Book Symposium ‘Proxy Warfare’: How Does Proxy Warfare Challenge International Humanitarian Law? Concluding Remarks, Part I

About the author(s):

Eugénie Duss is a Research Fellow at the Academy of International Humanitarian Law and Human Rights.

See French version below.

Acknowledgement: The author would like to express her sincere gratitude to Dr. Pauline Lesaffre for reviewing this piece. Her insightful feedback and expertise were instrumental in improving the accuracy of this blog post.

Disclaimer: The author has used AI-driven tools, Grammarly, for language editing and grammar correction. The author, however, takes full responsibility for the accuracy and integrity of the content.

Proxy warfare challenges international humanitarian law (IHL) in several ways, raising both general and specific difficulties (for a brief overview, see here). Some of these challenges have been explored throughout this symposium by various IHL experts. Often, they relate, directly or indirectly, to the status of the proxy group and the concept of ‘belonging’ as found in Article 4(A) of the Geneva Convention III (GC III) on prisoners of war (POW). Others pertain to occupation by proxy. These points relating respectively to the status of combatant and POW and to military occupation illustrate that applying IHL norms specific to international armed conflicts (IAC) – i.e., absent from the law of non-international armed conflict  (NIAC) – in the context of proxy warfare is quite complex because, on the ground, an armed group is fighting against a State. Drawing on the insights provided by the contributors, I will now propose certain solutions derived from my book (Part I). Building on this overview, I will then briefly explore alternative avenues (Part II).

Part I

The Concept of ‘Belonging’ and Its Implications

As mentioned in the introductory post, situations where an organised armed group is under the overall control of a first State (sponsor) and fights against another State (target) are to be classified as IACs. If we consider that both the target and sponsor States are parties to such IACs, and that the proxy belongs to its sponsor, it follows that the proxy is bound by the law of IAC and may be equated with the (irregular) armed forces of its sponsor. In particular, this reasoning, namely the idea that the proxy belongs to its sponsor, carries several implications for detention and the status of the proxy under IHL, as described below.  

Regarding the deprivation of liberty, as highlighted by Marco Sassòli, applying the rules governing the status and treatment of POW in IACs by proxy presents both conceptual and practical challenges. Resorting to the concept of ‘belonging’  in such situations offers various advantages. Firstly, it allows for the identification of the sponsor as the de jure detaining power, while the proxy acts as the sponsor’s irregular forces. Therefore, members of the target State’s armed forces who are detained by the proxy have actually fallen into the power of the sponsor State. However, since the proxy is often the sole actor engaged in hostilities against the target State, relevant IAC rules must be adapted to the non-State nature of the proxy. 

Secondly, the concept of ‘belonging’ also permits that proxy members potentially benefit from POW status when detained by the target State. In my view, this extends to cases when proxy members are nationals of that State. In traditional IHL, while the text of GC III does not explicitly specify a nationality criterion, nationals of the detaining power are excluded from POW status (majority view) or may at least face prosecution for treason (minority view). This means that proxy members would not enjoy POW status in the hands of the target State when they are nationals of this State. To overcome this difficulty, one might replace the nationality criterion with the criteria of allegiance, as developed by the International Criminal Tribunal for the Former Yugoslavia (ICTY) (here, para 163-169). Another solution would be to resort to adversity and determine it through the concept of ‘belonging’, i.e., establish that the proxy is fighting for the sponsor. 

Concerning the status of the proxy, the requirement of ‘belonging’ initially serves as a collective conditionthat the proxy must fulfil for its members to benefit from the POW status under Article 4(A)(2) GC III. However, proxy members would only enjoy POW status if the proxy meets the four additional conditions outlined in Article 4(A)(2), and if they fulfil the individual obligation to distinguish themselves from the civilian population. These additional conditions include conducting ‘operations in accordance with the laws and customs of war,’ which in turn encourages adherence to IHL. The concept of belonging also helps to more broadly identify which entities are operating within an IAC. In particular, as discussed by Jelena Aparac, the concepts of ‘belonging’ and ‘overall control’ might assist in assessing the status of ambiguous actors or entities, such as private military and security companies (PMSC) or the opaque Wagner Group, under IHL. 

