The Conflict in Eastern DRC and the State Responsibility of Rwanda and Uganda
In late January 2025, a rebel alliance involving the militia group M23 seized control over the town of Goma, the provincial capital of North Kivu in the Democratic Republic of the Congo (DRC). Notwithstanding their declaration of a unilateral ceasefire, at the time of writing the rebels appear to be advancing towards Bukavu, the provincial capital of South Kivu. Recent events represent a significant escalation of hostilities in the region.
The conflict could escalate further due to external support provided to the conflict parties. M23 operates with assistance from both Rwanda and Uganda, including direct combat support from the Rwandan armed forces. Meanwhile, Burundian forces bolster the DRC’s own troops, together with a mission from the Southern African Development Community (SADC). South Africa is one of the contributing nations to the SADC mission and has lost several of its soldiers during the recent hostilities.
An additional concern is the impact of the hostilities on the local civilian population. Since M23 first formed in 2012, humanitarian organizations and the UN have repeatedly highlighted the group’s egregious conduct in potential violation of the law of armed conflict (LOAC). During the group’s first military campaign, the UN High Commissioner for Human Rights identified M23’s leaders as “among the worst perpetrators of human rights violations in the DRC, or in the world.” The group’s conduct has not improved during its current operations. Recent allegations against the group include indiscriminate killing, rape, torture, and forced conscription.
This post considers the potential State responsibility of Rwanda and Uganda for M23’s conduct in potential violation of LOAC. It focuses first on Rwanda, due to Kigali’s more significant support to the group, considering the potential grounds on which M23’s conduct in violation of LOAC could be attributed to the State. It then assesses Uganda’s relationship with M23 and that State’s potential responsibility for M23 fighters’ conduct in breach of international law. The post does not consider Rwanda or Uganda’s responsibility for their own organs’ conduct in supporting M23, which may violate the prohibition on the use of force and other norms of international law.
M23 and Rwanda
The decades-long hostilities in the DRC began in the 1990s, triggered in part by the reverberating effects of the genocide in neighboring Rwanda. From the outset, the conflict has been characterized by the participation of numerous armed groups, acting with the benefit of significant external State support, and the commission of egregious abuses against the civilian population.
The rebel group M23 first emerged in 2012. Its full name is the Movement of 23 March; a reference to an agreement formalized on that date in 2009. The agreement granted another armed group, the National Congress for the Defense of the People (CNDP), the status of a political party and appointed its senior officers to key positions within the DRC’s national army. In early 2012, two ex-CNDP officers founded M23, arguing that the DRC government had failed to live up to the terms of the 2009 deal.
Rwanda was instrumental in the group’s creation and its early success. The State’s principal rationale for backing the group was its distrust of the Congolese authorities’ ability to adequately secure the border and protect Rwanda from the threat posed by ethnic Hutu militias (p. 14-15). In 2012, a UN Group of Experts tasked with examining the situation in the DRC produced two detailed and damning reports regarding Rwandan involvement in the conflict. The experts documented systematic State support for M23, including facilitation of the rebels’ transportation within Rwanda and into the DRC (p. 4-7), assistance with recruitment (p. 7-9), and instruction of recruits at M23 training camps (p. 11, para. 25).
For the remainder of 2012, violence escalated and M23 gained control over significant territory in eastern DRC including Goma. The following year, however, the group split into factions and was ultimately defeated. At that time, most M23 fighters fled to Uganda and Rwanda. The group then reemerged in November 2021, conducting attacks on Congolese army positions (p. 17). M23 once again made rapid progress, winning control over significant territory in eastern DRC. As with its previous military campaign, the group committed numerous acts in potential violation of international law. UN reports indicate that it also continued to receive significant assistance from Rwanda.
Importantly, during both of its military campaigns, M23 received direct combat support from the Rwandan Defense Forces (RDF). Recent UN reports assert, for instance, that the RDF “provided troop reinforcements to M23 with the aim of seizing or reinforcing strategic areas” (p. 16, para. 57) and employed advanced military equipment in support of M23 operations (p. 12-13). The UN concluded in December 2024 that 3,000 to 4,000 RDF troops remained deployed in eastern DRC in support of M23 operations (p. 12, para. 49).
Are M23 Fighters Acting under Rwanda’s Instructions, Direction or Control?
The UN Group of Experts’ reports document not only the significant assistance Rwanda provides to M23 but also the numerous abuses that M23 fighters have perpetrated against civilians, including killings, arbitrary arrests, torture, rape, and forced labor (p. 18-20). It is likely that many of these acts amount to war crimes, for which the fighters involved could bear individual criminal responsibility.
These same acts may also engage Rwanda’s international responsibility if the relevant conduct is attributable to it under the law of State responsibility. If the acts are attributable to Rwanda and they violate the State’s international legal obligations, they amount to internationally wrongful acts. Potential legal consequences include an obligation on Rwanda to cease the harmful conduct and to pay reparations for the injury caused (Articles on State Responsibility (ASR), arts. 28-39).
There are several grounds on which conduct performed by a non-State group can be attributed to a State (ASR, arts. 4, 5, 8-11). In the case of Rwanda and M23, perhaps the most obvious rule of attribution that could apply is that reflected in Article 8 ASR. This provides for the attribution to a State of conduct performed on its instructions, or under its direction or control.
The International Law Commission’s commentary to the ASR indicates that instructions, direction, and control are separate grounds of attribution; “it is sufficient to establish any one of them” (ASR, art. 8, commentary para. 7). First, State instructions can lead to attribution of conduct to a State if there is evidence that State officials issued some form of order or command to a non-State actor to act in a manner that violates the State’s international legal obligations. One recent UN report seems to support attribution on this basis, stating that M23’s commander received “instructions … from RDF and Rwandan intelligence” (p. 11, para. 39). However, evidence that Rwandan officials provided general instructions to the group is insufficient to lead to attribution; the instructions must relate to the particular conduct on the part of the non-State actor that violates LOAC (ASR, art. 8, commentary para. 7). Thus, for the revenge killings M23 conducted against civilians to be attributable to Rwanda on this basis (p. 18, para. 66), evidence would be needed that Rwandan officials instructed M23 fighters to commit these specific acts.
In practice, evidence that State officials instructed members of an armed group to act in violation of international law is rarely available. It may therefore be easier for injured States to establish that the fighters were acting under Rwanda’s direction or control at the relevant time. Again, the UN’s reporting provides some support for attribution on this basis, asserting that “Every M23 unit was supervised and supported by RDF special forces” (p. 12, para. 49).
The appropriate control test that applies to determine attribution on this basis has been the subject of considerable debate. On one view, the test should be the same as that used to decide when a conflict is internationalized through a State’s support to a non-State armed group. This overall control test (paras. 116-145) focuses on the level of support the State provides to the group and the State’s involvement in the coordination and planning of its activities.
A previous post on Articles of War considered whether the non-international armed conflict (NIAC) between M23 and the DRC has been internationalized through Rwanda’s support to the group. The authors concluded based on public information available at the time that Rwanda did exercise overall control over M23, meaning that the NIAC between M23 and the DRC was transformed into an international armed conflict between the DRC and Rwanda. More recent reporting supports this conclusion. Therefore, if overall control is also the appropriate test to determine attribution for the purposes of Article 8 ASR, all M23’s conduct performed under Rwanda’s overall control is attributable the State.
The International Court of Justice (ICJ), however, has rejected this test for the purposes of attribution. In its Bosnian Genocide judgment (paras. 396-406), the ICJ reaffirmed the test of effective control it had previously established in Paramilitary Activities (para. 115). This is a much stricter test that requires evidence of State control over the specific acts on the part of the non-State actor that violate international law.
To illustrate, consider the killings by M23 fighters of individuals who were clearly hors de combat, as they had their arms tied behind their backs (p. 19, para. 67). If members of the RDF exerted tactical control over M23 during those attacks, the fighters’ conduct, including their acts in violation of LOAC, are attributable to the State. But if Rwandan armed forces’ involvement was limited to a broader level of oversight over the fighters’ use of force, or they did not exert tactical control over the particular attacks during which these killings occurred, the effective control threshold might not be satisfied. That is the case even if Rwandan officers played a significant role in the planning of the operation, they knew that the fighters might act in a manner that violated LOAC, and they had the authority to intervene and stop such violations but failed to do so.
As this example demonstrates, the effective control threshold is very high, and injured States will often experience difficulties in securing the necessary evidence to satisfy its requirements. Given the very strong relationship between Rwanda and M23, it may be that some of the group’s conduct in violation of LOAC is attributable to Rwanda on this basis. However, this attribution assessment would need to be carried out on a case-by-case basis, regarding each and every M23 operation during which the group acted in a manner that violated international law. It follows that there might be insufficient evidence to support attribution regarding all M23’s acts in potential violation of LOAC.
