conceptual “disgrace”’.³⁴ For example, referring to the debate on the mental element for accessorial liability, more particularly the competing rationale for the purpose/knowledge/recklessness standards, Stewart states:

On closer inspection, none of the three highly debated standards (purpose, knowledge, recklessness) is theoretically justifiable. Like other modes of liability in international criminal justice, all three violate the principle of culpability in certain circumstances because they all tolerate the imposition of a crime’s stigma in situations in which the person convicted of the offence did not make the blameworthy choice necessary to be found guilty of that particular offence. Many point out the perversity of using JCE III to escalate blame for genocide in this manner, but what about instances in which complicity has an identical effect? With accessorial liability, individuals are also held responsible for genocide where they knew or were merely aware that genocide was one of a number of crimes that would probably be committed. These scenarios, which are actually more common in practice, violate culpability too. Tellingly, these violations are explicitly based on examples drawn from a host of Western systems.³⁵

As a consequence, Stewart argues that complicity ‘should collapse along with all other modes of liability into a single broad notion of perpetration’,³⁶ where a principal is any participant who ‘made a substantial causal contribution to a prohibited harm while harboring the mental element necessary to make him responsible for that crime’.³⁷ In line with the unitary model, the accomplice’s contribution to the crime can be accounted for at the sentencing stage.³⁸

However, several well respected academics in area of analysis, including Jackson, Ohlin, Robinson, Werle and Burghardt have rejected this radical proposal and reiterated their support for a differentiated system of responsibility for international crimes.³⁹ While agreeing with Stewart’s criticism that the current interpretation of accessorial modes of liability in international criminal law is far from perfect, they consider that distinguishing between principal perpetrators and accomplices remains important to international criminal law, especially in the identification of the masterminds behind the crimes.⁴⁰

According to Werle and Burghardt, Jackson’s arguments that the principles of culpability and fair labelling are conflicted do not hold water.⁴¹ They suggest that both models (unitary and differentiated) may be appropriate for international criminal law. Nevertheless, they conclude that, ‘certain normative and empirical features of international criminal law, in general, and of the system of the ICC Statute in particular, weigh heavily in favor of a differentiation model, where modes of participation are indicative of the degree of criminal responsibility’.⁴² To their minds, international criminal law is charged with developing ‘normative criteria for gradation of responsibility’ insofar as the discipline deals with ‘the most serious crimes committed by a large number of persons in complex factual scenarios’.⁴³

Accordingly, they consider that modes of participation are necessary indicators of the degree of individual criminal responsibility.⁴⁴ The differentiation model involves key procedural consequences such as the obligation of the prosecution to set out facts and legal elements of the charges in detail; the application of different legal thresholds to the different modes of participation; and a more transparent and predictable sentencing process.⁴⁵ By contrast, the unitary model avoids the ‘thorny issue of normative gradation for the purpose of a guilty verdict, only to find it again at the sentencing stage’.⁴⁶

No matter how difficult is the task of defining the criteria that may be used to establish the degree of blameworthiness, it is one we cannot shy away from without abandoning the constitutive idea of international criminal law itself – the idea of individual criminal responsibility.⁴⁷

For Jackson, Stewart’s proposal is also flawed and the expressive benefits of a unitary model of responsibility are more illusory than real.⁴⁸ According to Jackson, even if there were some benefit to Stewart’s proposal, it overrides the fundamental principles of culpability and fair labelling that underpin a differentiated model of participation in crime.⁴⁹ Jackson argues that eliminating complicity would potentially violate the principle of fair labelling in criminal law, which requires that ‘wrongdoing is labelled accurately and, with a sufficient degree of specificity to distinguish law-breaking of a different kind or gravity’.⁵⁰ Jackson and Ohlin highlighted that otherwise ‘some participants’ responsibility would be radically over-weighted, others radically under-weighted, and the system would tell us virtually nothing about what the wrongdoer did’.⁵¹

Jackson eloquently summarizes the importance of the differentiated model:

A unitary model of participation is inconsistent with how we do, and ought to, think about responsibility. To borrow Darryl Robinson’s example, the groom, bartender, and guest are all participants in a wedding. Indeed, they may all causally contribute to it. But we would not deny profound differences in their roles. Likewise, in the context of wrongdoing, complicity is a necessary element of a complete account of morality and responsibility. Gardner argues that ‘the distinction between principals and accomplices is embedded in the structure of rational agency. As rational beings, we cannot live without it’. There are two elements to Gardner’s account. The first concerns the wrongness of complicity: we should be concerned with not only the harms we do ourselves but also those we help or influence others to do. The second concerns the scope of that wrong: both principals and accomplices should be responsible for their own actions.⁵²

Finally, according to Robinson, the differentiation model has an expressive function by reflecting ‘meaningful moral differences between those who cause or control the crime and those who made blameworthy but minor and secondary contributions’.⁵³ As an illustrative example, Robinson explains that the ‘should have known’ standard of command responsibility has been accepted as a justifiable element in the context of a command relationship. The unitary model will not provide such flexibility and would have to either allow the ‘should have known’ standard in all contexts or prohibit it entirely.⁵⁴

Therefore, unsurprisingly perhaps, the ad hocs have tended to interpret their Statutes to distinguish between principals and accomplices and have to a greater or lesser degree adopted the ‘differential participation model’. The classic principal modes of liability at these tribunals are commission and joint commission. Among the accessory modes of liability, there are two main ways in which an individual may act as an accomplice; either ordering, planning, and instigating (which describe proximity between the perpetrator and the commission of the crime), or aiding and abetting (which generally entails a subsidiary contribution to the criminal act).

However, the ad hocs’ Statutes do not contain these express distinctions and instead place principal liability at the same level and within the same category as accessorial liability.⁵⁵ Instead, these distinctions have largely evolved through a process of incremental interpretation and jurisprudential development. As is now part of international justice’s well known legacy, it was only in 1999, when grappling with the complexity of how to define a joint criminal enterprise (JCE) liability to cope with contributions to collective action, that the Tadić ICTY Appeals Chamber distinguished between principal and accessory liability:

In light of the preceding propositions it is now appropriate to distinguish between acting in pursuance of a common purpose or design to commit a crime, and aiding and abetting.

1. (i) The aider and abettor is always an accessory to a crime perpetrated by another person, the principal. (…)⁵⁶

In 2003, the ICTY Appeal Decision in the Milutinović case further clarified that, under customary international law (and therefore under Article 7 of the ICTY Statute), the doctrine of JCE gave rise to principal liability.⁵⁷ Subsequently, the ICTY has relied upon these two decisions as a basis upon which they could distinguish between principal and accessory liability.⁵⁸ Similarly, at the ICTR, the Court relied upon the Tadić Decision to hold that Article 6 of their Statute (which mirrors Article 7 of the ICTY Statute) expresses a distinction between principal and accessory liability.⁵⁹ Consistent with the maintenance of these distinctions, accessory modes of responsibility generally attract a lower sentence than those resulting from responsibility as a co-perpetrator.⁶⁰

This distinction is more apparent on the face of the Rome Statute. However, judicial interpretation has created a degree of uncertainty concerning the nature of the distinction that is yet to be resolved. In the first place, the Rome Statute expressly enumerates four types of criminal responsibility: (1) committing a crime – perpetration and co-perpetration⁶¹; (2) ordering and instigating⁶²; (3) aiding and abetting⁶³; and (4) contributing to the commission of a crime by a group of persons acting with a common purpose.⁶⁴ In Lubanga, the first case at the ICC, Pre-Trial Chamber I drew several distinctions between these modes of liability, noting that:

[The Rome Statute] distinguished between (i) the commission stricto sensu of a crime by a person as an individual, jointly with another or through another person within the meaning of Article 25(3)(a) of the Statute, and (ii) the responsibility of superiors under Article 28 of the Statue and ‘any other forms of accessory, as opposed to principal liability provided for in Article 25(3)(b) to (d) of the Statute’ [ordering, soliciting and inducing, aiding and abetting and contribution].⁶⁵