However, these concepts only apply in the presence of organised armed groups (see here, para 120 and 137, and the four conditions of Article 4(A)(2) GC III). Therefore, it is crucial to determine whether such actors as PMSCs, or certain components of them, qualify as organised armed groups. Yet Jelena Aparac aptly underlines that this task can be challenging in practice due to the diversity, multifaceted profile and structural complexities of PMSCs and entities such as the Wagner Group. A potential solution could involve determining whether the entity or one of its components conducts hostilities on behalf of a party to the conflict – i.e. belongs to it – and whether its members, taken individually, perform a continuous function corresponding to that collectively exercised by the entity as a whole. The International Committee of the Red Cross (ICRC) (pp 32-5) has proposed such an approach, mainly for targeting purposes in NIACs, and defined organised armed groups as composed of individuals performing a continuous combat function. In my book, I also applied this approach to identify individuals who belong to the proxy and may claim POW status. Adopting a single criterion to determine both individual and collective belonging would also help ensure the overall consistency of the legal framework.

Additionally, it is important to note that ‘overall control’ and ‘belonging’ are two distinct concepts. The former requires that the sponsor ‘has a role in organising, coordinating or planning the military actions of the military armed group, in addition to financing, training and equipping or providing operational support to that group’ (here, para 137). By contrast, the latter involves a subjective element, as the State ‘must accept both the fighting role of the group and the fact that the fighting is done on its behalf’ (here, para 1005). Thus, the concept of ‘belonging’ requires that a party to an IAC, i.e., a State, agrees, even tacitly – in other words, does not reject – that the group fights on its behalf (here, para 1007).

However, as previously explained in the introductory post, neither the sponsor nor the proxy usually acknowledges their relationship publicly. Accordingly, their denial would prevent the proxy from being considered as belonging to its sponsor. Nevertheless, IHL is a pragmatic body of law whose application does not depend on the declared will of States, but on objective and factual criteria (here para 194, 209 and 211-7). A formal refusal that the proxy is fighting on its behalf should not allow the sponsor State to deny individuals the protections they are entitled to under the law of IACs (including POW status, provided they meet the necessary conditions), in light of the actual situation prevailing (the proxy operates within an IAC and fights on behalf of the State). Therefore, as further discussed in my book, even if the sponsor denies that the proxy is fighting on its behalf, the proxy still belongs to its sponsor as long as the sponsor maintains overall control. 

Lastly, the question of whether a proxy belongs to its sponsor has significant implications for attribution purposes, in particular in determining whether it acts as an agent of its sponsor. This issue is not purely academic, but is of direct practical relevance, for example with regards to the application of the Arms Trade Treaty (ATT). Article 2(3) ATT stipulates that this treaty does not apply to ‘the international movement of conventional arms by, or on behalf of, a State Party for its use provided that the conventional arms remain under that State Party’s ownership’. This means that situations where States transfer arms to their own agents, possibly the sponsor States to their proxies, do not fall within the ATT scope of application. 

This said, as explained by Stuart Casey-Maslen, the prevailing view under the general rules of State responsibility is that groups under the overall control of a State cannot be considered agents of that State. This aligns with the interpretation of the International Court of Justice (ICJ). Accordingly, a higher threshold is required for a group to be equated with a de facto organ or agent of the State (total dependence), or even to hold the State accountable for its conduct (effective control) (here, para 392-3 and 396-400).

This prevailing view seems to be in contradiction with the idea that, under IHL, States are responsible for the conduct of both their regular and irregular forces (Articles 43(1) and 91 of Additional Protocol I (1977)). Further, the ICRC (para 306) and the ICTY (para 91-2, 104 and 120) clearly assert that the sponsor is responsible for the conduct of its proxy, treating it as the sponsor’s agent. 

In my book, I discuss a way to reconcile both approaches: while the sponsor cannot generally be held accountable based on the criterion of ‘overall control’, there is a special responsibility rule under IHL (i.e. a lex specialis) that applies exclusively in armed conflict situations. This rule allows for the actions of an organised armed group under a State’s overall control to be attributed to that State. In this context, the proxy must indeed be treated as a de facto agent of the State.