This outcome is potentially problematic, particularly if the conflict between M23 and the DRC has been internationalized through Rwanda’s support to the group. It would mean that Rwanda is deemed to be acting through the group for the purposes of conflict classification, meaning that all the rules of LOAC that are relevant to international armed conflicts apply, yet M23’s conduct in violation of Rwanda’s LOAC obligations might not be attributable to the State.
It is therefore relevant to consider whether any of the other rules of attribution reflected in the ASR could attribute M23’s conduct to Rwanda. While the rule enshrined in Article 8 ASR has a narrow scope of application, attributing to a State only those acts that are performed under its instructions, direction or control, other rules have a broader focus, which is replicated in the effects of attribution. Given the very close ties between M23 and Rwanda, it is pertinent to assess whether M23 could qualify as a de facto State organ, meaning that all its conduct performed in that capacity is attributable to the State.
Is M23 a de facto State Organ of Rwanda?
Normally, organs of a State are designated as such under the State’s domestic laws. However, the commentary to Article 4 ASR, which relates to the attribution to a State of the conduct of its organs, provides that “a State cannot avoid responsibility for the conduct of a body which does in truth act as one of its organs merely by denying it that status under its own law” (ASR, art. 4, commentary para. 11). Accordingly, there are two categories of State organ: de jure organs, which have that status under the State’s internal laws; and de facto organs, which in fact act as an organ of State. While M23 has not been integrated into the RDF as a matter of law, it is relevant to consider whether the group could nevertheless qualify as a de facto organ of Rwanda.
The ICJ addressed de facto State organ status in its Paramilitary Activities and Bosnian Genocide judgments. To qualify as such, the Court stated that the relationship between the entity and the State must be “so much one of dependence on the one side and control on the other” that it would be right to equate the entity, for legal purposes, with an organ of State (Paramilitary Activities, para. 109). In other words, the non-State forces must be “mere instruments through which the [State] was acting, and … lacking any real autonomy” (Bosnian Genocide, para. 394).
The ICJ referred to several factual circumstances that may assist in determining whether an entity has gained the status of a de facto State organ, some of which are evident in the relationship between Rwanda and M23. The first is that the State created the entity (Paramilitary Activities, paras. 93-94). In this respect, the UN Group of Experts concluded that “Rwandan officials … coordinated the creation of the rebel movement and its major military operations” (p. 6). A second factor of relevance is that the State selected, installed, and paid the group’s political leaders (Paramilitary Activities, para. 112). Here, the UN reports found that “Rwandan officials nominated the political leadership and government of M23,” and raised funds for the group within Rwanda (p. 12).
A further factor of relevance is the State’s involvement in devising the strategy and directing the tactics of the group (Paramilitary Activities, para. 110). Regarding M23’s first campaign, the UN Group of experts concluded that “Rwandan officials [exercised] overall command and strategic planning for M23” (p. 13). The UN’s more recent reporting refers to a similarly high level of Rwandan control over M23 military operations (p. 12, para. 49). Rwanda not only supports the group logistically, with significant supplies of military and non-military equipment, but the RDF is also heavily involved in the group’s military operations.
These circumstances demonstrate the close links between Rwanda and M23, but they may be insufficient to conclude that M23 qualifies as a de facto State organ. The ICJ’s test requires a relationship of complete dependence and strict control, meaning that M23 is entirely reliant upon Rwanda and has no real autonomy of its own. Thus, M23 can only qualify as a de facto State organ if it is, in all respects other than its legal status, effectively a branch of the Rwandan armed forces, subject to the same degree and quality of State control as the RDF, both in terms of its subjection to orders and its reliance on the State.
While in many respects it appears that this stringent threshold may be satisfied, other factors point away from that conclusion. In particular, reporting indicates that M23 has other sources of income, including from the illegal exploitation of natural resources in eastern DRC. In addition, the group receives assistance not only from Rwanda but also from Uganda. These facts suggest that M23 may have “some qualified, but real, margin of independence” from Rwanda (Bosnian Genocide, para. 394). On balance, therefore, despite the significant support the Rwandan government provides to M23 and the high level of control Kigali exercises over the group’s operations, it is far from certain that the group qualifies as a de facto State organ of Rwanda. If it does not, M23’s conduct is not attributable to Rwanda on this basis.
Does Uganda Bear State Responsibility for M23’s Actions?
In common with its neighbor, Rwanda, Uganda’s support to rebels in eastern DRC has spanned decades. Through its interventions, Uganda sought to minimize the threat from hostile armed groups operating near its western border and also profited economically from the DRC’s vast mineral wealth. Although Uganda’s support to M23 is less significant than that provided by Rwanda, the authorities in Kampala have nevertheless played a critical role in facilitating the group’s activities in eastern DRC.
During M23’s first military campaign, Ugandan officials permitted M23 to establish a permanent presence in Kampala and provided the group with political advice and technical assistance. In addition, the Ugandan military provided significant support to the group, including assistance with recruitment, participation in the planning of M23’s operations, and the provision of military advice (p. 13-18). When the conflict reignited in 2021, Uganda again assisted the group by allowing M23 fighters to freely move across its territory (p. 12). Officials from the Ugandan military and military intelligence also provided more active support, including by hosting meetings attended by M23’s leaders.
When considering Uganda’s potential responsibility for M23’s conduct in violation of LOAC, the same considerations apply as outlined above in respect of Rwanda. In the case of Uganda, however, it is less likely that the relevant attribution thresholds can be satisfied due to the more limited evidence regarding Uganda’s involvement in the group’s military operations. Thus, it is highly unlikely that M23 qualifies as a de facto State organ of Uganda due to the absence of a relationship of complete dependence and strict control. It may also be difficult for an injured State to secure evidence that M23 fighters were acting on the instructions of Ugandan officials at the time they committed acts in violation of LOAC, or under the State’s effective control. In contrast with Rwanda, Uganda does not exert control over M23’s activities but rather provides logistical and other support towards its military campaign.
Through its assistance to the group, it is likely that Uganda knowingly facilitated M23’s conduct in violation of LOAC. However, the ASR does not include any mechanism by which a State can bear responsibility for its aid or assistance to a non-State actor. The rule reflected in Article 16 ASR applies only to a State’s support to another State. This provides that a State which aids or assists another State in the commission of an internationally wrongful act bears responsibility for doing so if it has knowledge of the circumstances of the act, and the act would be internationally wrongful for that State. Therefore, if any of M23’s conduct in violation of LOAC is attributable to Rwanda, Uganda could bear international responsibility for its own contributions towards those acts. However, if the relevant conduct is not attributable to Rwanda (or to any other State), there is no internationally wrongful act for Uganda to assist, meaning that Uganda’s responsibility is not engaged.
Concluding Thoughts
The conflict in eastern DRC highlights the prevalence of armed groups on the battlefield, as well as the harm that those groups frequently cause. It also provides an example of a particularly strong relationship between a State and a non-State armed group. Often, States seek to obfuscate their links to armed groups engaged in conflict, and Rwanda’s President has repeatedly denied or evaded questions regarding the State’s support to M23. Nevertheless, the UN reports provide evidence of significant Rwandan involvement in the conflict. If substantiated, this could lead to Rwanda’s international responsibility for conduct on the part of M23 that violates LOAC.
The hostilities in eastern DRC nevertheless bring to light concerns regarding international law’s effectiveness in holding States to account when they choose to act through armed groups. One issue relates to Uganda’s role in the conflict and the State’s potential responsibility for its assistance to M23. If M23’s conduct in violation of LOAC is attributable to Rwanda, Uganda could bear international responsibility for its aid and assistance towards those acts. However, if the relevant conduct cannot be attributed to Rwanda or another State, the law of State responsibility as expressed in the ASR does not provide any mechanism by which Uganda can be held to account for its assistance to M23. This makes little logical sense. Why should States bear international responsibility for their assistance to other States in violating international law but not for the similar support they provide to a non-State armed group? While customary international law may be evolving to address this concern (see Bosnian Genocide, para. 419-420), the State-centric focus of the ASR fails to account for the ever-increasing importance of armed groups, both in the conflict in the DRC and elsewhere around the world.
Another concern is the lack of coherence between the tests for determining attribution under the law of State responsibility and the internationalization of conflicts under LOAC. If it is correct that Rwanda exercises overall control over M23, the conflict is international in character and the full panoply of LOAC obligations applies to the conflict parties. Yet because different, more stringent, tests apply to determine the attribution of M23’s conduct to the State, Rwanda might not bear international responsibility for all the group’s acts that violate its duties under LOAC. This again makes little practical sense and leads to arguments (such as here, paras. 265-73) that the same test should apply for both purposes.
***
Jennifer Maddocks is an Assistant Professor in the Department of Law and Philosophy at the United States Military Academy, West Point and the Managing Editor for Articles of War.
Photo credit: Al Jazeera