Similarly, in 2010, the Mbarushimana Pre-Trial Chamber found that Article 25(3) entailed a hierarchy of responsibility and described the modes of liability as being arranged in accordance with ‘a value oriented hierarchy of participation in a crime under international law’, where the ‘control over the crime decreases’ as one moves down the sub-paragraphs.⁶⁶ This is consistent with a value-oriented hierarchy of participation in a crime that places commission as the highest degree of individual responsibility and contribution to a group crime as the ‘weakest mode of participation’.⁶⁷

However, with specific focus upon sentencing principles, the Katanga Trial Chamber appears to have chipped away at this erstwhile clarity. Whilst using the ‘differential participation model’ to classify principals and accessories, the Katanga Trial Chamber rejected the Mbarushimana Pre-Trial Chamber’s decision and held that the distinction between the different modes of liability as principal or accomplice did not amount to a hierarchy of blameworthiness. The Chamber also stated that there was no rule in the Statute or the Rules of Procedure that necessitated the imposition of lower sentences for accomplices as against principals. Referring to national criminal codes, such as that operative in Germany, where the sentence for each is identical (even if mitigation may lower the eventual sentence of the aider⁶⁸), the Chamber concluded that there was no automatic correlation between modes of liability and penalty.⁶⁹ Accordingly, a person responsible as an instigator may incur a penalty akin or even identical to that handed down against a person found responsible as a perpetrator of the same crime.⁷⁰

In sum, the ICCs’ precise approach to these foundational issues remains a work in progress. Although the modes of liability in the Rome Statute appear embedded in a differential participation model, there is still plenty of room for manoeuvre before the Appeals Chamber proffers some certainty to these issues, especially with regard to the sentencing provisions.⁷¹

2. Article 28N’s Approach

Turning to Article 28N of the AU Statute and which of the two models are intended, it begins with the phrase: ‘An offence is committed by any person who, in relation to any of the crimes or offences provided for in this Statute: [sub-paragraphs]’.⁷² The three subsequent sub-paragraphs contain a broad mix of overlapping modes that include both principal and accessorial modes of liability. Sub-paragraph (i) lists a series of liabilities, namely, ‘incites, instigates, organizes, directs, facilitates, finances, counsels or participates as a principal, co-principal, agent or accomplice in any of the offences set forth in the present Statute’. The second and third sub-paragraphs respectively refer to accessorial liability: ‘aiding and abetting’ and a mode of joint liability for anyone who ‘is an accessory before or after the fact or in any other manner participates in a collaboration or conspiracy’.

As may be seen, this construction is anything but straightforward. On the face of the pleading, Article 28N adopts a unitary participation model whereby anyone who contributes to the crime is to be held liable as principal. The overarching definition of commission suggests that Article 28N entails only one main mode of liability (i.e. ‘commission’) that is sub-divided into several forms (as outlined in the three sub-paragraphs referenced above). In the AU Statute, the imputation of the conduct of a principal to an accomplice is achieved by including complicit conduct within the terms of commission. This approach appears to be broadly consistent with the approach taken by certain African States (e.g. Kenya,⁷³ Tanzania,⁷⁴ Nigeria⁷⁵ and Zambia⁷⁶) whereby commission is elaborated as a catchall category into which falls all manner of other forms: for example, the person who aided and abetted, counselled or procured a person to commit the act is regarded as the person who committed the act.

As outlined above, the AU Statute’s approach contrasts with the current trend in international criminal law. Consequently, the adoption of the unitary participation model in the proposed AU Statute may have adopted a path that veers away from a strict adherence to principles of culpability and fair labelling. As discussed above, in international criminal law, these principles suggest differentiation amongst participants in crime, not only at the sentencing stage, but also in the attribution of responsibility. Accordingly, a statute where modes of participation are indicative of the nature and degree of individual criminal responsibility may more accurately and adequately reflect the complex factual situations, the large number of perpetrators, the variance in involvement, and the seriousness of the crimes that may be considered by the new Court. Therefore, as with the ICC, the guarantee of a distinction between these different forms of individual responsibility may be important to ensure the legitimacy of the work of the AU Court.⁷⁷

A closer examination of Article 28N’s unitary participation formulation raises even more doubts about its ability to achieve some of these aims. Not only does Article 28N define the commission of an offence as the same as an attempted commission of an offence (an ‘offence is committed by a person who, in relation to any of the crimes or offences provided for the Statute also attempts to commit any of the offences set forth’), many of the forms of commission, beyond reassuring the anxious drafter that every conceivable direct or indirect act or omission falls within its terms, appear to serve little or no useful purpose.

As will be further discussed below,⁷⁸ irrespective of whether they are forms of direct or indirect commission, the appearance of a number of these overlapping modes of liability serves only to confuse rather than clarify or enhance effective assessment of individual culpability. A striking illustration of this potential may be seen in the use of ‘co-principal’ in sub-paragraph (i) which suggests that Article 28N is designed to encompass a mode of joint commission. However, this reference appears to overlap with, or even mirror, the notion of ‘collaboration’ in sub-paragraph (iii). Similarly, incitement and instigation are both included in Article 28 as forms of commission. However, as instigating requires acts that influence the direct perpetrator by inciting, soliciting or otherwise inducing him to commit the crime,⁷⁹ Article 28N’s inclusion of both incitement and instigation may therefore have created unhelpful overlap or even duplication.

Therefore, although Article 28N may reflect an intention to adopt a ‘unitary participation model’, this intention is made less clear by the myriad of additional modes of liability. As unlikely as it may seem, it may be that the drafter simply stumbled into adopting the unitary participation model but then departed from this model through a determination not to be caught short.

Moreover, other aspects of the Statute, such as Article 43A addressing sentencing, fails to proffer any decisive clarification of these important questions. In stating that ‘[i]n imposing the sentences, the Trial Chambers shall take into account such factors as the gravity of the offence and the individual circumstances of the convicted persons’⁸⁰ it reproduces the provision of the ICTY and ICC Statutes. As discussed above, the ICTY has interpreted these provisions as reflective of the differential model,⁸¹ whereas the ICC Trial Chamber in Katanga has done otherwise.

3. Overloading the Statute with Multiple Modes of Liability

As mentioned above, Article 28N introduces an array of modes of responsibility that have not been part of modern international criminal law statutes. While reproducing many of the modes of liability that are the norm in the ad hocs’ Statutes, Article 28N includes new forms of complicity, namely organizing, directing, facilitating, financing, counselling and ‘accessory before and after the fact’.

However, these ‘new’ and overlapping modes of liability may not be necessary, let alone useful, to cope with the rigours of practical criminal process and adjudication. Whether one accepts the (persuasive) arguments of commentators such as Ambos, who have recommended a radical reduction in the range of modes of liability and ‘a rule limiting complicity (secondary participation) to inducement/instigation and other assistance (“aiding and abetting”)’,⁸² it is difficult to understand the purpose of loading the AU Statute in this manner. In particular, it is already difficult to delineate some of the modes of accessory liability contained in the ad hocs’ and ICC Statutes: planning, ordering, instigating, aiding and abetting and contributing are in practice almost impossible to separate from each other,⁸³ especially when viewed through (anticipated) complexity of a range of concurrent criminal and non-criminal action. There are many such overlaps, including between abetting, ordering and inducing,⁸⁴ as well as a lack of a clear demarcation between soliciting and inducing, that each appear to encompass a situation where a person is influenced by another to commit a crime.⁸⁵

In reality, the tendencies of most international prosecutors to overload and plead as vague an indictment as loose pleading standards allow, often leads to indictments and trials at the international level that suffer from a multitude of overlapping liabilities that play little role in the proceedings other than to distract from the core issues in contention. As will be discussed below, there is little to suggest that several of the liabilities that constitute ‘commission’ in the AU Statute do not equally foreshadow a level of distraction that may serve to undermine the precision and the accuracy of the adjudication.

