The spread of proxy warfare in the twenty-first century has fundamentally altered the landscape of armed conflict, presenting unprecedented challenges to international law and accountability. States increasingly prefer indirect engagement through non-state armed groups, enabling them to pursue strategic objectives while avoiding the political, military and human costs of direct intervention.1 Africa has been particularly susceptible to this form of warfare, with empirical research finding that 96% of African states have sponsored proxies outside their own territories.2
The Democratic Republic of the Congo (DRC) has endured nearly three decades of protracted conflict characterised by the involvement of multiple armed groups and foreign state actors.3 The M23 rebellion, which emerged in 2012 and has since become one of the most significant armed groups in eastern DRC, has, according to the UN Security Council and the UN Group of Experts, received substantial military, logistical and strategic support from the Rwanda Defence Force (RDF), which the Council has called on to cease that support and withdraw from Congolese territory.4 The Group of Experts has documented this support extensively, including direct troop involvement, weapons supply, training, intelligence sharing and strategic coordination.5
Despite this extensive evidentiary record, the current legal framework presents significant barriers to holding sponsoring states accountable for violations committed through proxy forces. Article 8 of the Articles on Responsibility of States for Internationally Wrongful Acts (ARSIWA) provides that “[t]he conduct of a person or group of persons shall be considered an act of a State under international law if the person or group of persons is in fact acting on the instructions of, or under the direction or control of, that State in carrying out the conduct.”6 However, the provision deliberately refrains from specifying the degree of control required, which has generated significant jurisprudential fragmentation between competing attribution tests.7
This article examines how sponsoring states exploit these attribution thresholds to evade accountability, focusing on the Rwanda-M23 relationship in eastern DRC. It evaluates the effectiveness of the current legal framework in addressing contemporary proxy warfare and proposes reforms to close the accountability gap.
This article employs a doctrinal legal research methodology, a systematic approach that involves the exposition, analysis and critique of the legal rules and principles governing a particular legal category.8 Doctrinal research has been defined as a systematic exposition of the rules governing a particular legal category, analysis of the relationship between rules, explanation of areas of difficulty and prediction of future developments.9 It relies primarily on logical analysis of legal provisions and judicial decisions, conducted as desk or library research without empirical fieldwork.10 The choice of this methodology is justified for three reasons. First, the research questions are questions of legal interpretation rather than of human behaviour, and doctrinal analysis is designed to answer such questions.11 Second, the purpose of the research is to interpret the rules of attribution, identify inconsistencies between competing control tests and propose legal reforms, which aligns squarely with the aims of doctrinal research.12 Third, the ongoing conflict and insecure conditions in eastern DRC made field access impossible, and all the necessary material exists in published judicial decisions and United Nations reports accessible through library databases, making doctrinal methodology the only practical approach.13
The primary sources analysed include ARSIWA, particularly Article 8, which provides the legal basis for attributing the conduct of non-state armed groups to states.14 Judicial decisions of the International Court of Justice, including Nicaragua15 and Bosnia and Herzegovina v. Serbia and Montenegro,16 and of the International Criminal Tribunal for the former Yugoslavia, including Tadić,17 constitute the core jurisprudential sources. Secondary sources include scholarly articles, books and the reports of the United Nations Group of Experts on the DRC documenting Rwandan support for the M23 rebellion.
The effective control test emerged from the judgment of the International Court of Justice (ICJ) in Nicaragua v. United States. The Court held that, for the conduct of a non-state armed group to give rise to the legal responsibility of a state, “it would in principle have to be proved that that State had effective control of the military or paramilitary operations in the course of which the alleged violations were committed.”18 This required demonstrating that the state had issued specific instructions for, or directed, each individual operation.19
The ICJ’s effective control test established a high threshold for attribution. The Court distinguished between general support for a rebel group and the specific control required to hold a state responsible for particular violations.20 This distinction reflected the Court’s concern that states should be held responsible only for conduct with which there is a sufficiently close link.21 The Court subsequently reaffirmed the effective control test in Bosnia and Herzegovina v. Serbia and Montenegro in 2007, expressly rejecting the lower threshold developed by the International Criminal Tribunal for the former Yugoslavia (ICTY).22
The overall control test was developed by the Appeals Chamber of the ICTY in Prosecutor v. Tadić. The Appeals Chamber rejected the ICJ’s effective control standard as inconsistent with the logic of the law of state responsibility and at variance with judicial and state practice.23 Instead, it held that the required degree of control is lower where organised military or paramilitary groups are concerned. Under the Tadić standard, a group will be considered a de facto organ of a state when “the State wields overall control over the group, not only by equipping and financing the group, but also by coordinating or helping in the general planning of its military activity.”24 This test does not require proof of instructions for each specific operation.25
The significance of Tadić extends beyond the jurisprudence of the ICTY. It fundamentally altered the debate on attributing a proxy’s conduct to the sponsoring state by proposing a more flexible approach to attribution that better reflects the organisational realities of proxy warfare.26 While the ICJ focused on operational control over specific military actions, the ICTY emphasised the broader relationship between a sponsoring state and an organised armed group.