Occupation by Proxy: Conceptual and Practical Considerations

From a conceptual standpoint, occupation by proxy raises questions about the degree of control, first, that the proxy must have over the concerned territory and, second,  that the sponsor must exercise over the proxy. Regarding the second,  I agree with Martha M. Bradley, Alina-Camille Berdefy, and Joshua Joseph Niyo that occupation by proxy does not require a higher degree of control from the sponsor over its proxy. 

However, since territorial control is a prerequisite for both military occupation (see e.g. here, para 301-4) and the application of Additional Protocol II (1977) (AP II), it is worth examining whether these two forms of territorial control are identical and what that might imply. According to Martha Bradley and Camille Berdefy, the requirement of ‘functional control’ under AP II differs from the one of effective control in cases of occupation by proxy. Consequently, a pre-existing AP II-type NIAC may continue to exist between the same belligerents alongside the occupation by proxy when the group controlling the territory is under the overall control of a third State. By contrast, I believe that the two forms of control are similar. In both scenarios, the territory must be controlled to the exclusion of government forces (see here, para 373-5, and here, para 4467). Further, if we consider that a parallel NIAC persists and is not absorbed by the IAC by proxy, this would imply, first, reverting to double classification instead of internationalising the conflict as a whole. Second, to identify the applicable legal framework, it would be essential to determine whether a specific conduct occurred in relation to the NIAC or the IAC. This can be quite challenging when, in both cases – military occupation and an AP II-type NIAC – the proxy effectively controls the territory while being under the sponsor’s overall control.

In practice, when a situation of occupation by proxy is recognised, difficulties nevertheless persist, as Joshua Niyo highlights. Once we establish that the sponsor is acting as the occupying power through its proxy, which in turn functions as the occupying forces and exercises authority over the territory, a challenge arises when the proxy operates with a ‘double hat’ – meaning the proxy was already administering the territory before being controlled by its sponsor. This ‘layered authority’ necessitates clearly defining the roles of the proxy and distinguishing when it acts as an occupying force and when it operates as a local authority, thereby enjoying a broader margin of action. 

In such situations, I argue that only the military wing of the proxy – consisting of individuals performing a continuous combat function – may serve as the occupying forces. Meanwhile, the other branches of the proxy, e.g. political, may only act as local authorities. Additionally, regarding the military wing of the proxy, I believe it is important to assess whether the proxy had been fulfilling the role of occupying forces prior to being placed under the sponsor’s overall control. If it had, the proxy may continue to function as a local authority; if it had not, it must be considered as acting only as the occupying forces and must adhere to the law of occupation. This solution preserves the status quo ante and complies with the principle that local authorities operate under the authority of the occupying forces (see here and Dinstein, para 154 and 280). 

Another issue pertains precisely to the preservation of the status quo ante. As developed in my book, I am of the opinion that this principle must be applied strictly in situations of occupation by proxy. However, obviously, this does not mean that the proxy is fully exempt from its positive obligations under the law of occupation. Rather, in line with the context-dependent approach developed for the application of international human rights law to non-state actors (Murray, pp 179-83), this means that the proxy, in fulfilling its obligations as the occupying forces, must primarily allow local institutions, authorities, and establishments to continue functioning (respect). It is only when these entities fail to operate in accordance with the law of occupation that the proxy may step in to ‘facilitate’ their operations (facilitation). If facilitation proves insufficient, the proxy must either perform these functions directly, potentially using local human and material resources, or seek assistance from third parties (substitution).

A third and last challenge concerns clarifying the legal obligations, practical capabilities, and responsibilities between the sponsor and its proxy. In general, while the sponsor assumes formal obligations under the law of occupation, the proxy is actually the one who has to meet these obligations on the ground. I believe that, even though the proxy fulfils the duties of the occupying power in practice, the official occupying power, i.e. the sponsor, retains responsibility for violations of the law of occupation. This situation actually mirrors that of a traditional military occupation: a State is an abstract concept and can only act through its organs and agents, here, its occupying forces. Moreover, compared to traditional occupations, occupation by proxy can occur with the sponsor’s regular armed forces being present (as seen with M23 and Rwanda in the Eastern Democratic Republic of Congo) or absent (as in the case of Nagorno-Karabakh until September 2024). As Joshua Niyo points out, this diversity must be taken into account when delimiting specific obligations.