B. The New Modes of Liability

As noted above, Article 28N sub-paragraph I includes organizing, directing, facilitating, financing and counselling as forms of commission. These are modes of liability that have not been deployed at the ad hocs or the ICC. Indeed, most of these new modes of liability seem to have been derived or adopted from the UN Convention on Transnational Organized Crime (UNTOC) which requires State parties to adopt legislative measures to establish as specific criminal offences the following conduct: organizing, directing, aiding, abetting, facilitating or counselling the commission of serious crime involving an organized criminal group.⁸⁶

1. Organizing and Directing

As noted, ‘organizing’ and ‘directing’ are forms of liability that are not employed at the ad hocs or at the ICC. Nevertheless, their insertion in Article 28N appears to have been inspired by international and regional instruments that seek to address and criminalize terrorism and organized crime. Organizing and directing, along with facilitating (see below), are contained in UNTOC Article 5(1)(b) that lists modes of liability in relation to an array of organized crime. The Council of Europe Convention on the Prevention of Terrorism also requires state parties to adopt such measures as may be necessary to establish as a criminal offence ‘organizing or directing others to commit’ the offences of public provocation to commit a terrorist offence, recruitment for terrorism and training for terrorism.⁸⁷ Research, however, suggests that most states have not chosen to incorporate these particular modes of liability into their criminal legislation and the majority have instead elected to utilize more classic modes of liability such as aiding and abetting.⁸⁸

Undoubtedly, although there are lessons to be learnt from the ad hocs and the ICC, these ‘new’ modes of liability will require novel and extensive judicial interpretation if they are to be useful. Although organizing is not a form of liability at the ad hocs or the ICC, the conduct encapsulated appears to be the same as, or closely resembles, that captured by the ‘planning’ mode of liability deployed at the ad hocs. Organizing is commonly defined as making arrangements for something to happen.⁸⁹ According to the ad hocs’ jurisprudence, an individual may be held liable when he did not physically commit a crime but participated in its planning. Planning is defined as one or several persons contemplating the commission of a crime at both the preparation and execution phases.⁹⁰ The actus reus of planning requires that one or more persons design the criminal conduct that is later perpetrated. The planning should have been a factor substantially contributing to the criminal conduct.⁹¹ The required mens rea is the intent to plan the commission of the crime or, at a minimum, the awareness of a substantial likelihood that a crime will be committed in the execution of that plan.⁹² An accused cannot be charged with both planning and committing (or ordering) on the same facts. However, planning may be considered an aggravating circumstance.⁹³

Similar convergence and overlaps may be seen with regard to Article 28N’s ‘directing’ mode of liability. Directing appears to be the same as ordering someone, especially officially,⁹⁴ and this appears to be similar to, or the same as, ‘ordering’ (as commonly applied at the ad hocs and the ICC). At the ad hocs, responsibility for ordering requires proof that a person in a position of authority used that authority (de jure or de facto) to instruct another to either commit an offence that in fact occurs or is attempted or perform an act or omission in the execution of which a crime is carried out.⁹⁵ The order must have been a factor substantially contributing to the physical perpetration of a crime or underlying offence.⁹⁶ The ICC has taken a similar approach.⁹⁷

In addition, the accused need only instruct another to carry out an act or engage in an omission – and not necessarily a crime or underlying offence per se – if he has the intent that a crime or underlying offence be committed in the execution of the order, or if he is aware of the substantial likelihood that a crime or underlying offence will be committed.⁹⁸ The ICC similarly requires the person to be at least aware that the crime would be committed in the ordinary course of events as a consequence of the execution or implementation of the order.⁹⁹

2. Facilitation

In international criminal law, facilitation does not constitute a stand-alone mode of liability but is closely related to, or the same as, the concept of aiding and abetting: ‘mere’ facilitation may suffice for aiding and abetting.¹⁰⁰

Similar to the inclusion of ‘organizing’ and ‘directing’, Article 28N’s adoption of ‘facilitation’ appears to have been inspired by UNTOC and the need for modes of responsibility suited for the prosecution of specified crimes such as terrorism, trafficking in persons or drugs. As noted above, Article 5(1)(b) states that each State Party shall adopt such legislative and other measures as may be necessary to establish as criminal offences, when committed intentionally: (…) (b) Organizing, directing, aiding, abetting, facilitating or counselling the commission of serious crime involving an organized criminal group. However, likewise, the Convention does not offer any insight into the essential constituent elements, preferring to allow States a degree of flexibility in transposing the provision into their domestic legislation. A study of a selection of 15 countries suggests that very few countries have opted to rely upon facilitation as a specific mode of liability and instead rely on the aiding and abetting mode of liability.¹⁰¹

3. Financing

In international criminal law, financing is not considered as a mode of liability per se. On the contrary, it is generally an act or conduct that constitutes a way or form of aiding and abetting the crime. An emblematic example may be seen in the Stanišić and Simatović case at the ICTY wherein the accused are charged with aiding and abetting war crimes and crimes against humanity for, inter alia, allegedly financing training camps and special units of the Republic of Serbia State Security and other Serb Forces.¹⁰²

Similar to the above-mentioned modes of liability, the inclusion of ‘financing’ within Article 28N appears to be inspired by the introduction of an array of economic crimes within the AU Statute, such as terrorism, trafficking in humans or drugs, and piracy. The UN Convention for the Suppression of the Financing of Terrorism observes that the ‘financing of terrorism is a matter of grave concern to the international community as a whole’ and states that ‘any person commits an offence within the meaning of this Convention if that person by any means, directly or indirectly, unlawfully and willfully, provides or collects funds with the intention that they should be used or in the knowledge that they are to be used…’.¹⁰³ Recently, on 31 March 2017, the European Union took a similar approach; publishing Directive 2017/541 on combating terrorism, thereby imposing on member states the obligation to criminalize the financing of terrorism.¹⁰⁴ The Directive states, inter alia, that ‘criminalization should cover not only the financing of terrorist acts, but also the financing of a terrorist group, as well as other offences related to terrorist activities, such as the recruitment and training, or travel for the purpose of terrorism, with a view to disrupting the support structures facilitating the commission of terrorist offences.’¹⁰⁵

Financing terrorism is defined as providing or collecting funds, by any means, directly or indirectly, with the intention that they be used, or in the knowledge that they are to be used, in full or in part, to commit, or to contribute to the commission of, any of specific offences such as terrorist offences and offences related to a terrorist group and offences related to terrorist activities (i.e. public provocation to commit a terrorist offence, recruitment or providing training for terrorism).¹⁰⁶ For certain offences, such as terrorist offences, it is not necessary that the funds be in fact used, in full or in part, to commit, or to contribute to the commission of, any of those offences, nor is it required that the offender knows for which specific offence or offences the funds are to be used.¹⁰⁷

Thus, one can see where the drafter of Article 28N was headed. However, less clear is its value – particularly in light of the implicit incorporation of financing at the ad hocs as one of a range of similarly incriminating acts (alongside such acts as the provision of logistics, training or propaganda) alleged to support terrorism. It appears as if the drafters, without considering need or utility, merely adopted the literal terms of the UN Convention for the Suppression of the Financing of Terrorism and other similar international agreements thereby creating a ?new? liability that is at best duplicative.

4. Counselling

Although this mode of liability is new in international criminal law, it has been widely used by national jurisdictions,¹⁰⁸ and is also contained in several African Criminal Codes (e.g. Ghana,¹⁰⁹ Kenya,¹¹⁰ Tanzania,¹¹¹ Nigeria¹¹² and Zambia¹¹³). For example, under the United Kingdom (UK) Serious Crime Act 2007, a person may become a party to a crime as a secondary party (who aids, abets, counsels or procures the commission of an offence).¹¹⁴

However, the mode of liability appears to overlap substantially with abetting as well as instigating that, as argued above, might itself be considered as an umbrella term that also encompasses incitement and encouragement. UK courts, for example, have accepted that counselling and abetting were very similar. In Attorney General’s Reference (No 1 of 1975), it was held that a meeting of minds between two persons was necessary to hold someone liable for abetting or counselling a crime.¹¹⁵ While abetting involves some form of encouragement communicated to and known by the principal to commit the crime (before or during the act), counselling refers to conduct prior to the commission of the crime such as advising on an offence or supplying information necessary to commit the offence.¹¹⁶ Counselling involves ‘advising, soliciting, encouraging, or threatening the principal to commit an offence’. In Canada, a similar approach has been taken: counselling involves ‘actively inducing’.¹¹⁷ It includes procuring, soliciting or inciting.¹¹⁸

Nevertheless, it might be argued that the Article 28N’s term ‘counselling’ evokes a particular type of instigation and therefore may in turn have a useful delineating and expressive purpose. It may help to capture conduct and express specific wrongdoing that is particularly relevant for the new financial crimes, such as the liability of a lawyer or accountant who knowingly provides advice in furtherance of money laundering or corruption. However, in the context of Article 28N and the many new modes of liability, some doubt must arise whether another mode of liability adds to the confusion or will prove to be of real benefit in delineating and prosecuting specific conduct or otherwise promoting fair labelling. As with many of these concerns, only time and practical adjudication will tell.

