Beyond the 'mercenary' label: Why international law is failing to regulate modern private armies

Since 2017, Russian private military actors – particularly the Wagner Group and, more recently, its successor often referred to as the “Africa Corps” – have become embedded in several African conflict settings. Their deployments have been linked to serious allegations of abuse, including torture, arbitrary detention, mass killings, and indiscriminate attacks on civilians in Mali, the Central African Republic (CAR), Sudan, Libya, and Mozambique.

Yet, securing accountability has proved difficult. Many alleged victims are citizens of the very states that invited these forces, making domestic prosecutions politically unlikely and shifting attention towards international mechanisms. In October 2025, for example, Larochelle Avocats filed an Article 15 submission to the International Criminal Court on behalf of Central African citizens, alleging that Wagner and the regime of former Prime Minister Faustin-Archange Touadéra committed war crimes and crimes against humanity as part of a state policy to keep Touadéra in power in exchange for control over mining sites.

Wagner is not the only private military company linked to human rights abuses or serving foreign policy ends of intervening states. Nonetheless, the scale and political centrality of its operations have reignited a foundational question: who counts as a mercenary today?

The Return of the Mercenary Question

The mercenary enterprise predates the emergence of national armies in the modern Westphalian state. Debates about mercenarism in Africa can be traced back at least to 1960, when UN General Assembly Resolution 1514 on decolonization heightened concerns among African states pursuing self-determination. Mercenarism was seen as intertwined with anti-liberation efforts, reinforcing the perceived need to restrain such armed actors.

Even as private force evolved into contemporary private military and security companies (PMSCs), the core legal and political dilemmas did not disappear. Instead, they sharpened: who should be tried, where, and under whose responsibility? The most persistent fault line has been definitional. When mercenarism is framed as a crime under dedicated anti-mercenary treaties – most notably the 1977 OAU Convention for the Elimination of Mercenarism in Africa and the 1989 International Convention against the Recruitment, Use, Financing and Training of Mercenaries – enforcement rises or falls on a threshold question: who qualifies as a “mercenary”?

Although important differences exist between mercenaries and PMSCs, the treaty-based definition is the product of compromise among states and is central to enforcement. Even so, it often fails to capture modern private military actors. In practice, these actors can structure their operations to fall outside the definition while still performing mercenary-like functions. In our research reassessing the existing regulatory framework, with particular attention to African contexts where these actors are most visible, we argue that this definitional problem, together with the limited scope and uneven ratification of these conventions, underscores the need for a more comprehensive and inclusive approach.

While the treaty definition is a brittle foundation for prosecution across the diverse conflict and post-conflict settings in which PMSC now operate, other pathways to accountability exist, including individual criminal liability, civil and corporate liability, and state responsibility. In practice, however, these pathways remain underused. They depend on political will that is frequently absent when powerful interests are implicated or when host states perceive such actors as strategically useful.

Meanwhile, prominent self-regulatory initiatives favored by many Western states, such as the Montreux Document and the ICoC/ICoCA ecosystem, have not closed the accountability gap. While they offer valuable standards and oversight expectations, their largely non-binding character and uneven uptake limit their capacity to constrain harmful conduct or deliver remedies where violations occur.

The central point is this: a narrow definition weakens accountability by shrinking the reach of mercenary law. As a result, groups such as the Wagner Group continue to operate without leash, a well-documented reality.

Crucially, this definitional problem has not been accidental. Like the broader deficit in prosecution, the narrowness of the legal definition reflects the politics of law itself. Law, in its origin, development, and application is rarely neutral; it is often shaped by contested interests. Put simply, how can we diagnose “enforcement failures” when we have not yet agreed on what, exactly, is meant to be enforced?

Politics in Legal Framing

Historical mercenary trials and UN deliberations during the decolonization era demonstrate the inseparable relationship between politics and law. Consider first Rolf Steiner, the German-born soldier tried in Sudan in 1971 in what the Sudanese government called “the Nuremberg trial of Africa.” Although the proceedings invoked the language of international wrongdoing, the Khartoum People’s Court acknowledged that mercenarism was not, at that time, an international crime and convicted Steiner under domestic law. He was found guilty of preparing to wage war against the Sudanese government along with related offenses. Notably, the Sudanese court did not treat the “mercenary” label as decisive: it stated that the provisions it was applying began with “whosoever,” meaning they could apply to a mercenary or anyone else. Thus, in the court’s view, the legal question turned on what Steiner did, not on his motives.