Remerciements : L’auteure tient à exprimer sa sincère gratitude à Dr Pauline Lesaffre et à M. Guillaume Guez Maillard pour leur relecture de cette contribution. Leurs remarques pertinentes et leur expertise ont été déterminantes pour améliorer la qualité de ce billet, tant sur le fond que sur la forme.

Les conflits armés internationaux (CAI) par proxy mettent le droit international humanitaire (DIH) à l’épreuve à plusieurs égards, soulevant des difficultés à la fois générales et spécifiques (pour un bref aperçu, voir ici). Certaines de ces difficultés ont été examinées dans ce symposium par plusieurs experts en DIH. Celles-ci concernent souvent, directement ou indirectement, le statut du groupe proxy et au concept d’« appartenance » tel qu’il figure à l’art. 4 (A) (2) de la Troisième Convention de Genève (CG III) relative aux prisonniers de guerre (PG). D’autres concernent l’occupation par proxy. Qu’il s’agisse des règles relatives au statut de combattant et de PG ou de celles ayant trait à l’occupation militaire, l’application de ces deux ensembles de normes du DIH propres aux CAI et sans équivalent en droit des conflits armés non internationaux (CANI), se révèle particulièrement délicate lorsque, sur le terrain, un groupe armé combat un État. En s’appuyant sur les analyses des contributeurs, ce billet propose plusieurs solutions issues de l’ouvrage de l’auteure (partie I). Fort de ce tour d’horizon, ce billet explore ensuite brièvement des pistes alternatives (partie II).

Partie I

Le concept d’« appartenance » et ses implications

Comme l’a souligné le billet introductif, les situations où un groupe armé organisé (proxy) qui combat un État (cible) est contrôlé globalement par un autre État (sponsor) doivent être qualifiées de CAI. Si l’on considère que l’État cible et l’État sponsor sont tous deux parties à ce CAI, et que le proxy appartient à son sponsor, il en découle que le proxy est lié par le droit des CAI et peut être assimilé aux forces armées irrégulières de son sponsor. Considérer que le proxy appartient à son sponsor n’est pas sans incidence sur la détention et le statut du proxy en DIH.

En matière de privation de liberté, comme l’a relevé Marco Sassòli, l’application des règles régissant le statut et le traitement des PG dans les CAI par proxy soulève des difficultés tant conceptuelles que pratiques. Le recours au concept d’« appartenance » dans de telles situations présente plusieurs avantages. Premièrement, il permet d’identifier le sponsor comme puissance détentrice de jure, tandis que le proxy agit en tant que forces irrégulières de ce dernier. Par conséquent, les membres des forces armées de l’État cible détenus par le proxy sont, en réalité, tombés au pouvoir de l’État sponsor. Toutefois, dans la mesure où le proxy est souvent le seul acteur engagé dans les hostilités contre l’État cible, les règles pertinentes du droit des CAI doivent être adaptées à sa nature non étatique.

Deuxièmement, le concept d’« appartenance » permet également aux membres du proxy de bénéficier potentiellement du statut de PG lorsqu’ils sont détenus par l’État cible. À mon sens, cela s’étend aux cas où les membres du proxy sont des ressortissants de cet État. En DIH, bien que le texte de la CG III ne pose pas expressément d’exigence de nationalité, la doctrine majoritaire exclut traditionnellement les ressortissants de la puissance détentrice du statut de PG, ou les considère, selon l’opinion minoritaire, au moins passibles de poursuites pour trahison. Cela signifie que les membres du proxy au pouvoir de l’État cible ne bénéficieraient pas du statut de PG lorsqu’ils sont ressortissants de cet État (ce qui arrive souvent en pratique). Pour surmonter cette difficulté, il serait possible de substituer au critère de nationalité celui de l’allégeance, tel que développé par le Tribunal pénal international pour l’ex-Yougoslavie (TPIY) (ici, par. 163-169). Une autre solution consisterait à recourir à la notion d’adversité, en la déterminant au moyen du concept d’« appartenance », c’est-à-dire en établissant que le proxy combat pour le sponsor.