C. The Classic Modes of Liability in International Criminal Law

1. Aiding and Abetting

Article 28N includes the aiding and abetting mode of liability. It does not elaborate on the constituent elements. Its definition is limited to the statement that an offence is committed by any person who, in relation to any of the crimes or offences provided for in the Statute, aids or abets the commission of any of the offences set forth in the Statute.

Aiding and abetting has been commonly used by the prosecution at the ad hocs and will likely be frequently used at the ICC in its future trials. Similarly, the Statutes of the ad hocs and the ICC include aiding and abetting as a form of liability but without elaboration. However, those terms and the way in which the ad hocs have approached the liability provide a number of lessons for any future AU Court.

The ad hocs and the ICC Statutes define the mode of liability differently. The ad hocs’ Statutes consider that a person who aided and abetted in the planning, preparation or execution of a crime shall be individually responsible for the crime.¹¹⁹ The Rome Statute states that:

In accordance with this Statute, a person shall be criminally responsible and liable for punishment for a crime within the jurisdiction of the Court if that person:

1. (c) For the purpose of facilitating the commission of such a crime, aids, abets or otherwise assists in its commission or its attempted commission, including providing the means for its commission.¹²⁰

The Rome Statute definition appears to contain additional elements: ‘for the purpose of facilitating’, ‘otherwise assists’, ‘attempted’ and ‘including providing the means’. There is little guidance thus far concerning how the ICC will interpret these defining elements. Although two accused have been recently convicted of aiding and abetting or otherwise assisting the commission of the offence of presenting false evidence and corruptly influencing witnesses in the recent contempt case in Prosecutor v. Bemba Gombo et al., the ICC judges did not provide any real insight into Article 25(3)(c).¹²¹. In Blé Goudé, the ICC provided the following clarification: ‘In essence, what is required for this form of responsibility is that the person provides assistance to the commission of a crime and that, in engaging in this conduct, he or she intends to facilitate the commission of the crime.’¹²²

In contrast, the ad hocs have clarified the basic elements of this mode of liability. The ad hocs define the actus reus of aiding and abetting as carrying out acts to assist, encourage or lend moral support to the commission of a certain specific crime and this support has a substantial effect upon the perpetration of the crime.¹²³ The criminal participation must have a direct and substantial effect on the commission of the offence.¹²⁴ In other words, ‘the criminal act most probably would not have occurred in the same way had not someone acted in the role that the accused in fact assumed’.¹²⁵ The ICTY has interpreted the notion of ‘substantial contribution’ in a broad way by including encouragement of the perpetrator or tacit approval.¹²⁶

However, as the debate concerning whether ‘specific direction’ was part of international customary law and an element of aiding and abetting shows, the development, or clarification, of the elements of aiding and abetting at the ad hocs has not been without controversy. On the contrary, in the Perišić case, the ICTY Appeals Chamber considered whether specific direction was an element of aiding and abetting. After reviewing the ICTY and ICTR case law, it concluded that specific direction was an element of the actus reus of aiding and abetting.¹²⁷ As the Chamber explained, the element of specific direction establishes a culpable link between assistance provided by an accused and the crimes of principal perpetrators.¹²⁸ The Chamber further explained that for acts geographically or otherwise proximate to the crimes of principal perpetrators, specific direction might be demonstrated implicitly through discussion of other elements of aiding and abetting liability, such as substantial contribution. However, where an aider and abettor is remote from the crimes the other elements of aiding and abetting may not be sufficient to establish specific direction. In such cases, specific direction should be specifically considered.¹²⁹

However, this decision was highly controversial and subsequently reversed by ICTY Appeals Chamber decisions. In 2014, in the Šainović et al. case, the Appeals Judgment concluded that ‘specific direction’ was not an element of aiding and abetting liability ‘accurately reflecting customary international law and the legal standard that has been constantly and consistently applied in determining aiding and abetting liability.’¹³⁰ The Appeals Chamber noted that, prior to the Perišić Appeals Judgment, ‘no independent specific direction requirement was applied by the Appeals Chamber to the facts of any case before it’.¹³¹ The Appeals Chamber affirmed that ‘under customary international law, the actus reus of aiding and abetting consists of practical assistance, encouragement, or moral support with a substantial effect on the perpetration of the crime’.¹³² Recently, the ICTY Appeals Chamber in the Stanišić and Simatović case re-affirmed the Šainović ruling overturning decisions that rested upon the application of this element.¹³³

With regards to the mens rea, the ad hocs determined that an aider and abettor should have known that his acts would assist in the commission of the crime by the principal perpetrator and must be aware of the ‘essential elements’ of the crime. It does not require that he share the intention of the principal perpetrator of such crime.¹³⁴ The ICTY recognized that knowledge is an element of aiding and abetting under customary international law.¹³⁵ However, it is not necessary that the aider and abettor knew the precise crime that was intended or actually committed, as long as he was aware that one or a number of crimes would probably be committed, and one of these crimes was in fact committed.¹³⁶

As noted above, Article 25(3)(c) of the Rome Statute further requires that the assistance be made ‘for the purpose of facilitating the commission of [the] crime’, thus introducing an additional subjective threshold to the ordinary mens rea requirement of aiding and abetting.¹³⁷ This new element departs from customary international law as considered and determined in the ad hocs’ case law detailed above. As outlined above, in the Blé Goudé case, the ICC stated that: ‘what is required for this form of responsibility is that the person intends to facilitate the commission of the crime.’¹³⁸

Whether this interpretation of ‘purpose’ is wholly justified will remain an ongoing debate for many years. However, what is plain is that ‘purpose’ implies a subjective element stricter than mere knowledge that the accomplice aided or abetted the commission of the crime.¹³⁹ This higher threshold however only refers to the act of facilitation, not the crime itself. Accordingly, this version of aiding and abetting requires a double mental element: one for the act of assistance and one for the crime.¹⁴⁰

Given Article 28N’s failure to elaborate on the elements of the aiding and abetting mode of liability, it is not clear what path will be taken by the AU Court to its constituent elements. As the experience at the ICTY has shown, international courts have considerable discretion in interpreting the plain words of a statute. Given that Article 28N fails to proffer any meaningful insight into the constituent elements of aiding and abetting, any future AU Court have considerable room to decide whether to opt for the ICC’s more demanding approach – requiring a demonstration of the purpose of facilitation of the crime – or the ICTY’s ‘purposeless’ approach.

Instigating – Inciting

Like the ad hocs and ICC statutes, Article 28N includes both incitement and instigation. However, although the ad hocs’ and ICC statutes provide for both concepts, they draw a distinction between ‘incitement or instigation generally and direct and public incitement to genocide’.¹⁴¹

While the first category (incitement/instigation)¹⁴² is considered as encompassing accessory modes of liability (punishable only where it leads to the actual commission of an offence intended by the instigator¹⁴³), the second category (direct and public incitement)¹⁴⁴ has been held to be an inchoate crime only applicable to the crime of genocide.¹⁴⁵

The AU Statute fails to draw these distinctions, namely incitement is only included as a mode of liability (first category). Unlike the AU Statute, the ad hocs and the ICC statutes expressly refer to the inchoate crime of direct and public incitement to commit genocide.¹⁴⁶ Incitement is only mentioned once in the AU Statute, as the first in a long line of modes of liability that include instigating, organizing, facilitating and financing.