The Angolan proceedings of July 1976 extended this conduct-based logic further. Thirteen soldiers were tried by the People’s Revolutionary Court of Angola as mercenaries, resulting in prison sentences for nine and death sentences for four, amid denunciations that mercenarism “must be finished on this planet.” The trial drew intense Western criticism, with objections focusing on the lack of a presumption of innocence, lack of due process, and the argument that “mercenarism” was not a clearly established crime in international or Angolan law at the time. Yet the proceedings are significant for how they foregrounded the conduct question – who committed the crime of mercenarism? Under that framing, liability need not be confined to those who fight. It can extend to organizers, financiers, trainers, suppliers, recruiters – and potentially state actors who enable or direct such activity. The Diplock Report, which examined the principal legal issues surrounding mercenaries and their recruitment, adopted a similar stance: mercenaries “can only be defined by reference to what they do, and not by reference to why they do it.” This conduct-centered logic became an important template for Africa’s treaty approach, culminating in the Organization of African Unity (OAU) Convention for the Elimination of Mercenarism in Africa (1977).

From status to crime

Sadly, Africa’s emerging effort to treat mercenarism as a serious international crime soon became entangled in hegemonic contestation. During negotiations leading to Article 47 of Additional Protocol I (AP I) to the Geneva Conventions (1977), adopted a month before the OAU Convention, states settled on the first widely recognized international definition of a “mercenary,” but stopped short of creating a corresponding international crime. This was the first legislative bottleneck: the law addressed the status question without attaching direct criminal liability. Later instruments, including the 1989 International Convention against the Recruitment, Use, Financing and Training of Mercenaries, did criminalize mercenarism. Yet their practical reach has been weakened by limited ratification and inadequate oversight.

More fundamentally, liability often depends on a cumulative six-part test that is notoriously difficult to establish. A “mercenary” must be specially recruited to fight, participate directly in hostilities, be motivated essentially by private gain (typically linked to material compensation), be foreign to the conflict (neither a national nor a resident of the relevant territory), and not be formally incorporated into the armed forces of a party to the conflict, nor act under official state orders.

Some elements are easier to establish than others. The “foreignness” and “non-membership” criteria are straightforward, though even these conditions can be circumvented with relative ease. Papua New Guinea’s 1997 Sandline contract which required personnel to be enrolled as ‘Special Constables’ and carry ID cards illustrates how deputizing them into a state security role helped blunt the ‘mercenary’ label.”

The more significant obstacles lie elsewhere. First, proving that participation was motivated “essentially” by private gain is inherently challenging, as it requires adjudicating subjective motivation that may overlap with ideological commitment, political purpose, group loyalty, or adventurism. Second, demonstrating the requisite “material compensation” or benefit is increasingly complex in contemporary proxy and corporate arrangements, where remuneration may be indirect, opaque, or channelled through intermediaries.

Finally, where plausible claims of state direction or quasi-official deployment exist, the “official state orders” exclusion becomes difficult to litigate. The cumulative effect is that the law built a label with a lock – then made the key difficult to use. The more pressing question, then, is how this can be addressed.

Rethinking mercenarism: a conduct-focused definition

Calls for a broader international instrument to regulate PMSCs reflect growing recognition that the existing framework is inadequate. Since 2017, this has been taken up through an OHCHR-hosted intergovernmental working group, most recently in Geneva in December 2025 which culminated in the fifth draft of an international regulatory framework for PMSCs and mercenary-related actors. Yet these renewed discussions reopen a more fundamental question: how should mercenarism be defined?

If the priority is to balance the state’s monopoly over legitimate force, protect civilians – especially in African contexts – and enable meaningful accountability, a starting point is to acknowledge that mercenarism is not a fixed category but a socially constructed one. This requires making explicit the normative choices that underpin law-making – namely, which uses of force are acceptable, which may legitimately be outsourced, and which should remain beyond outsourcing.

On this view, the definition should address the regulatory “mischief:” the market-enabled outsourcing of organized coercive force that enables political violence, undermines sovereignty and self-determination, and increases the risk of serious abuses while evading responsibility. This suggests a conduct-based offense targeting the enabling system, rather than a fragile inquiry into whether a particular actor fits a narrow definition. Such an approach aligns with earlier African mercenary trials that prioritised conduct over labels.

Institutionally, reform pathways exist. Additional Protocol I includes an amendment procedure in Article 97. While the OHCHR-hosted intergovernmental working group cannot amend AP I, it can catalyse such a state-driven reconsideration of how mercenarism is framed in international law. Decades on, states once again have an opportunity to prove that tackling mercenarism is more than just rhetoric.

Reform without collapse 

Alternatively, if the existing test is retained, reducing evidential paralysis is possible. For instance, rather than removing nationality and residency requirements, “foreignness” could adopt a coherent limiting logic that targets the specific phenomenon at issue: market-driven participation in organised violence outside accountable state command.

“Private gain” could be clarified to encompass material benefit in a broader sense, including indirect profit streams – such as mining rights, revenue-sharing arrangements, political influence convertible into economic advantage, and comparable inducements. These would more accurately reflect contemporary incentive structures without abandoning the requirement of evidentiary discipline.

“Direct participation” could be clarified, for instance by referencing the 
International Committee of the Red Cross’s interpretive guidance. However, care must be taken to avoid stretching “direct participation” so far that it collapses the distinction between fighters and support personnel. The OAU Convention already criminalises in Article 1(2), the organising, financing, training, assistance, and other enabling conduct, capturing much of the operational ecosystem without redefining “fighting” itself.

Finally, the cumulative test can be made more workable by recalibrating the burden of evidence – without undermining the presumption of innocence. If the prosecution proves beyond reasonable doubt a set of objective indicators – such as special recruitment to fight, direct participation in hostilities, and foreignness (if relevant – then the accused may be required to present evidence that an exclusion applies, such as formal membership in the armed forces or deployment on official duty by a state. If the accused meets that evidentiary threshold, the prosecution should remain obliged to disprove the exclusion beyond reasonable doubt. This structure maintains the prosecution’s ultimate burden and the accused’s entitlement to the benefit of doubt, while reducing the practical impossibility of proving “status negatives” in situations where actors operate through deniable, proxy-driven contexts.

Notwithstanding these suggestions, accountability for past conduct need not depend on the mercenary label. Finland’s 2025 exercise of universal jurisdiction to convict Yan Petrovsky for war crimes is instructive. Finnish Criminal Code does not criminalise mercenarism as a standalone offence and Finland is not party to the International Convention against the Recruitment, Use, Financing and Training of Mercenaries. Still, Finland is only one jurisdiction among many, and it remains uncertain whether other states will replicate this approach at scale.

Looking ahead, the most urgent task is prospective (de lege ferenda) reform – particularly for African states disproportionately affected by mercenary activity and externally sponsored private-force interventions.

Definitional reform is not merely technical. It depends on how law is framed, whose interests that framing serves, and which forms of outsourced violence are made visible to regulation and prosecution. Future work must be complemented by a research agenda drawing on Critical Legal Studies and Legal Realism to examine how the “mercenary” definition is constructed, contested, and operationalised across treaty negotiations, domestic proceedings, and regulatory initiatives, and how these framing contests distribute stigma, immunity, and exposure to liability among states, intermediaries, corporate actors, and personnel. Ultimately, the issue involves not only definitional refinement but also the legal allocation of responsibility for violence committed under the shadow of state authority.

***

Oluwaseyi Mike Bamigboye is a lawyer and a DUPRS-funded doctoral researcher at Deakin University, Australia. His research focuses on mercenaries, private military and security companies (PMSCs), and the outsourcing of military logistics. He is also a Junior Fellow at the Bill Graham Centre for Contemporary International History at Trinity College, University of Toronto.

Christopher Kinsey is Professor of International Security in the Defence Studies Department at King’s College London.