S’agissant du statut du proxy, l’exigence d’« appartenance » sert initialement de condition collective que le proxy doit remplir pour que ses membres puissent bénéficier du statut de PG en vertu de l’article 4(A)(2) CG III. Toutefois, les membres du proxy ne bénéficieraient du statut de PG que si le proxy remplit les quatre conditions supplémentaires énoncées à l’article 4(A)(2), et s’ils respectent eux-mêmes l’obligation individuelle de se distinguer de la population civile. Ces conditions incluent le fait de mener des « opérations conformément aux lois et coutumes de la guerre », ce qui encourage à son tour le respect du DIH. Le concept d’appartenance contribue également, de manière plus générale, à identifier les entités qui opèrent dans un CAI. En particulier, comme l’examine Jelena Aparac, les concepts d’« appartenance » et de « contrôle global » peuvent aider à apprécier le statut, au regard du DIH, d’acteurs ou d’entités ambigus, tels que les sociétés militaires et de sécurité privées (SMSP) ou l’opaque groupe Wagner.

Toutefois, ces concepts ne s’appliquent qu’en présence de groupes armés organisés (voir ici, par. 120 et 137, et les quatre conditions de l’article 4(A)(2) CG III). Il est donc essentiel de déterminer si des acteurs tels que les SMSP, ou certaines de leurs composantes, peuvent être qualifiés de groupes armés organisés. Or, Jelena Aparac souligne à juste titre que cette tâche peut s’avérer difficile en pratique en raison de la diversité, du profil multiforme et des complexités structurelles des SMSP et d’entités telles que le groupe Wagner. Une solution consisterait à établir si l’entité, ou l’une de ses composantes, mène des hostilités pour le compte d’une partie au conflit, c’est-à-dire si elle lui appartient, et si ses membres, pris individuellement, exercent une fonction continue correspondant à celle exercée collectivement par l’entité dans son ensemble. Le Comité international de la Croix-Rouge (CICR) (p. 34-7) a proposé une telle approche, principalement à des fins de ciblage dans les CANI, et a défini les groupes armés organisés comme étant composés d’individus exerçant une fonction continue de combat. Dans mon ouvrage, j’ai aussi appliqué cette approche pour identifier les individus qui appartiennent au proxy et peuvent prétendre au statut de PG. Adopter un critère unique afin de déterminer l’appartenance, tant individuelle que collective, permettrait aussi d’assurer la cohérence générale de l’édifice juridique.

En outre, il importe de souligner que le « contrôle global » et l’« appartenance » sont deux concepts distincts. Le premier exige que le sponsor « joue un rôle dans l’organisation, la coordination ou la planification des actions militaires du groupe militaire, en plus de le financer, l’entraîner, l’équiper ou lui apporter son soutien opérationnel » (ici, par. 137). En revanche, le second comporte un élément subjectif, dans la mesure où l’État « doit accepter à la fois la fonction combattante du groupe et le fait que le combat soit mené en son nom » (ici, par. 1005). Ainsi, le concept d’« appartenance » exige qu’une partie à un CAI, c’est-à-dire un État, accepte, même tacitement – autrement dit, ne rejette pas – que le groupe combatte pour son compte (ici, par. 1007).

Toutefois, comme expliqué précédemment, aucun sponsor ou proxy n’a jamais reconnu publiquement leur relation. En conséquence, leur dénégation empêcherait que le proxy soit considéré comme appartenant à son sponsor. Néanmoins, le DIH est un droit pragmatique dont l’application ne dépend pas de la volonté déclarée des États, mais de critères objectifs et factuels (ici, para. 194, 209 et 211-7). Un refus de façade ne devrait pas permettre à l’État sponsor de priver des individus des protections auxquelles ils ont droit en vertu du droit des CAI (y compris le statut de PG pour autant qu’ils remplissent les conditions nécessaires), à la lumière de la situation réelle qui prévaut (le proxy opère dans un CAI et combat pour le compte de l’État). Dès lors, comme cela est discuté plus avant dans mon ouvrage, même si le sponsor nie que le proxy combatte pour son compte, le proxy appartient néanmoins à son sponsor aussi longtemps que celui-ci le contrôle globalement.