Turning to the potential interpretation of this mode of liability, the ad hocs’ jurisprudence determines that conduct that constitutes incitement is also encompassed by instigation. Instigating has been defined at the ad hocs as ‘prompting’, ‘urging or encouraging’ another to commit an offence.¹⁴⁷ In sum, instigating requires acts that influence the direct perpetrator by inciting, soliciting or otherwise inducing him to commit the crime.¹⁴⁸

As noted above, although the Rome Statute does not expressly refer to instigating, inducing and soliciting have been interpreted as substantially covering the same ground.¹⁴⁹ In Harun, the ICC considered inducing equivalent to inciting.¹⁵⁰

Whilst it is sufficient at the ad hocs to demonstrate that the instigation was a factor substantially contributing to the conduct of another person committing the crime,¹⁵¹ the ICC requires the inducement (instigation) to involve exertion of influence over another person to commit a crime and the existence of a direct effect on the commission of the crime.¹⁵² An analysis of the case law at the ICC suggests that the ‘direct effect’ criterion is the same as the ad hoc’s ‘substantial effect’ criterion.¹⁵³ At both the ad hocs and the ICC, it needs to be shown that the accused should have been aware of the likelihood that the commission of a crime would be a probable consequence of his acts.¹⁵⁴

In sum, the case law of the ad hocs and the ICC suggests that instigating (as a mode of liability) may be considered to be an umbrella term that includes inciting/inducing the crime. The inclusion of both incitement and instigation in Article 28N appears to disregard this jurisprudential history in favour of more duplication.

2. Joint (Principal and Accessory) Liability

Article 28N appears to address crimes committed as part of joint plans involving various masterminds and physical perpetrators. It states, inter alia, that:

An offence is committed by any person who, in relation to any of the crimes or offences provided for in this Statute:

1. i. Incites, instigates, organizes, directs, facilitates, finances, counsels or participates as a principal, co-principal, agent or accomplice in any of the offences set forth in the present Statute;

   (…)

2. iii. Is an accessory before or after the fact or in any other manner participates in a collaboration or conspiracy to commit any of the offences set forth in the present Statute.¹⁵⁵

This aspect of Article 28N may represent some form of tacit recognition of the experience of the ad hocs and the ICC and international criminal law in general, namely that in most instances cases are likely to be largely focused upon crimes and accountability involving criminal plans, collective action and the examination of ‘a multi-perpetrator setting’.¹⁵⁶ Modern international criminal law has continuously wrestled with the collective nature of crime involving multiple masterminds and many physical perpetrators that make it difficult to isolate the conduct of each accused.¹⁵⁷

As a consequence, this area of international criminal law has given rise to a degree of judicial innovation that has led to understandable critique and controversy.¹⁵⁸ Indeed, arguably, this area is the most contentious area of substantive international criminal law.¹⁵⁹ In sum, to take into account the manner in which superiors or individuals remote from the crimes actually operate, the ad hocs and the ICC have sought to develop expansive interpretations of the notion of commission. However, these hand-made developments have raised legitimate due process concerns, such as those revolving around fundamental principles of law such as nullum crimen sine lege and nulla poena sine lege.¹⁶⁰ The AU Court will have to grapple with these same issues. As will be discussed below, in light of the drafting, the path through many of these thorny issues is far from clear.

Firstly, it is important to note that international criminal law has not arrived at a universally accepted doctrine or approach to these collective criminal actions. Each attempt has been widely criticized and little agreement seems to exist on the most appropriate model to prosecute collective crimes. There are three main doctrines that have been used at the international level: conspiracy (inchoate crime), JCE and co-perpetration (modes of liability). These will be briefly considered below.

Conspiracy was introduced into international criminal law through the Nuremberg and Tokyo Charters.¹⁶¹ However, conspiracy is an inchoate crime (and not a mode of liability).¹⁶² It was a crime that assisted in linking ‘several individuals in one general criminal scheme, facilitating their prosecution and making it easier to obtain convictions against the alleged defendants.’¹⁶³ The Tokyo tribunal defined conspiracy to wage aggressive or unlawful war as an agreement by two or more persons to commit this crime.¹⁶⁴ The accused must have participated or contributed in the aggressive war. Additionally, the accused must have had knowledge of the conspiracy’s aggressive aims and the special intention to support the objects of the conspiracy.¹⁶⁵ Both the Nuremberg and Tokyo tribunals restricted conspiracy to crimes against peace and rejected its application to other crimes.¹⁶⁶ Conspiracy was later introduced in the ad hocs Statutes in relation to the crime of genocide.¹⁶⁷

In addition, Articles 25(3)(d) of the Rome Statute provides for a new accessory mode of liability for collective actions: the contribution to the commission or attempted commission of such a crime by a group of persons acting with a common purpose. This paragraph was adopted as a compromise with conspiracy and was taken from the 1998 Anti-terrorism Convention.¹⁶⁸ The drafters of the Rome Statute rejected the concept of conspiracy as an inchoate crime and instead adopted a concept of complicity in a group crime as a mode of participation in crime. Conspiracy was deemed as a ‘very divisive issue’ by the drafters.¹⁶⁹ In the Katanga judgement, the Trial Chamber noted that this accessory mode of liability was introduced in the Rome Statute in order to ensure that the accomplices whose conduct do not amount to aiding and abetting are prosecuted before the ICC.¹⁷⁰ It further explained that this mode of liability is not a form of JCE in so far as the accused is only liable for the crimes he contributed to the commission of and not all the crimes part of the common plan.¹⁷¹ Regarding the level of contribution, the ICC found that the individual criminal responsibility under Article 25(3)(d) needed to reach ‘a certain threshold of significance below which responsibility under this provision [did] not arise’.¹⁷² It further held that the contribution must be at least significant.¹⁷³

To hold criminally liable individuals committing collective crimes, the ad hocs developed a new mode of participation, the concept of JCE, a common law influenced doctrine¹⁷⁴ that attempted to capture the collective nature of international crimes. It is a form of commission to assign responsibility to individuals, who did not physically commit the criminal acts but acted with the intent to aid those who did, that arose from an expansive interpretation of the word ‘committing’ under Article 7(1) of the ICTY Statute.¹⁷⁵ In summary, ‘[w]hoever contributes to the commission of crimes by the group of persons or some members of the group, in execution of a common criminal purpose, may be held to be criminally liable, subject to certain conditions.’¹⁷⁶ All the participants are equally guilty of the crime regardless of the role each played in its commission.¹⁷⁷

The doctrine of JCE comprises three forms where accused have associated with other criminal persons, intended to commit a crime, joined others to achieve this goal and made a significant contribution to the commission of the crime. Thus, an individual can be held liable for the actions of other JCE members, or individuals used by them, that further the common criminal purpose (first category of JCE -basic) or criminal system (second category of JCE – systemic or ‘concentration camp cases’), or that are a natural and foreseeable consequence of the carrying out of this crime (third category of JCE – extended).

The three forms of JCE share the same actus reus, namely (i) a plurality of persons (ii) the existence of a common plan, design or purpose which amounts to or involves the commission of a crime provided for in the Statute (iii) the participation of the accused in the common plan involving the perpetration of one of the crimes provided for in the Statute (physical participation, assistance in, or contribution to, the execution of the common plan or purpose).¹⁷⁸

Regarding the mens rea, each form requires its own elements: JCE I requires proof that all participants shared the same criminal intent. It is necessary to establish that the accused voluntarily participated in the enterprise and intended the criminal result.¹⁷⁹ JCE II requires that the accused must have personal knowledge of the system of ill-treatment (whether proven by express testimony or inferred from the accused’s position of authority), as well as the intent to further this concerted system of ill-treatment.¹⁸⁰

For JCE III, a member of the joint criminal enterprise may be held liable for a crime or crimes which he did not physically perpetrate if, having the intent to participate in and further a common criminal design or enterprise, the commission of other criminal acts was a natural and foreseeable consequence of the execution of that enterprise, and, with the awareness that such crimes were a ‘natural and foreseeable’ consequence of the execution of that enterprise, he participated in that enterprise.¹⁸¹

Finally, the ICC has taken a different approach to these ‘joint action’ challenges. Instead of conspiracy as an inchoate crime or JCE as a mode of liability, the ICC has enunciated the notion of co-perpetration using the concept of control over the crime. This implies that principals to a crime are not limited to those who physically carry out the objective elements of the offence, but also include those who, in spite of being removed from the scene of the crime, control or mastermind its commission because they decide whether and how the offence will be committed.¹⁸² The ICC has also expanded this collective mode of liability to include indirect co-perpetration to capture the relationship between co-perpetrators who controlled separate militias, each committing crimes that were part of the common plan.¹⁸³

An in-depth analysis of the merits of each approach to joint action crimes is outside the confines of this Chapter. However, as noted above, there is extensive commentary examining each approach with critics of each and every approach.¹⁸⁴

In sum, conspiracy was extensively criticized and rapidly abandoned. The records of the international tribunals show that the prosecution of conspiracy proved to be a difficult task.¹⁸⁵ As a result, the tribunals adopted a strict approach to the conspiracy charge.¹⁸⁶ These narrow definitions failed to comprehensively encompass the criminal conduct and arguably created a system that permitted defendants to evade criminal responsibility for conduct deserving of it.¹⁸⁷

Regarding JCE, the lack of distinction between principals and accessories and the foreseeability requirement at the centre of JCE III are considered to be major problems. Commentators argue that the JCE doctrine systematically eviscerates the distinction between principals and accessories. All accused will be convicted of the same thing if they intended to contribute to the common plan.¹⁸⁸ Furthermore, it is correctly argued, JCE III endangers the principle of individual and culpable responsibility by introducing a form of collective liability, or guilt by association.¹⁸⁹ Convictions ultimately rest upon a lowered mens rea – a type of recklessness (dolus eventualis) and not a clear intent that the crimes be committed or awareness that those crimes were going to be committed.

The ICC concept of (indirect) co-perpetration is thought to reflect a more objective rationale than JCE.¹⁹⁰ Commentators argue that the participants’ contribution to a criminal endeavour is defined more precisely.¹⁹¹ Moreover, the concept maintains a distinction between principal and accessory.¹⁹² Only those who had control over the crime would be held liable as perpetrators, the others will be liable as accomplices.

However, a closer examination of these apparent benefits raises serious questions concerning whether the control theory has really improved upon the JCE doctrine in these culpability and legality aspects. Critics argue that the ‘control over the crime’ approach requires an ‘essential’ contribution of the perpetrator to the crimes, departing from the ‘significant’ contribution required for JCE. Commentators have highlighted the difficulty of assessing what constitutes the ‘essential contribution’, particularly that this ‘requires a hypothetical and nearly impossible counterfactual inquiry into whether the defendant’s behavior constituted an essential contribution to the crime’.¹⁹³

The mens rea requirements of co-perpetration also raise serious culpability issues that mirror some of the concerns with JCE III. Co-perpetrators ‘intend’ the crime if they are aware of the risk that the physical perpetrators will commit the offence and the co-perpetrators reconcile themselves to this risk or consent to it.¹⁹⁴ As stated by Ohlin, at most this is a form of recklessness/dolus eventualis, which closely resembles JCE III.¹⁹⁵ As discussed above, mere awareness even of a high risk that the crime will occur is not sufficient to found liability under JCE I and II.¹⁹⁶ As Ohlin has also correctly concluded, this approach consists of a ‘combination of awareness of joint control over the crime with an intentionality requirement that is so watered down that the control requirements appears to be doing all the heavy lifting’.¹⁹⁷

As may be seen from this brief discussion concerning commonly held due process critiques with regard to conspiracy, JCE and co-perpetration, the AU will be required to steer a path through these various approaches to design an appropriate and practical liability that links individuals to crimes of this nature whilst avoiding this entangled history of due process concerns.

However, the drafters of Article 28N have not provided the basis for a firm beginning. Article 28N appears to suffer from a range of problems that provides fertile ground for a range of confused judicial responses to these most complex problems. First, Article 28N does not appear to expressly opt for, or favour, any of these aforementioned approaches. As the ad hocs and the ICC have demonstrated, this fact alone is not an obstacle to developing expanded notions of commission to deal with joint action crimes. However, on the face of Article 28N, the drafters have hamstrung any future deliberation by failing provide a clear indication of what was intended or which option might best be employed at the future AU Court. Instead, the drafting leaves the door open for all of the above.

As discussed above, unlike the ICC, the ad hocs’ Statutes failed to articulate any mode of liability that encompassed joint action crimes. Instead, JCE was read into the statutes through a series of creative decisions at the trial and appellate level. Article 28N appears to suffer from the opposite problem and includes a range of definitions that might (or might not) be referencing expanded notions of commission or accessory with a view to encompassing joint action conduct. These include, participation as ‘a principal, co-principal, agent or accomplice’; as an ‘accessory before or after the fact’; or any individual that ‘in any other manner participates in a collaboration or conspiracy to commit any of the offences set forth in the present Statute’. It is therefore difficult, if not impossible, to assess what was in the drafters’ mind. The duplication of distinct notions such as the civil law notion of collaboration (also known as association) and the common law notion of conspiracy sitting alongside the conduct of ‘principals, co-principals, agent or accomplice’ is likely to challenge even the best of jurists and academicians, let alone those advocates struggling in the trenches of a future court room.

It appears that the inclusion of conspiracy and collaboration (association) in Article 28N was at least in part heavily influenced by Article 6(1)(b)(ii) of the UNTOC. Article 6(1)(b) states that State Parties shall establish as criminal offences the ‘participation in, association with or conspiracy to commit, attempts to commit and aiding, abetting, facilitating and counselling the commission of [laundering of proceeds of crime].’ However, the interpretative guide of the UNTOC explains that the two approaches were not introduced into the instrument with the expectation that both would be transposed into the same domestic law. It was to reflect the fact that some countries had conspiracy in their law, while others had criminal association (association de malfaiteurs) laws and effective transposition of the Convention at the domestic level involved respect for respective legal tradition and culture.¹⁹⁸

As a means of incorporating differing legal traditions, this latitude makes practical sense. However, including both in a statute, less so. The two concepts have different elements but essentially cover the same conduct. Conspiracy may be shown through mere proof of an intentional agreement to commit serious crimes for the purpose of obtaining a financial or other material benefit. Since most civil law countries do not recognize conspiracy or do not allow the criminalization of a mere agreement to commit an offence, association focuses on the conduct of the accused. It requires proof of the participation in criminal activities and the general knowledge of the criminal nature of the group or of at least one of its criminal activities or objectives.¹⁹⁹ If a person takes part in non-criminal action that nonetheless may be supportive of criminal activities, the knowledge that such involvement will contribute to the achievement of a criminal aim of the group will also need to be established.²⁰⁰

In addition to this duplication, the current reference in Article 28N of the AU Statute to ‘accessory before or after the fact’ and ‘participation in any other manner’ adds more repetition. Accessory before the fact traditionally encompasses ordering, soliciting or inducing.²⁰¹ Therefore, this provision appears to reiterate the accessory modes of liability already detailed in paragraph (i) of Article 28N.

In sum, not only is Article 28N duplicative and confused, it fails to offer any clarity as to what joint liabilities were intended or are favoured. It appears to do little more than leave the entirety of the interpretation of these complex issues to the (unfortunate) judges who will be forced to grapple with these issues in the course of future proceedings with little or no guidance of the drafters’ intent.

D. Corporate Criminal Liability

The AU Statute is the first to introduce the concept of corporate criminal liability in international criminal law.

Article 46C

Corporate Criminal Liability

1. 1. For the purpose of this Statute, the Court shall have jurisdiction over legal persons, with the exception of States.

2. 2. Corporate intention to commit an offence may be established by proof that it was the policy of the corporation to do the act which constituted the offence.

3. 3. A policy may be attributed to a corporation where it provides the most reasonable explanation of the conduct of that corporation.