The divergence between the effective control and overall control tests creates an accountability gap that sponsoring states exploit. The two conflicting tests have created a legal gap that gives states considerable freedom to fight wars by means of proxies while avoiding responsibility. The existence of multiple standards, each with different requirements and interpretations, allows sponsoring states to design their proxy relationships to fall between the cracks of the legal framework.27
The ICJ in Bosnia and Herzegovina v. Serbia and Montenegro expressly rejected the overall control test, holding that it “has the major drawback of broadening the scope of State responsibility well beyond the fundamental principle governing the law of international responsibility: a State is responsible only for its own conduct.”28 This rejection deepens the jurisprudential division and creates significant legal uncertainty for accountability in contemporary proxy warfare.
The primary strategy employed by sponsoring states to evade attribution is the maintenance of indirect control.29 This involves structuring proxy relationships to avoid the kinds of direct operational command that would trigger attribution under the effective control test while retaining sufficient influence to achieve strategic objectives.30
This is achieved through several mechanisms. First, states avoid formal agreements or contracts with proxy groups that would establish explicit command-and-control relationships.31 Second, states use intermediaries, including intelligence agencies, private military contractors and allied third states, to channel support to proxies, concealing the direct link between the sponsoring state and the armed group. Third, states provide support through covert means that are difficult to detect and verify, including encrypted communications, clandestine supply chains and deniable assets.32
The UN Group of Experts has documented Rwandan support for M23 through both covert channels and formal state institutions, including the Rwandan army and its intelligence services.33 This dual structure complicates attribution: although the involvement is extensive, its covert dimensions make it difficult to establish the direct operational link required by the effective control test.34
The effective control test imposes an extremely high evidentiary burden on those seeking to attribute conduct to a sponsoring state. The ICJ’s requirement of proof of effective control over the specific operations in the course of which the alleged violations were committed, and of direction of each individual operation, is almost impossible to satisfy in the context of covert proxy warfare, where states deliberately avoid creating evidence of their involvement.35
Sponsoring states exploit this evidentiary burden by ensuring that their support to proxies operates through channels that minimise documentary evidence and through deniable intermediaries. The Group of Experts has reported, for example, that after the capture of Goma in January 2025 the footprint of the Rwanda Defence Force was reduced to the minimum necessary in order to maintain discretion about the involvement of Rwanda, and that RDF troops remaining in M23-controlled territory were deployed at strategic positions, such as border crossings and airstrips, often wearing official Congolese uniforms.36 Similarly, support is often provided through informal networks rather than formal agreements, making it difficult to establish a clear chain of command.37
The fragmentation between the effective control and overall control tests creates opportunities for sponsoring states to exploit legal uncertainty. By structuring their support to avoid the requirements of both tests, states can ensure that no legal standard is clearly met, thereby evading attribution.38 This fragmentation is particularly pronounced in the DRC context, where the support documented by the Group of Experts could satisfy the overall control test, while no court has yet determined whether it would meet the stricter effective control test as the ICJ applies it to each operation. The uncertainty regarding which test applies, and under what circumstances, provides a legal shield for sponsoring states. States have learned to adapt their methods of intervention to exploit the ambiguity between the two tests.39
Finally, sponsoring states employ strategic denial and counter-narratives to undermine the evidentiary record and maintain plausible deniability. This involves challenging the credibility of international investigations, presenting alternative interpretations of the evidence and mobilising diplomatic support to protect against accountability.40
According to an AFP report of July 2024 on the Group of Experts’ findings, Rwandan authorities publicly denied until the end of 2023 that their troops were operating alongside M23 in North Kivu, and Kigali had since then no longer commented directly on such accusations; the Government of Rwanda had also continued to deny, including in response to the Group’s own requests for information, that it provided support to M23.41 Responses of this kind help to maintain legal ambiguity and to hinder attribution, despite the weight of the evidentiary record.42
The effective control test, while providing doctrinal clarity, is fundamentally inadequate for addressing the realities of contemporary proxy warfare in the DRC. Its requirement of specific operational direction creates an evidentiary burden that is very difficult to discharge.43 The test’s focus on operational control over individual operations is fundamentally mismatched with the way modern proxy conflicts are conducted. Requiring proof that a sponsoring state directed each individual military operation therefore ignores the manner in which military organisations ordinarily function.44
The effective control test also fails to account for the sophisticated strategies states employ to evade attribution. By maintaining plausible deniability, exploiting evidentiary burdens and utilising de facto structures, states can ensure that their proxy relationships fall below the effective control threshold while maintaining sufficient operational influence to achieve their strategic objectives.45
The overall control test’s focus on broader patterns of support and coordination makes it better suited to the realities of modern proxy warfare.46 The ICTY in Tadić developed the overall control test to address organised armed groups of the kind that characterise conflicts such as the one in eastern DRC. The test requires proof that a state coordinates or helps in the general planning of an armed group’s military activity, not merely that it equips and finances the group. This threshold may be met in the Rwanda-M23 relationship, given the extensive evidence of strategic coordination, intelligence support and operational integration documented by the Group of Experts.
However, while more flexible than the effective control test, the overall control test is not without its own limitations.47 The test has been criticised by the ICJ for expanding state responsibility beyond acceptable limits.48 There are genuine concerns that the test could lead to attribution in cases where states exercise only minimal influence over armed groups, creating a risk of over-attribution.49 There is also uncertainty regarding the precise threshold for overall control, as the ICTY did not provide detailed guidance on how the test should be applied.50
The discrepancy between the two tests creates an accountability gap that states supporting proxies can exploit. In the DRC context, the evidence of Rwandan support for M23 may satisfy the overall control test, but that test has not been accepted by the ICJ51 and so cannot ground responsibility before it. On the other side, the standard applied by the ICJ, the principal judicial organ of the United Nations, is effective control. The Group of Experts has itself described the RDF’s role in terms of “de facto direction and effective control” over M23 operations, but that is the characterisation of an investigative body rather than a judicial finding, and no court has determined whether the evidence would satisfy the effective control test as the ICJ applies it to each operation.52 The existence of two competing standards, each with different evidentiary requirements, allows a state in Rwanda’s alleged position to avoid a finding of responsibility. As Kiyirembera observes, “[t]he stringent evidentiary requirements pose a significant challenge in the context of proxy warfare, where State sponsors engage in indirect and covert interventions precisely to circumvent the political, strategic, and legal repercussions of direct engagement.” This observation captures the central dilemma of the current legal framework: it is designed for a world of formal state-to-state relations that no longer reflects the reality of contemporary conflict.53 This makes it difficult for a state such as Rwanda, if the allegations against it are established, to be held accountable for support to M23, perpetuating impunity and the suffering of civilians in the region.
The attribution gap has devastating consequences for civilians in eastern DRC. The inability to hold sponsoring states accountable for the conduct of their proxies perpetuates a cycle of violence that continues to kill and displace civilians on a large scale, and the Group of Experts records violations by both belligerent parties, including Congolese armed forces and allied Wazalendo groups, and finds that armed groups commit such violations most often with complete impunity.54 Attacks on civilian populations, sexual violence and displacement continue with impunity, as armed groups operate with state support while their sponsors remain protected from legal consequences.55
The M23 rebellion has been particularly devastating for civilians. The group has been implicated in numerous violations of international humanitarian law, including attacks on civilian populations, summary executions, sexual violence and forced displacement.56 The Group of Experts has documented that M23 has carried out attacks against civilian populations, including shelling of populated areas, extrajudicial killings and acts of sexual violence.57
The resurgence of M23 in late 2021 intensified an already severe displacement crisis. By late 2022, an estimated 5.6 million Congolese were internally displaced, making it the largest internal displacement crisis in Africa.58
This article concludes that the existing international legal framework is insufficient to address contemporary proxy warfare. Although Article 8 of ARSIWA provides the basis for attributing the conduct of non-state armed groups to states, the continuing uncertainty between the effective control and overall control tests creates an accountability gap. The Rwanda-M23 relationship, as documented by the UN Group of Experts, illustrates how states can exercise substantial influence through indirect and covert forms of support while avoiding the threshold required for attribution.
The following recommendations are designed to address the accountability gap identified in this article, so that sponsoring states can no longer evade responsibility for violations committed through proxy armed groups.
Article 8 should be amended to provide greater clarity on the relationship between effective control and overall control. The assessment should take into account the nature, duration, intensity and purpose of the relationship between the sponsoring state and the proxy, while still requiring a meaningful connection between the state and the wrongful conduct.
Courts should consider the combined effect of military assistance, intelligence sharing, strategic planning, training and operational coordination rather than examining each form of support separately. This would better reflect the way contemporary proxy relationships operate and reduce opportunities for states to avoid attribution by distributing their support across different channels.
International investigations should be able to rely on converging circumstantial evidence where direct orders, formal command structures or other documentary evidence are deliberately concealed. This would help address the evidentiary difficulties created by covert operations and plausible deniability.
Taken together, these reforms would strengthen the law of attribution and reduce the ability of states to use proxy forces while avoiding responsibility. They would also make international law more responsive to contemporary forms of armed conflict and strengthen accountability for violations committed through proxy forces.
*****
1. Amos C. Fox, In Pursuit of a General Theory of Proxy Warfare 1–24 (Ass’n of the U.S. Army, Land Warfare Paper No. 123, 2019), https://www.ausa.org/sites/default/files/publications/LWP-123-In-Pursuit-of-a-General-Theory-of-Proxy-Warfare.pdf.
2. Dylan Craig, Proxy War by African States, 1950–2010, at iii (2012) (Ph.D. dissertation, American University).
3. Roc Thomas Kiyirembera, Blurred Lines: Conflict Classification, Proxy Militias, and the Quest for Accountability in the Democratic Republic of the Congo, 2 Law & Criminology J. 1, 1–21 (2025), https://doi.org/10.21825/lcj.91572.
4. S.C. Res. 2773, ¶¶ 1, 4 (Feb. 21, 2025).
5. U.N. Group of Experts on the Democratic Republic of the Congo, Final Report of the Group of Experts on the DRC Submitted in Accordance with Paragraph 4 of Security Council Resolution 2021 (2011), U.N. Doc. S/2012/843 (Nov. 15, 2012).
6. Int’l L. Comm’n, Draft Articles on Responsibility of States for Internationally Wrongful Acts, art. 8, U.N. Doc. A/56/10 (2001) [hereinafter ARSIWA].
7. James Crawford, State Responsibility: The General Part 141–61 (2013).
8. Bethel Uzoma Ihugba, Introduction to Legal Research Method and Legal Writing 28 (2020).
9. Terry Hutchinson & Nigel Duncan, Defining and Describing What We Do: Doctrinal Legal Research, 17 Deakin L. Rev. 83, 101–08 (2012), https://www.austlii.edu.au/cgi-bin/viewdoc/au/journals/DeakinLawRw/2012/4.html.
10. Id.
11. Id.
12. Research Methods for Law ch. 1 (Michael McConville & Wing Hong Chui eds., 2d ed. 2017).
13. Id.
14. ARSIWA, supra note 6, art. 8.
15. Military and Paramilitary Activities in and Against Nicaragua (Nicar. v. U.S.), Judgment, 1986 I.C.J. 14 (June 27) [hereinafter Nicaragua].
16. Application of the Convention on the Prevention and Punishment of the Crime of Genocide (Bosn. & Herz. v. Serb. & Montenegro), Judgment, 2007 I.C.J. 43 (Feb. 26) [hereinafter Bosnian Genocide].
17. Prosecutor v. Tadić, Case No. IT-94-1-A, Judgement (Int’l Crim. Trib. for the Former Yugoslavia July 15, 1999).
18. Nicaragua, 1986 I.C.J. ¶ 115.
19. Id. ¶¶ 109–10, 115; Bosnian Genocide, 2007 I.C.J. ¶ 400.
20. Nicaragua, 1986 I.C.J. ¶¶ 115–16.
21. Elena Laura Álvarez Ortega, The Attribution of International Responsibility to a State for Conduct of Private Individuals Within the Territory of Another State, InDret, no. 1, 2015, at 1–20.
22. Bosnian Genocide, 2007 I.C.J. ¶¶ 400, 406.
23. Tadić, Case No. IT-94-1-A, ¶¶ 115–20, 124–37.
24. Id. ¶ 131.
25. Antonio Cassese, The Nicaragua and Tadić Tests Revisited in Light of the ICJ Judgment on Genocide in Bosnia, 18 Eur. J. Int’l L. 649, 657 (2007), https://doi.org/10.1093/ejil/chm034.
26. Crawford, supra note 7, at 141.
27. Jennifer Maddocks, State Responsibility for International Law Violations Involving Non-State Actors in Armed Conflict (2021) (Ph.D. thesis, University of Reading), https://centaur.reading.ac.uk/107450/.
28. Bosnian Genocide, 2007 I.C.J. ¶ 406.
29. Laurie R. Blank, The Law and Ethics of Arming Rebel Proxies 2–18 (May 17, 2022) (unpublished manuscript), https://ssrn.com/abstract=4112524, published as Laurie R. Blank, Arming Proxies: Ethical and Legal Dilemmas, in Routledge Handbook of Proxy Wars 246 (Assaf Moghadam, Vladimir Rauta & Michel Wyss eds., 2023), https://doi.org/10.4324/9781003174066-23.
30. Mariia Rudko, “Little Green Men” in a Legal Gray Area: International Responsibility for Proxy Wars 42–44 (2020) (LL.M. thesis, Lund University), https://lup.lub.lu.se/student-papers/search/publication/9019695.
31. Blank, supra note 29, at 1.
32. Blank, supra note 29.
33. U.N. Group of Experts on the Democratic Republic of the Congo, Final Report of the Group of Experts on the Democratic Republic of the Congo, ¶¶ 36–37, 40–41 & annex 21, U.N. Doc. S/2025/446 (July 3, 2025) [hereinafter 2025 Group of Experts Report].
34. Jennifer Maddocks, The Conflict in Eastern DRC and the State Responsibility of Rwanda and Uganda, Articles of War (Feb. 6, 2025), https://lieber.westpoint.edu/conflict-eastern-drc-state-responsibility-rwanda-uganda/.
35. Will Rowe, Note, In Search of Lost Crime: The International Court of Justice, Non-State Armed Groups, and State Responsibility, 56 Geo. J. Int’l L. 529 (2025).
36. 2025 Group of Experts Report, supra note 33, ¶¶ 36–37.
37. Erica Dreyfus Borghard, Friends with Benefits? Power and Influence in Proxy Warfare 9 (2014) (Ph.D. dissertation, Columbia University), https://doi.org/10.7916/D8Q81B7Z.
38. Rudko, supra note 30.
39. Id. at 37, 42–44.
40. Scott Williamson, Do Proxies Provide Plausible Deniability? Evidence from Experiments on Three Surveys, 68 J. Conflict Resol. 322 (2024), https://doi.org/10.1177/00220027231170562.
41. Rwandan Troops Fought Alongside M23 Rebels in DR Congo: UN Experts, Arab News (July 8, 2024), https://www.arabnews.com/world/rwandan-troops-fought-alongside-m23-rebels-in-dr-congo-un-experts-2545601; U.N. Group of Experts on the Democratic Republic of the Congo, Final Report of the Group of Experts on the Democratic Republic of the Congo, ¶ 54, U.N. Doc. S/2023/431 (June 13, 2023) [hereinafter 2023 Group of Experts Report].
42. Kiyirembera, supra note 3, at 1–21.
43. Cassese, supra note 25, at 649–68.
44. Id. at 654, 657.
45. Rudko, supra note 30, at 7.
46. Cassese, supra note 25, at 657–58.
47. See Prosecutor v. Blaškić, Case No. IT-95-14-T, Judgement, Declaration of Judge Shahabuddeen (Int’l Crim. Trib. for the Former Yugoslavia Mar. 3, 2000); but see Cassese, supra note 25, at 662.
48. Bosnian Genocide, 2007 I.C.J. ¶ 406.
49. Id.
50. Blaškić, Declaration of Judge Shahabuddeen, supra note 47, ¶¶ 2–3.
51. Bosnian Genocide, 2007 I.C.J. ¶ 406.
52. U.N. Charter art. 92; 2025 Group of Experts Report, supra note 33, ¶¶ 18, 44; see also Bosnian Genocide, 2007 I.C.J. ¶ 400.
53. Kiyirembera, supra note 3, at 18, 20.
54. U.N. Group of Experts on the Democratic Republic of the Congo, Final Report of the Group of Experts on the Democratic Republic of the Congo, ¶¶ 90–93, U.N. Doc. S/2024/432 (June 4, 2024) [hereinafter 2024 Group of Experts Report].
55. Human Rights Watch, Democratic Republic of Congo: Ending Impunity for Sexual Violence (June 10, 2014), https://www.hrw.org/news/2014/06/10/democratic-republic-congo-ending-impunity-sexual-violence.
56. 2024 Group of Experts Report, supra note 54, ¶¶ 90, 94; 2023 Group of Experts Report, supra note 41, ¶¶ 65–69.
57. 2023 Group of Experts Report, supra note 41, ¶¶ 63–69; 2024 Group of Experts Report, supra note 54, ¶¶ 91, 93–94.
58. U.N. High Comm’r for Refugees, UNHCR Position on Returns to North Kivu, South Kivu, Ituri and Adjacent Areas in the Democratic Republic of Congo Affected by Ongoing Conflict and Violence – Update III ¶ 23 (Nov. 2022), https://www.refworld.org/policy/countrypos/unhcr/2022/124181.