Enfin, la question de savoir si un proxy appartient à son sponsor emporte des conséquences importantes aux fins de l’attribution, en particulier lorsqu’il s’agit de déterminer s’il agit comme agent de son sponsor. Cette question n’est pas purement académique, mais revêt une pertinence pratique directe, par exemple s’agissant de l’application du Traité sur le commerce des armes (TCA). L’art.  2(3) TCA prévoit que ce traité ne s’applique pas au « au transport international par tout État Partie ou pour son compte d’armes classiques destinées à son usage, pour autant que ces armes restent sa propriété ». Autrement dit, les transferts d’armes d’un État à ses propres agents, potentiellement d’un sponsor à son proxy, échappent au champ d’application du TCA.

Cela étant, comme l’explique Stuart Casey-Maslen, l’opinion majoritaire rejette l’idée que les groupes placés sous le contrôle global d’un État puissent être considérés comme des agents de cet État en vertu des règles générales de la responsabilité de l’État. Cette position correspond à celle de la Cour internationale de Justice (CIJ), qui requiert un seuil plus élevé pour qu’un groupe soit assimilé à un organe ou à un agent de facto de l’État (totale dépendance) ou qu’il puisse simplement engager la responsabilité de cet État pour ses agissements (contrôle effectif) (ici, par. 392-3 et 396-400).

Cette position se heurte à l’idée selon laquelle, en DIH, les États sont responsables du comportement de leurs forces armées régulières et irrégulières (articles 43(1) et 91 du Protocole additionnel I (1977)). En outre, le CICR (par. 306) et le TPIY (par. 91-2, 104 et 120) affirment clairement que le sponsor est responsable du comportement de son proxy, en l’assimilant à un agent de son sponsor.

Dans mon ouvrage, je propose une réconciliation de ces deux approches : si le sponsor ne peut généralement pas voir sa responsabilité engagée sur le fondement du critère du « contrôle global », il existe une règle spéciale de responsabilité en DIH, c’est-à-dire une lex specialis, qui s’applique exclusivement dans les situations de conflit armé. Cette règle permet d’attribuer à l’État les actes d’un groupe armé organisé placé sous son contrôle global. Dans ce contexte, le proxy doit effectivement être traité comme un agent de facto de l’État.

Occupation par proxy : considérations conceptuelles et pratiques

D’un point de vue conceptuel, l’occupation par proxy soulève deux questions fondamentales : quel degré de contrôle le proxy doit-il exercer sur le territoire concerné ? Quel  degré de contrôle le sponsor doit-il exercer sur le proxy ? S’agissant du second aspect, je rejoins Martha M. Bradley, Alina-Camille Berdefy et Joshua Joseph Niyo lorsqu’ils considèrent que l’occupation par proxy ne requiert pas un degré de contrôle plus élevé du sponsor sur son proxy.

Toutefois, dans la mesure où le contrôle territorial constitue une condition préalable tant de l’occupation militaire (voir, par exemple, ici, par. 301-4) que de l’application du Protocole additionnel II (1977) (PA II), il convient d’examiner si ces deux formes de contrôle territorial sont identiques et ce que cela pourrait impliquer. Selon Martha Bradley et Camille Berdefy, l’exigence de « contrôle fonctionnel » au titre du PA II diffère de celle du contrôle effectif dans les situations d’occupation par proxy. Par conséquent, un CANI préexistant relevant du PA II pourrait continuer d’exister entre les mêmes belligérants parallèlement à l’occupation par proxy lorsque le groupe contrôlant le territoire est placé sous le contrôle global d’un État tiers. À mon sens, les deux formes de contrôle sont toutefois similaires. Dans les deux cas, le territoire doit être contrôlé à l’exclusion des forces gouvernementales (voir ici, par. 373-5 et ici, par. 4467). En outre, si l’on considère qu’un CANI persiste en parallèle et n’est pas absorbé par le CAI par proxy, cela impliquerait, premièrement, de revenir à une double classification plutôt que d’internationaliser le conflit dans son ensemble. Deuxièmement, afin d’identifier le cadre juridique applicable, il serait essentiel de déterminer si un comportement spécifique est intervenu en lien avec le CANI ou le CAI. Cela peut s’avérer particulièrement difficile lorsque, dans les deux cas, à savoir l’occupation militaire et un CANI relevant du PA II, le proxy exerce un contrôle effectif sur le territoire tout en étant placé sous le contrôle global du sponsor.