4. 4. Corporate knowledge of the commission of an offence may be established by proof that the actual or constructive knowledge of the relevant information was possessed within the corporation.

5. 5. Knowledge may be possessed within a corporation even though the relevant information is divided between corporate personnel.

6. 6. The criminal responsibility of legal persons shall not exclude the criminal responsibility of natural persons who are perpetrators or accomplices in the same crimes.

Although it was discussed during the negotiation of the Rome Statute, the French proposal to include corporate criminal liability was rejected by the States.²⁰² However, several domestic regimes have granted their courts jurisdiction over international crimes committed by corporations. Corporate criminal liability has been recognized in the Anglo-American legal systems since the mid-90s and there has been progressive adoption of laws extending the court’s jurisdiction to companies in other legal systems in the last decades.²⁰³ Two surveys of national jurisdictions revealed that over twenty states in America, Europe, Asia, and Oceania (e.g. Australia, Belgium, Canada, France, India, Japan, the Netherlands, Norway, the United Kingdom and the United States) have adopted laws allowing the prosecution of corporate entities.²⁰⁴ Several African states have also adopted corporate criminal liability provisions, such as Ethiopia,²⁰⁵ Botswana,²⁰⁶ Kenya,²⁰⁷ Malawi,²⁰⁸ Namibia,²⁰⁹ Rwanda,²¹⁰ South Africa²¹¹ and Zimbabwe.²¹² As will be discussed below, there have been a variety of approaches with regard to the form and scope of the liability adopted, in sum, vicarious liability, the identification model and the ‘organizational’ liability framework. An analysis of the various models of criminal liability suggests that the drafters of the AU Statute appear to have intended to design a mode of liability that is close to the Australian ‘corporate culture’ approach which is a variant of the organizational liability approach. These issues will be discussed below.

1. The Various Models of Corporate Liability in Domestic Legislations

In the common and civil law legal systems, three main types of corporate liability may be distinguished. The common law variant is the vicarious liability, or respondeat superior, used in Austria, Ethiopia, Namibia, South Africa, the United States and Zimbabwe. Under this model, any crime committed by individual employees or agents are directly imputed to the corporation provided that the offence was committed in the course of their duties, and intended to benefit the corporation.²¹³ The actus reus and mens rea are therefore related to the employee and not the company. A company may however avoid liability by demonstrating that they put in place effective due diligence programmes.²¹⁴ For example, in Ethiopia, a corporation can be held liable if a crime has been committed by one of its director or employee in connection with the activities of the corporation.²¹⁵ The act of the director or the employee should have been committed with the intent of promoting the interest of the corporation by using unlawful means, by violating its legal duty or by unduly using the corporation as a means.²¹⁶

Another model is the identification model used in Canada, Rwanda and the United Kingdom. Under this model, only the crimes committed by individual senior officers and employees may be imputed to the corporation. The conduct and state of mind of these senior officers and employees is considered as that of the corporation. The definition of senior officer or employee, however, varies between the countries. For example, in the United Kingdom, directors and senior managers are the corporation’s ‘directing mind and will’.²¹⁷ These individuals are considered to be the embodiment of the company.²¹⁸ This theory has been widely criticized for being too restrictive and not representative of the horizontal or decentralized decision-making structure of many companies.²¹⁹

The final model is the ‘organizational’ liability. Under this model, ‘a corporation is liable because its “culture”, policies, practices, management or other characteristics encouraged or permitted the commission of the offence’.²²⁰ The liability of the company is not only limited to the acts of its employees, senior officials or agents but also applies to the ‘corporate culture’. These provisions are ‘arguably the most sophisticated model of corporate criminal liability in the world’.²²¹ Our research did not identify any African states countries with similar models. Australia appears to be the best example of this model,²²² and will be discussed further below in an attempt to shine light on Article 46C.

Under the Criminal Code of Australia, where an employee, agent or officer of a body corporate, acting within the actual or apparent scope of their employment, or within their actual or apparent authority, commits a crime, the actus reus must also be attributed to the body corporate.²²³ If intention, knowledge or recklessness is the requisite subjective element, it will only be attributed to the body corporate if that body corporate expressly, tacitly or impliedly authorized or permitted the commission of the offence.²²⁴ Authorization or permission for the commission of a crime may be established on four bases, including where ‘a corporate culture existed within the body corporate that directed, encouraged, tolerated or led to non-compliance’.²²⁵

The ‘corporate culture’ model seems to encompass the notion of policy included in the AU Statute. As noted, Article 46C of the AU Statute states that: ‘Corporate intention to commit an offence may be established by proof that it was the policy of the corporation to do the act which constituted the offence.’ This appears to suggest that a company will be directly liable for any criminal act committed in furtherance of the corporate policy.

In this regard, companies may be involved as a perpetrator in international crimes in various contexts. First, direct liability will exist where a company may directly take part in the crime as a perpetrator when the company’s general goal is to commit a crime (e.g. money laundering or trafficking in drugs) or indirectly, when, in accomplishing its economic objective, the company commits a crime with intent or knowledge (e.g. corruption, trafficking in persons).²²⁶ These will be discussed below.

2. Principal Corporate Liability

With regards to direct liability, the role of a company usually does not raise complex legal issues.²²⁷ The general principle of the principal liability developed in international criminal law may easily be applied. The company should have directly perpetrated the crime (through its employees, agents or officials) with knowledge and intent.

Intent may be established by proving that it was ‘the policy of the corporation to do the act which constituted the offence’.²²⁸ Policy is defined as ‘the most reasonable explanation of the conduct of that corporation’.²²⁹ As may be seen, this AU Statute requirement, however, raises a number of potential due process issues. The policy of a company may prove to be difficult to identify. Whilst a wide interpretation of the concept of policy may facilitate the prosecution of corporations, an overly expansive interpretation or acceptance of any reasonable explanation proffered by the Prosecution, will have a substantial impact on fair trial rights and ultimately the legitimacy of such prosecutions.

The Australian Criminal Code provides an interesting way of interpreting the notion of policy so as to ameliorate some of these concerns. To attribute the crime to the corporate culture, the authority to commit an offence should have been given by a high managerial agent of the body corporate. If not, the employee, agent or officer of the body corporate who committed the offence should have believed on reasonable grounds, or entertained a reasonable expectation, that a high managerial agent of the company would have authorised or permitted the commission of the offence.²³⁰

Turning now to the AU concept of knowledge, Article 46C states ‘corporate knowledge of the commission of an offence may be established by evidence that the actual or constructive knowledge of the relevant information was possessed within the corporation’. While the Statute requires the corporation to be aware of the crime, this knowledge does not have to be centralized and can be ‘divided between corporate personnel’.²³¹ This last characteristic of the knowledge seems to capture the reality of modern corporate decision-making, which tends to be more horizontal and decentralized. One aspect needs to be further considered: the recipient of the information. Article 46C does not define who should have the information: whether the concepts of ‘corporation’ and ‘corporate personnel’ may equate to a mere employee or must be senior officials. Another important question concerns whether it is enough if only one person possesses the required information. The Statute has left these essential points unanswered.

The Australian experience offers some insight into these issues. As outlined in the Criminal Code, a high managerial agent, and not merely any employee in the company, should possess knowledge – except if the latter reported the commission of the crime to a higher ranked agent or the information is widely known among the employees.²³²

Apart from these issues, Article 46C fails to define the physical element of corporate criminal liability. It seems to only require that the conduct reflected the corporation’s policy in order for it to be attributable to the company. However, it fails to explain whose action within the corporation may be attributable to the company (employees, agents, board of directors etc.) and the conditions for the attribution of responsibility (whether the particular actor acted in the course of their employment duties, etc.,). This vagueness may be contrasted with Division 12.2 of the Australian Criminal Code, which provides for a degree of specificity on these critical issues: in sum, the physical element of an offence committed by an employee, agent or officer of a body corporate acting within the actual or apparent scope of his or her employment, or within his or her actual or apparent authority, can be attributed to the body corporate.

3. Accomplice Corporate Liability

As a general proposition, apart from principal liability, companies may also be involved in international crimes as accomplices. Generally speaking, a company may contribute to a crime as an accomplice in three different ways. First, the company may act as a direct accomplice when it assists the perpetrators in the commission of the crime (e.g. assistance in the transportation of trafficked hazardous waste, financial contribution, or providing (raw or military) material or arms that will or are likely to be used for the commission a crime); as a beneficial accomplice when the company benefits from the crimes committed by the perpetrators (e.g. buying diamond, oil or any product whose production or extraction involved the commission of a crime)²³³; and finally, companies may act as a silent accomplice when they fail to ‘raise systematic or continuous human rights abuses with the appropriate authorities’²³⁴ (e.g. doing business with a government that has unconstitutionally taken power or with a group involved in drug trafficking). This last category will not be further considered since this complicity has more in common with moral rather than legal culpability. Such complicity does not generally engage criminal liability since the company is not involved in any manner in the commission of the crime.²³⁵

A comprehensive discussion of each form of accomplice liability as they might relate to corporations is outside the confines of this Chapter. However, as discussed, actus reus and mens rea requirements vary according to the particular mode of liability. In 2006, the International Commission of Jurists asked eight experts to explore when companies and their officials could be held legally responsible on the basis of accomplice liability. They concluded that aiding and abetting was the form of accomplice liability most relevant to the question of corporate conduct.²³⁶ As outlined above, this chapter seeks to open the discussion and identify preliminary concerns with regard to the various modes of liability in the AU Statute. Therefore, we will briefly discuss this vital accessory mode of liability and some of the problems that may arise in relation to holding corporations to account as accomplices on the basis of Article 46 C.

As discussed, customary international law requires that the aider and abettor made at least a substantial contribution to the principal’s act. The act of assistance must have had a substantial effect on the perpetration of the crime e.g., in the case of a company that provides weapons or logistics that enable the perpetrator to commit the crime. This type of action appears relatively straightforward but in practice has not proven to be so. In reality the term ‘substantial contribution’ is a ‘very indeterminate concept’ and the identification of the ‘relevant proximate causes (of the crimes) among the many causes will depend upon several aspects, including policy decisions’.²³⁷

This intermingling of causes increases the more the accused’s assistance is remote from the crimes. In contradiction to the specificity and certainty that is essential to ensuring respect for the principle of individual culpability, the accused risks being held liable on the basis of the effect of his assistance – namely on what use the principal makes of the aid given – rather than on the basis of his own acts and control. Therefore, convictions may rest on how much the principal used the accused’s assistance, which may be entirely beyond the aider’s control. The aider will be criminally liable if the perpetrator made significant use of his assistance – no matter how general and removed the assistance was from the criminality, even if the aider took all reasonable steps to prevent the aid being used in furtherance of criminality or intended it to promote only the lawful activities of the principal.

Any new AU Court interpreting Article 46C will need to address these thorny issues, not least of which will be whether, in cases where the aider and abettor is remote from the crimes, the ‘specific direction’ assessment (discussed above²³⁸) is an appropriate means of ensuring respect for the principle of individual culpability. As discussed above, according to many experienced commentators and courts, including the present authors, this element is required in cases of remote assistance to enable general assistance to the perpetrator and assistance that is directed specifically at the commission of the crime to be properly distinguished in the confines of complex trial processes. Accordingly, the actus reus of aiding and abetting may require sufficient proximity and the direct linkage between the aid provided and the relevant crimes.²³⁹

As discussed in this chapter, this debate touches on whether knowledge that the acts contribute to the commission of the crimes is the only mental element required to establish aiding and abetting. For example, if the AU Court adopts the ICTY and ICTR interpretation (when interpreting Article 46C), then a company officer that knows that the products he sells are likely to be used by the buyer to commit a war crime will be held liable as an aider and abettor, even if he did not intend to commit the crimes. On the other hand, the ICC’s mental element for aiding and abetting appears to require (at least something close to) intent and knowledge.²⁴⁰ Mere awareness that the accused’s assistance will be used for the commission of the crime will not be sufficient.²⁴¹ Adoption of the ICC’s approach will therefore place additional demands upon the Prosecution and make a conviction less likely. On the other hand, as the following domestic cases discussed below show, the application of the knowledge threshold to Article 46C would ensure a more all-encompassing liability approach but not necessarily one that stays on the right side of the principle of culpability.

The Dutch case of van Anraat is of particular relevance. Although the accused was the businessman and not the company, the findings of the Court provides an interesting insight into these issues and potential manifestations of corporate criminal liability. Van Anraat was charged with complicity in war crimes. He was accused of selling thiodiglycol (TDG) to Saddam Hussein’s regime – a chemical used to produce mustard gas. Van Anraat claimed that his chemicals were intended for the textile industry. While no findings about the purpose of facilitating the use of chemical weapons against civilians were found,²⁴² the Court, applying the ‘knowledge standard’ only, held that van Anraat ‘knew that the TDG which was supplied by him would serve for the production of poison/mustard gas in Iraq and that efforts were made to conceal that purpose’.²⁴³ Had the Dutch Court applied the ICC’s intent and knowledge standard, van Anraat would certainly not have been convicted.

In contrast, in the Presbyterian Church of Sudan v. Talisman Energy, Inc.²⁴⁴ and Aziz v. Alcolac cases,²⁴⁵ US courts have required both purpose and knowledge. In the Talisman case, a Canadian company was charged with aiding and abetting the Government of Sudan to advance human rights abuses that facilitated the development of Sudanese oil concessions by Talisman affiliates. The lower court held that it could not be established that Talisman acted with the intention to assist the violation of international human rights. On appeal, the Court relied on the elements of aiding and abetting under international law, as defined in Article 25(3)(c) of the Rome Statute, and concluded that ‘applying international law, we hold that the mens rea standard for aiding and abetting liability in ATS actions is purpose rather than knowledge alone’.²⁴⁶ Accordingly, the knowledge standard might have caused the court to reach a different conclusion.²⁴⁷ As noted by Finnin, ‘reliance on Article 25(3)(c) [instead of customary international law as defined by the ad hocs] is having a real and immediate impact on the scope of corporate liability for aiding and abetting international crimes’.²⁴⁸

In sum, the AU judges will be required to grapple with these difficult and oft argued issues and craft innovative answers to questions that will arise in the application of Article 46C. As this brief sojourn through the immediate issues shows, there is no certainty concerning the precise actus reus and mens rea elements and creative and thoughtful decisions are required if Article 46C is to live up to its exciting potential. Given the scale of the challenges, and the experience at the ICC and ad hocs to date, it is difficult to be too optimistic: Article 46C may well prove, at least for the early years of any AU Court, to be a triumph of good intention and hope over fairness and utility.

2. Conclusion

As discussed throughout this Chapter, the AU drafters have taken an extravagant approach to their enumeration of modes of liability. In an attempt to avoid accountability gaps, the AU Statute attempts to do too much and what emerges is a degree of imprecision and duplication that creates a high risk of unhelpful complexity and confusion. International courts need to learn a number of salutary lessons. In particular, as experience has shown, effective and efficient criminal adjudication of international crimes (or complex trials more generally) require clear, precise and distinct modes of liability. Anxious prosecutors will always use whatever is at their disposal, whether it makes for an efficient or fair process. Providing them with modes of liability beyond those that are strictly necessary may seem like a useful “belts and braces’ approach, but as experience has shown, it unlikely to assist with these essential objectives.

In this regard, Article 28N would undoubtedly benefit from a paired down approach informed by close attention to years of experience at the ad hocs and some from the ICC. Whilst the historic introduction of the concept of corporate criminal liability into international justice by way of Article 46C represents and exciting innovation at the international level, there is not much, if anything, to be gained by many of the other (additional) proposed modes of liability. Conversely, if efficient adjudication and judicial economy and consistency are worthy goals achievable through concrete and careful judicial process orientated steps, in many instances, there is much to be lost.

However, as noted above, much will also depend upon the skills and determination of the judges of the new AU Court. Inevitably, they must grapple with the challenge of interpreting their respective modes of liability in light of the objectives and principles of international criminal justice. One thing is for certain; the drafters have left them with a formidable task.