Comme le souligne Joshua Niyo, même lorsqu’une situation d’occupation par proxy est reconnue, des difficultés pratiques subsistent. Un premier obstacle survient lorsque le proxy opère avec une « double casquette », c’est-à-dire lorsqu’il administrait déjà le territoire avant d’être contrôlé par son sponsor. Cette « autorité stratifiée » impose de définir clairement les rôles du proxy et de distinguer les situations dans lesquelles il agit comme forces occupantes de celles dans lesquelles il opère comme autorité locale, jouissant ainsi d’une marge d’action plus large.

Dans de telles situations, je propose la solution suivante : seule l’aile militaire du proxy, soit celle composée d’individus exerçant une fonction continue de combat, peut agir comme forces occupantes. Les autres branches du proxy, par exemple politiques, ne doivent agir qu’en qualité d’autorités locales. En outre, lorsque les agissements du proxy sont régis par le droit de l’occupation, il convient d’établir s’il exerçait déjà ces fonctions avant d’être placé sous le contrôle global du sponsor. Si tel était le cas, le proxy peut continuer à fonctionner comme autorité locale. Dans le cas contraire, il doit être considéré comme agissant en tant que forces occupantes et se conformer au droit de l’occupation. Cette solution préserve le statu quo ante et respecte le principe selon lequel les autorités locales opèrent sous l’autorité des forces occupantes (voir ici et Dinstein, para. 154 et 280).

Une autre question concerne la préservation du statu quo ante. Comme je l’ai développé dans mon ouvrage, le principe du statu quo ante doit être strictement appliqué dans les situations d’occupation par proxy. Cela ne signifie pas, cependant, que le proxy est entièrement dispensé de ses obligations positives en vertu du droit de l’occupation. Au contraire, conformément à l’approche contextuelle développée dans le cadre de l’application du droit international des droits humains aux acteurs non étatiques (Murray, p. 179-83), cela signifie que le proxy, dans l’exécution de ses obligations en tant que forces occupantes, doit avant tout laisser les institutions, autorités et établissements locaux continuer de fonctionner (respect). Ce n’est que lorsque ces entités ne fonctionnent pas conformément au droit de l’occupation que le proxy peut intervenir pour « faciliter » leur fonctionnement (facilitation). Si la facilitation s’avère insuffisante, le proxy doit soit exercer directement ces fonctions, éventuellement en recourant aux ressources humaines et matérielles locales, soit solliciter l’assistance de tiers (substitution).

Une troisième difficulté concerne la répartition des obligations juridiques, des capacités pratiques et des responsabilités entre le sponsor et son proxy. En général, tandis que le sponsor assume les obligations formelles découlant du droit de l’occupation, c’est en réalité le proxy qui doit les satisfaire sur le terrain. De mon point de vue, même si le proxy remplit en pratique les devoirs de la puissance occupante, la puissance occupante officielle, c’est-à-dire le sponsor, demeure responsable des violations du droit de l’occupation. Cette situation reflète en réalité celle d’une occupation militaire traditionnelle : un État est un concept abstrait et ne peut agir que par l’intermédiaire de ses organes et agents, ici, ses forces d’occupation. En outre, par comparaison avec les occupations traditionnelles, l’occupation par proxy peut se produire lorsque les forces armées régulières du sponsor sont présentes, comme dans le cas du M23 et du Rwanda dans l’est de la République démocratique du Congo, ou absentes, comme dans le cas du Haut-Karabakh jusqu’à septembre 2024. Comme le souligne Joshua Niyo, cette diversité doit être prise en compte dans la délimitation des obligations spécifiques.

(Visited 10 times, 10 visits today)

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: