Shifting Responsibility on a Spectrum: The UK’s Responsibility for Externalised Border Control Operations

Abstract

The United Kingdom (UK) government continues exploring ways to externalise its border controls to deter people from travelling to its shores. States, including the UK, use externalised border controls to reduce responsibility and avoid legal obligations through distance-creation in a manifestation of ‘irresponsibilisation’. They argue that extraterritorial border controls do not trigger their obligations under international refugee and human rights law, which are primarily territorial in scope. Were such claims true, they would create accountability gaps, allowing states to evade responsibility through cooperation and offshoring their legal duties. This paper challenges this view. It introduces a ‘responsibility spectrum’ applicable to the UK and other states’ actions involving externalised border controls, especially offshore processing of asylum claims or returns. The argument demonstrates that responsibility can arise for breaches of negative obligations, aiding and assisting a state, or violations of positive obligations. It emphasises that, despite the difficulties posed by ‘irresponsibilisation’, international law will ensure the UK is held accountable for any breaches it facilitates through any future externalisation policies.

1. Introduction

Following legal challenges to the United Kingdom (UK)–Rwanda Migration Partnership, which sought to relocate ‘illegal’ migrants from the UK to Rwanda for processing, the incoming Labour government scrapped the Rwanda plan (BBC News 2024a). The Border, Security, Asylum and Immigration Bill repeals the Safety of Rwanda Act 2024 and the duty on the Secretary of State to remove asylum seekers who entered the UK ‘illegally’ (Border, Security, Asylum and immigration Bill 2025). Much of the Illegal Migration Act 2023 and the Nationality and Borders Act 2022 remain in place, including expanded detention powers. The new Bill continues to emphasise border securitisation, deterrence, and criminalising irregular entry.

The Labour government is now exploring alternatives to the Rwanda plan. Starmer has said that it aims to ‘tackle problems upstream, before they reach our shores’ (BBC News 2024b). One proposed policy involves a security agreement between the UK and Iraq, with Iraq receiving nearly £1 million to secure its borders and combat people smugglers (Gov.uk 2024). Other options have included outsourcing asylum processing to alternative ‘safe third countries’, in line with the Italy–Albania Protocol, though Albania made clear they were unwilling to partner with the UK (ECRE 2024).

Indeed, ‘return hubs’ are the latest manifestation of the UK’s efforts to externalise its border control operations to deter migrants from reaching its shores. This is consistent with the EU’s developments on the EU Return Regulation, which would allow an EU member to establish an agreement or arrangement with a third country for return purposes (Migration and Home Affairs 2025). The UK government explains that return hubs would serve as detention centres in third countries for failed asylum seekers who have exhausted all appeals but seek to delay deportation (The Times 2024). In practice, these centres would be used to forcibly transfer people targeted by expulsion measures. This could leave individuals without proper legal protections, with unclear accountability for rights violations, and subject to indefinite detention while awaiting deportation (Action Aid 2025). As legal challenges to the Rwanda plan appear have made full-scale third-country processing unfeasible, externalising the returns process appears to be an attempt to create distance between the UK and the return of failed asylum-seekers, often the most problematic part of refusing asylum claims, whilst contributing to the deterrence narrative. What is clear is that the UK is investigating options to externalise its border control operations and minimise its responsibility.

Externalisation is defined as ‘the process of shifting functions that are normally undertaken by a State within its own territory so that they take place, in part or in whole, outside its territory’ (Cantor et al. 2022, p. 122). Such processes manifest in several ways, with Dastyari, Nethery and Hirsch identifying six ‘methods’ of externalisation: maritime interception (encompassing ‘pushbacks’), extraterritorial processing and detention, third-country interception (including ‘pullbacks’), bilateral cooperation, public messaging campaigns, and multilateral cooperation (Dastyari et al. 2022). States view such border control as a legitimate exercise of state sovereignty to regulate the flows of people and goods into their territory (Moreno-Lax and Lemberg-Pedersen 2019, p. 15). Externalising border control relies on this right to regulate entry while disregarding the corresponding State obligations. As will be examined, such policies utilise ‘distance creation’ to ensure that the duties owed by the state to the individual do not apply because the individual is not physically present within the State’s territory. The UK, in exploring externalisation, is attempting to reduce its responsibility for migrants by extending its sovereign power abroad without a corresponding increase in responsibilities (Moreno-Lax 2024, pp. 29–30). This leads to a border designed to block unwanted migrants, with the so-called ‘meta-border’ shifting to maximise control while minimising accountability (Moreno-Lax 2024, pp. 29–34). Mann explains that externalised border enforcement is a strategic choice—intentionally shifting responsibility and delegating coercive authority to prevent asylum requests (Mann 2025). This involves international cooperation being used to transfer responsibility, deter arrivals and evade accountability through ‘architectures of repulsion’ (Nicolosi 2024, p. 6; FitzGerald 2020, p. 6).

Externalisation policies inherently entail cooperation between States, most often between a global North ‘sponsoring’ state that seeks to externalise some of its border controls to a global South ‘cooperating’ state (Hathaway and Gammeltoft-Hansen 2015; Ovacık and Spijkerboer 2024). Gammeltoft-Hansen and Vedsted-Hansen highlight the unprecedented degree of cooperation through bilateral and multilateral agreements that cede and pool sovereignty evident in externalisation practices (Gammeltoft-Hansen and Vedsted-Hansen 2016, p. 1). Cooperation takes many forms but specifically for our purposes, it can involve transferring individuals to a third country for processing or return, with a primary focus on containment, deterrence and security.

In light of this context, this article examines the challenge of holding states, particularly the UK, accountable for breaches of human and refugee rights that occur when externalised border control measures are implemented. It will focus on the responsibility that can arise from externalised or ‘offshore’ asylum processing or returns, as these are the forms of externalised border controls that the UK government has been considering for implementation. Return hubs and offshore asylum processing differ from externalisation practices that aim to prevent arrivals, such as pushbacks and maritime interdiction. The former involves removing those who have already arrived, whereas the latter seeks to prevent arrivals altogether. However, it is worth noting that all ‘methods’ of externalised border controls seek to deter arrivals, outsource obligations and avoid responsibilities. This article addresses the idea that states can shift responsibility to other actors by presenting a spectrum of responsibility that can shift with the different manifestations of cooperation that arise when a state’s obligations are externalised. The UK context serves as an illustrative example of how States attempt to distance themselves from responsibility; the spectrum of responsibility outlined here can be applied beyond this context to challenge externalisation practices in their various forms.

The argument will utilise Moreno-Lax’s elucidation of ‘irresponsibilisation’ and ‘distance-creation’ to understand the current phenomena (Moreno-Lax 2024; Moreno-Lax and Lemberg-Pedersen 2019). It complements discussions on ensuring accountability for externalisation practices by presenting the ‘spectrum of responsibilities’ that a state can incur for externalised border control policies (Tan and Gammeltoft-Hansen 2020; Pijnenburg and van der Pas 2022, 2025). In line with Fink’s work, which suggests a ‘spectrum’ of cooperative policies in externalised border practices (Fink 2016, p. 276), this paper proposes a corresponding ‘spectrum of responsibility’ that can apply depending upon the manifestation of the cooperation.

Fink’s ‘spectrum’ of cooperation ranges from where a ‘sponsoring’ State has de facto control of the other actors’ conduct to involvement that may amount to little more than mere incitement or political support (Fink 2016, p. 276). Such thinking is also demonstrated in Gammeltoft-Hansen and Hathway’s proposed typology of non-entrée policies, which illustrates the various mechanisms and forms of cooperation that states have developed to conduct deterrence (Hathaway and Gammeltoft-Hansen 2015, pp. 249–56). These range from diplomatic relations to direct financial initiatives, the provision of resources and training, the deployment of immigration officials, joint or shared enforcement, and direct migration control within the territory of the cooperating state, as well as joint operations. In line with Fink, it is evident that cooperation here is on a spectrum ranging from the most distant to joint conduct.

This paper argues that responsibility cannot be avoided through cooperation, distance-creation and meta-borders. It demonstrates, following Trapp, that state responsibility exists on a spectrum that can apply to UK conduct when it externalises the processing of asylum and returns. This spectrum assigns the level of responsibility correlative to contributions in rights violations (Trapp 2015, p. 245). At one end is direct responsibility for the breach of negative obligations, requiring the ‘sponsoring’ state, such as the UK, to maintain jurisdiction through contactless or functional control, making both states jointly responsible for breaches. When control is insufficient to establish jurisdiction, responsibility shifts to aiding and assisting in breaches if the sponsoring state is aware of them and significantly contributes to their occurrence. The spectrum also includes responsibility for failing to meet positive duties toward individuals affected during offshore operations. This spectrum addresses how states attempt to ‘irresponsibilise’ migration policy.

The contribution’s scholarly importance lies not only in consolidating and broadening existing discussions on how to determine state responsibility by situating it on a spectrum, but also in examining in greater detail the role of ‘aiding and assisting’ under Article 16 ARSIWA, advocating it as a crucial pathway to establishing that responsibility. Whilst the spectrum is presented in its entirety, the article emphasises the importance of establishing responsibility for ‘aiding and assisting,’ as this is an underutilised avenue for accountability. When states knowingly use externalised policies to avoid responsibility but maintain effective control, direct responsibility should be the preferred route for accountability. However, derived responsibility is an essential complement when a Court is unable to establish a jurisdictional link, serving as a strategic litigation tool. Even without effective control triggering jurisdiction, aiding and assisting can hold a state accountable for its knowing contribution to rights violations.

The article will proceed as follows: It will first (in Section 2) outline how the ‘irresponsibilisation’ of externalised border controls occurs and why States seek to utilise such policies. The article will then examine how irresponsibilisation impacts the ability of international law to ensure direct responsibility for breaches that occur within the context of the UK’s policies on offshore asylum processing and returns (Section 3). In response, it will explore how the rules prohibiting ‘aiding and assisting’ can be utilised to bolster attempts to hold the UK, and other States, that utilise externalisation practices, responsible for breaches of international obligations that occur (Section 4). The article will conclude by summarising how the responsibility spectrum can respond to attempts to shift responsibility away from ‘sponsoring’ or ‘destination’ states and prevent the irresponsibilisation of migration policies through distance creation.

2. The ‘Irresponsibilisation’ of Externalised Border Controls

2.1. Distance Creation Through Responsibility Diffusion and Denial

Externalisation agreements are de jure about strengthening cooperation in the field of migration but are de facto about deterring immigration (Muharremaj and Cami 2024, p. 1). States argue that they are legitimate if democratically approved and non-coercive, thereby protecting rights. Global North States avoid legal and practical scrutiny by shielding behind the sovereignty of third states. This raises challenges related to the application of refugee and human rights laws, accountability, shared and individual responsibilities and jurisdictional concerns (Pijnenburg et al. 2018).

This is by design. The logic of seeking to ‘externalise’ border controls is predicated on distance-creation, which attempts to reduce responsibility and avoid triggering legal responsibilities that have traditionally been viewed as territorial in scope (Nicolosi 2024, pp. 10–11). As such, extraterritorial agreements are predicated on the assumption that human rights won’t apply (Gibney 2021; Hathaway and Gammeltoft-Hansen 2015, pp. 249, 257). This focus on territoriality creates a double standard, allowing States to exploit geographical distance to justify ethical and legal detachment, contrary to the rule of law (Moreno-Lax 2024, p. 22). The resulting ‘irresponsibilisation’ is utilised to reconfigure the relationship between law and territory. Moreno-Lax characterises this as the meta-border, which extends sovereign power extraterritorially, while traditionally territorial conceptions of jurisdiction, state authority and responsibility contract legal obligations and rule of law safeguards (Moreno-Lax 2024, pp. 29–34). States use the law to organise irresponsibility and deflect accountability for harms suffered, which facilitates the dispersal, denial, and diffusion of responsibility (Veitch 2001; Cohen 2001).

Distance creation thus serves to shift responsibility, disperse legal duties, blur causation lines, and complicate attribution. There is both a spatial dimension, through the remoteness of the control, and a relational dimension, through the multiplicity of actors engaged in multi- or bi-lateral interactions that attempt to disperse the control and correlative responsibility (Moreno-Lax and Lemberg-Pedersen 2019, p. 5). As such, Moreno-Lax and Lemberg-Pedersen identify two levels of irresponsibilisation—responsibility diffusion and denial. Responsibility diffusion occurs in multi-actor contexts where causation and attribution become unclear and cooperation serves to generate a distance between the State and the migrant (Moreno-Lax and Lemberg-Pedersen 2019, pp. 19–25). This distance allows states to disclaim and disperse responsibility for breaches of international obligations by distancing themselves, both territorially and relationally, from the individual (Goodwin-Gill 2007; Davitti 2017). Responsibility denial involves rejecting accountability by claiming it belongs elsewhere and utilising distance to deny the triggering of international obligations (Moreno-Lax and Lemberg-Pedersen 2019, pp. 25–32). This ethical distance—both literal and figurative—removes migrants from sight, denying liability by implying a lack of intent and control.

2.2. The Problem of Irresponsibilisation and Externalisation

The problem with irresponsibilisation in this context is that externalisation practices inherently raise human rights concerns and endanger rights under the Refugee Convention. They threaten protections against torture, inhumane treatment, and arbitrary detention, while obscuring obligations to ensure proper procedures, access to family life, the right to leave and effective remedies (Pijnenburg 2020, 2021; McDonnell 2024; Frelick et al. 2016). Externalisation can also lead to denied healthcare, trafficking, enforced disappearances, sexual violence, and chain refoulement (Pijnenburg 2020, p. 315). The effects of externalisation often extend far beyond direct denials of an individual’s human rights, encompassing more wide-ranging yet significant breaches such as neglect, health-related harms, and even death. These consequences are argued to be by design, as a form of punishment that forces people to live in ‘state[s] of injury’ (Dehm et al. 2022). Loghan argues that ‘active neglect’ is a deliberate strategy to create hostile conditions, forcing refugees to abandon their claims (Loughan 2022). Negishi calls this ‘neo-refoulement’, where asylum seekers are forced to accept repatriation due to deprivation and lack of protection (Negishi 2024, p. 156). When externalisation involves neglect, like violence and subsequent human rights violations, it becomes considerably more difficult to attribute these to the State’s conduct.

The inherent link between externalisation and human rights breaches exists regardless of the state in which the externalised border control is conducted, though when a cooperating state has weaker human rights protection mechanisms, one can expect the threats to be heightened. This is due to the blurring of responsibility, the lack of oversight and accountability mechanisms and safeguards (Nicolosi 2024). Distance creation intends to obfuscate safeguards and responsibilities (Ghezelbash 2020). As such, externalised border control demonstrates a denial of States’ international obligations to protect the human rights of migrants. Externalisation lacks individual assessment and safeguards, increasing violence against migrants, human rights violations, summary expulsions, and ‘violent containment within territories where the risk of harm, persecution, or ill-treatment is high’ (Moreno-Lax 2024, p. 31). In this context, robust accountability mechanisms must be in place.

Regarding ‘return hubs’, the United Nations High Commissioner for Refugees (UNHCR) has stated that return hubs can improve the process of return and resettlement, provided they are developed in line with human rights standards (UN High Commissioner for Refugees (UNHCR) 2025). Indeed, extraterritorial processing has long been a part of the resettlement process; however, this new form of processing, aimed at deterring migrants from entering the territory, represents a shift in approach (Goodwin-Gill 2007, p. 26). The Refugee Council and Amnesty International have warned that return hubs will be unworkable and lead to human rights violations, with little evidence that rights will be upheld (The Times 2025; Amnesty International UK 2025). Crisp argues that sending people to return hubs will amount to forcible relocation and will coerce people into prison-like conditions for an unspecified length of time, despite the lack of a criminal offence (Crisp 2025). This likely amounts to arbitrary detention, undermining the proviso that return hubs would be implemented in line with human rights standards. Further, they impede appeals processes, which are vital in ensuring that individuals are not wrongly denied asylum and uphold the right to remedy.1 The aim of return hubs, like all externalised border controls, is to confine migrants and remove them from legal protections. This results in reduced access to legal safeguards, faster returns, and the marginalisation of migrants (Action Aid 2025). While guarantees are formally provided, they are largely denied in practice, and effective recourse is systematically hindered—using ‘return hubs’ risks weakening protections and processes.

Utilising externalised asylum processing and return hubs risks breaching refugee and human rights protections and leading to further rights abuses through chain refoulement, neglect, trafficking, slavery and sexual violence. Without rigorous democratic scrutiny and oversight to ensure accountability for rights violations, the use of externalised border controls risks creating dangerous responsibility gaps and systematic human rights abuses.

2.3. Responsibility on a Spectrum

Should breaches of the rights of individuals transferred to third, ‘cooperating’ countries by the UK government occur, the UK has argued vis-à-vis the Rwanda agreement that they will not have jurisdiction over these individuals, a point disputed by UNHCR (UN High Commissioner for Refugees (UNHCR) 2022). Given the extent of state engagement with externalised border controls through cooperation with third countries, the question then arises as to the processes whereby the responsibility of ‘sponsoring’ states can arise. Despite states’ attempts to utilise distance creation through externalised border controls, Gammeltoft-Hansen and Hathaway argue that the territorial delimitation of international legal obligations has been rejected; thus, States cannot avoid responsibility for conduct they undertake externally (Hathaway and Gammeltoft-Hansen 2015, pp. 244–47; Ammur v France, para. 52). Whilst the distancing of the State from the externalised border renders the application of international legal obligations complex, this article proposes that responsibility will still arise. Responsibility falls on a spectrum correlative to the level of cooperation and control maintained by the sponsoring State. As a result, distance-creation cannot erase responsibility.

3. Establishing Direct Responsibility for Externalised Border Controls

This section will first explore the establishment of responsibility for direct breaches of negative obligations, before presenting the responsibility framework for breaches of positive obligations. This article argues that establishing responsibility for direct violations of negative obligations is preferable, as it ensures a higher level of accountability in correlation with increased cooperation and control; however, externalisation policies pose challenges in establishing direct responsibility. Where distance-creation is such that the sponsoring state has little control over the cooperating state, cooperation will be minimal. However, responsibility can still arise for the state’s failure to uphold its duties to prevent breaches where it is in a position to do so.

3.1. Responsibility for Breaches of Negative Obligations

The first type of responsibility on the spectrum to consider is direct responsibility for breaching negative obligations under international and regional instruments. If the UK externalises its border controls, through third-country asylum processing or return hubs, this can ensure it is held responsible as the main, or joint, actor. The following will examine how direct responsibility can arise for the UK in the context of its externalised border controls. Under the law on state responsibility, the first requirement for responsibility to arise is that the conduct in question can be attributed to the State (ARSIWA Articles 2, 4–11). The second requirement is that the conduct breached one of the State’s international obligations (ARSIWA Articles 2, 12–15). These factors will be considered in turn to outline the challenges that externalisation policies raise when seeking the direct responsibility of the ‘sponsoring’ state.

3.2. Attribution

There must be conduct that caused or contributed to the breach that can be attributed to the state. As discussed, in the multi-actor context of externalised border controls, this can be complicated, especially when the State has intentionally distanced itself from the breach. The Articles on Responsibility of States for Internationally Wrongful Acts (ARSIWA) outline several ways in which attribution can be established. Most straightforward is that a state will always be held responsible for the actions and omissions of its own organs, even if they act ultra vires (ARSIWA Articles 4 and 7). For example, if the Home Office establishes or sends people to asylum processing centres in breach of non-refoulement rights, the UK is responsible. Equally, if UK officials oversee these centres and violate rights during the course of their management, this conduct is directly attributable to the UK. The direct level of control and cooperation makes attribution easier.

Given the distance creation inherent to externalised practices, it is unlikely that the UK’s conduct will be this direct. The ARSIWA does facilitate attribution in such contexts. This is derived from the notion that a State can control conduct, not just of its official agents or organs, but also of a private person or groups of persons if they are ‘acting on the direction, control or instructions of’ that State (ARSIWA Article 8, Commentary p. 48). Further, the State may instruct, direct or control agents of another state such that these actions are attributable to the directing State (ARSIWA Article 17; Pijnenburg 2021, pp. 199–203). These instances are more likely to be relevant where the UK maintains ‘contactless control’ over the asylum processing centre or return hub through the provision of funding and training of other officials, monitoring and strategic oversight and coordination of the centre and personnel (Moreno-Lax and Giuffré 2017). Here, should the UK maintain ‘effective control’ over the conduct of these organs or individuals, their conduct will be attributable to the UK (see for discussion of the effective versus overall control thresholds under Article 8: Cassese 2007; Milanović 2006).2 However, whilst it is possible that they would have contactless, effective control sufficient to fulfil article 8 ARSIWA, it is a high bar. Suppose the prevailing approach from the ICJ is followed. In that case, it must be demonstrated that the UK exercised a high level of strategic control and oversight, amounting to specific orders regarding the wrongful conduct. For example, the UK would need to have specific strategic oversight and control over the asylum processing centre or return hub, such that it is aware of and maintains conditions amounting to torture, inhumane or degrading treatment, or arbitrary detention.

3.3. Breach and Jurisdiction

The second aspect of an internationally wrongful act is that the attributable conduct must breach the State’s international obligations. As was discussed in Section 2, the use of externalised border controls is linked with breaches of numerous human and refugee rights. The UK is signatory to all the relevant human rights treaties, as well as the 1951 Refugee Convention, such that these obligations will apply to the UK’s conduct should its agents, or those acting with its authority, breach them.

There are two additional issues to consider when establishing a breach. First, how a breach can occur when multiple actors and actions may contribute to it. The ARSIWA acknowledges that a breach can consist of multiple acts or omissions that, in aggregate, constitute a single, composite act (ARSIWA Article 15). The wrongful act occurs when the conduct, in conjunction with other conduct, is sufficient to breach a primary obligation. Only when the conduct reaches the necessary threshold will the wrongful act be considered to have occurred. It then continues until the wrongful acts or omissions cease, as a continuing act (ARSIWA Article 14). Where conduct, by the UK and its partners, comes together to breach an individual’s human rights, for example, refoulement, torture or arbitrary detention, this can fall within the scope of a composite act. For example, where the UK sends an individual to an asylum processing centre in a third state, which it funds and manages, wherein the agents of the third state, trained by the UK, are committing acts of torture against detainees, the acts of both States would come together to constitute the breach. Each state could incur direct responsibility for its contribution to the composite act (Moreno-Lax and Lemberg-Pedersen 2019).

The second issue is the question of jurisdiction and the difficulty of establishing this in extraterritorial and multi-actor contexts. All relevant human rights conventions include a jurisdiction requirement, which triggers a state’s human rights obligations, primarily based on territorial principles. Gammeltoft-Hansen argues that the frameworks of human and refugee law presuppose that they do not have extraterritorial application (Gammeltoft-Hansen 2011, pp. 3–5). Conversely, externalised border controls are, by their nature, extraterritorial, and thus victims will not be in the territory of the sponsoring state. This is central to the logic behind externalisation policies, which create distance to deny the application of international law and thus responsibility. Applying human rights extraterritorially can ensure the continued application of a state’s obligations in external border control activities, especially where there is some ‘effective’ control. Externalisation emphasises the need to consider more composite notions of agency and responsibility that are decoupled from national territories (Moreno-Lax and Lemberg-Pedersen 2019, p. 15). These approaches will be examined to determine when the UK retains jurisdiction such that its human rights obligations would apply to externalised border controls.

There are divergent grounds for establishing extraterritorial jurisdiction in the jurisprudence of regional and international bodies.3 The first is effective control over some or all the territory abroad. This requires constant dominion over the foreign area to the extent that de facto jurisdiction is established and mirrors the de jure sovereignty of the State (Al-Skeini 2011, para. 138; Catan 2012, para. 106–7, 122–23; Moreno-Lax 2017, p. 276; Costello 2015, pp. 241–42). The threshold of ‘effective control’ is high, such that more remote cooperation between the sponsoring and cooperating states will not be sufficient for the UK’s jurisdiction to be triggered. However, if the UK maintains effective control over the return hub or asylum processing centre, its human and refugee rights obligations could continue to apply.4

The second ground is where the state exercises authority and control over individuals outside its borders (Al-Skeini 2011, para. 134–36; Issa and Others 2004, para. 71). This is so in three ways: the acts of diplomatic or consular agents when ‘they exert authority or control’ (Al-Skeini 2011, para. 134), when the state exercises ‘public powers normally exercised by the government’ (Al-Skeini 2011, para. 135) or the exercise of ‘physical power over persons’ (Al-Skeini 2011, para. 136) through arrest, detention, abduction or extradition. The personal model is echoed by the Human Rights Committee (HRC) in General Comment No. 31, which holds that States exercise jurisdiction over those ‘within the power or effective control of the forces of the State Party acting outside its territory’ (UNHCR GC 31, para. 10). The latter two aspects of the personal model can provide avenues for investigating the UK’s extraterritorial jurisdiction in regard to externalised border controls. The ‘physical power over persons’ model could be triggered should agents of the UK detain individuals within an asylum processing centre or return hub, under the control of its state agents (Hirsi 2012; Medvedyev and others v. France 2010; Xhavara and Others v. Italy and Albania 2001; Pijnenburg 2021, p. 146). This would require the UK to maintain a constant presence within the centre that is attributable to the UK.

The public powers model could also ensure the triggering of the UK’s jurisdiction, where it exercises all or some of the public powers normally exercised by that government (Hathaway and Gammeltoft-Hansen 2015, pp. 267–72). This will only be so if the acts in question are attributable to the UK, not the cooperating state, and must be exercised with the consent or invitation of that country’s government (Al-Skeini 2011, para. 135). As asylum and return processing are public powers that the UK can exercise with the cooperating state’s consent, they could do so on another state’s territory, either directly or through delegated agents; this may be enough to trigger jurisdiction.5 Both aspects of the personal model require the UK to maintain some physical presence extraterritorially and have direct control over the individual. This may not trigger the UK’s jurisdiction if they maintain ‘contactless’ control over externalised border controls.

The third ground for extraterritorial jurisdiction is the functional model, evident in HRC guidance6 and international jurisprudence (López Burgos v Uruguay; A.S. and others, HRC GC36; HRC GC31), but less so in the approach of the European Court of Human Rights (ECtHR) (M.N. and others v. Belgium 2020; Georgia v. Russia (II) 2021). A functional approach depends on the connection between ‘public powers’ and sovereign authority, which links the State to the individual through its exercise of governmental functions, regardless of where the authority is exercised (Moreno-Lax 2020; 2024, sec. 6; Giuffrè 2021). Exercising public power triggers the state’s jurisdiction, which in turn leads to responsibility for any resulting violations (Moreno-Lax 2024, p. 50; Pijnenburg 2021, p. 153). What is important is the public nature of the power being exercised. This model holds a State responsible for ‘sufficiently proximate repercussions’ of its actions on rights, including those perpetrated via proxy, even if these effects happen outside national borders (GC31, para. 10). The functional control model ensures responsibility when a breach occurs as a result of foreseeable consequences of exercising public powers. Such an approach would ensure that, even where the UK maintains contactless control of the return hub or asylum processing centre, the distance creation would not limit its obligations to uphold human rights when it exercises public powers, i.e., overseeing the asylum or return process within the cooperating state.

There is evidence of support for the functional approach in previous case law of the ECtHR, where the Court links human rights obligations to the State’s ability to realise human rights (Ilascu 2004, para. 392; Issa and Others 2004, para. 71). However, the functional approach was proposed in the recent SS case, and the Court’s inadmissibility ruling demonstrates the challenges of applying a functional model. The Court rejected the possibility of a ‘functional’ interpretation of jurisdiction. The court narrowed jurisdiction to cases where the State has effective control over territory and persons, negating the public powers model and undermining the functional control argument. As a result, effective control was confined to ‘physical’ control, meaning that ‘contactless control’ is also precluded (S.S. and Others v. Italy 2025, para. 78, 100–108). Further, the Court argued that the financial and technical support was insufficient to establish that the Libyan authorities were under the effective control and decisive influence of Italy per Article 8 ARSIWA.7

This restrictive approach is disappointing; the ECtHR could have clarified the law to ensure accountability when a State externalises its border controls and commits breaches of human rights violations by proxy. The approach diverges from the HRC, perpetuating a discrepancy between European and international human rights law. Mann argues that interpretive traditions constrained the court, causing it to defer to principles of sovereignty and territoriality (Mann 2025). Arguably, political pressure from States to protract its jurisdiction, especially regarding issues of migration, is at play in the Court’s restrictive approach (Italian Government Website 2025). Ultimately, the SS decision highlights the challenges that arise when seeking to ensure direct responsibility for externalised border controls, especially in the European context.

There are other approaches to extraterritorial jurisdiction proposed by scholars. Milanović argues in favour of a territorially unlimited approach to negative obligations (Milanović 2011) while Gammeltoft-Hansen recommends re-examining the extraterritorial effects doctrine when sponsoring States utilise distance-creation to avoid triggering jurisdiction while conducting border control (Gammeltoft-Hansen 2018; Gammeltoft-Hansen and Vedsted-Hansen 2016, p. 10; Pijnenburg 2021, p. 199). Effects jurisdiction may arise as a result of actions taking place within a State’s own territory, leading to human rights violations on the territory of another State. There is thus no requirement that the responsible State itself acts extraterritorially (Georgia Andreou v. Turkey 2008, p. 11; Nada v. Switzerland 2012). The ECtHR’s has vaguely held that ‘acts of the Contracting States performed, or producing effects, outside their territories can constitute an exercise of jurisdiction’ (Hirsi 2012, para. 72; Al-Skeini 2011, para. 131–33; Solomou v. Turkey 1999). The Court looks to whether a ‘direct and immediate’ relationship exists between the State’s own actions and the human rights violation suffered within another territory.8 The Inter-American Court on Human Rights (IACtHR) also found that a State exercises jurisdiction ‘if there is a causal link between the action that occurred within its territory and the negative impact on the human rights of persons outside its territory’ (Advisory Opinion on “The Environment and Human Rights” 2017, para. 104(h)). The effects doctrine supports the argument that the UK has jurisdiction over human or refugee rights violations occurring during externalised border controls, whether at a return hub or asylum processing centre, because the UK acts by sending individuals to these centres from its territory. This jurisdiction is further justified by the UK’s role in maintaining the centres through funding and training.

Finally, there is an argument that evolutionary approaches can ensure direct responsibility in the externalisation context. Jackson proposes reimagining Soering as a narrow preventive complicity rule and extending it to other forms of complicity in torture (Jackson 2016). In Soering, the Court established that the extraditing State is responsible under Article 3 European Convention on Human Rights (ECHR) if the claimant faces a ‘real risk’ of torture or inhumane treatment upon return, even if they are no longer under the State’s control. Responsibility thus depends on ‘foreseeable consequences outside the State’s jurisdiction’. Jackson argued that this logic can apply to other forms of complicity in Article 3 ECHR breaches, including offshore asylum processing. Like Jackson, Gammeltoft-Hansen and Hathaway propose an evolutionary approach, arguing that non-refoulement obligations can apply without territorial limits (Hathaway and Gammeltoft-Hansen 2015, p. 283).

Responsibility varies based on the nature of the cooperation. Direct responsibility arises when the state controls externalised border measures and wrongful conduct can be attributed. Employing a functional or effects approach to UK control over externalised border measures can ensure its responsibility for violations. While the ECtHR’s approach is narrow, the HRC offer alternative methods to hold the UK accountable. Responsibility cannot be avoided simply because of distance; it is tied to the nature of the cooperation. Nevertheless, SS highlights the complexities involved in assigning direct responsibility in externalised contexts. Notably, the Court did not address the issue of complicity, despite referencing Article 16 ARSIWA in its legal analysis. This suggests that it is a relevant legal avenue, but it has not been sufficiently explored, leading the Court to deem it unnecessary to investigate further. While Moreno-Lax correctly argues that externalisation policies cause rights violations by proxy (Moreno-Lax 2024, p. 49; Moreno-Lax and Lemberg-Pedersen 2019, p. 29)—an issue that warrants more than mere complicity—if the distancing effectively shields the State from the violation, then Article 16 ARSIWA offers a valuable legal tool for accountability (Pascale 2019; Mackenzie-Gray Scott 2019).

3.4. Responsibility for Breaches of Positive Obligations

At the opposite end of the responsibility spectrum are breaches of positive obligations. This approach can be effective when the level of distance-creation by the supporting State renders the evidentiary or legal hurdles for establishing responsibility for breaches of negative obligations or aiding and assisting insurmountable. Pursuing responsibility for violations of positive obligations responds to the distant relationship that cooperating States can utilise to implement externalisation policies, ensuring that the state remains responsible for failing to prevent violations of human and refugee rights, even when it maintains minimal control. Such cooperation in border controls might encompass offering funding, political backing, diplomatic support, and resources. Concerning the UK’s externalisation policies, this could mean providing funds and resources to establish asylum processing or return hubs in a partner country, without exercising any extraterritorial control or oversight.

Positive obligations, grounded in human rights law and often framed in terms of due diligence9, impose duties on states to take reasonable and timely measures to prevent, investigate, and punish violations committed by private actors or foreign agents (see for example, Yin 2022). The State can breach its human rights obligations not because of actions by private actors but due to its own omissions that contribute to the harm (Stoyanova 2023). The ECtHR cites Article 1 of the Convention, which requires Parties to secure the rights outlined in the ECHR. Such obligations arise due to proximity and ability to influence the occurrence of the breach (Ollino 2022, p. 3). In the context of the UK’s externalised border controls, breaches of the UK’s positive obligations would include failures to prevent the human rights abuses previously explored, in an asylum processing or return hub to which the UK has sent individuals in a third country, as well as due diligence obligations regarding investigating and punishing instances of such abuses by agents or actors within its control.

Liguori has explored the overlap between complicity and breaches of positive obligation, arguing that courts have preferred to focus on breaches of positive obligations to establish responsibility (Liguori 2022; den Heijer 2013). There are advantages to seeking responsibility for breaches of positive obligations. Crucially, the threshold for breach of positive obligations is significantly lower than that required to establish responsibility under Article 16 ARSIWA. The latter requires that a state provide aid or assistance with knowledge of the circumstances of the internationally wrongful act and with the intent to facilitate it. In contrast, positive obligations can be violated where a state merely ‘knew or ought to have known’ of the risk of serious harm and failed to act (Nasr and Ghali v Italy 2016). The Genocide case outlined the overlapping functions of the two categories (para. 430). Here, the ICJ declined to find Serbia aided and assisted in the Srebrenica genocide due to the lack of intent; however, the Court nonetheless held that Serbia had violated its positive obligation to prevent genocide, based on the due diligence standard (para. 430–38). The Court emphasised that Serbia was in a position of influence over the perpetrators, due to strong political and financial ties (para. 434–38). It concluded that while full knowledge was lacking, Serbia was or should have been aware of the serious risk and thus had a legal duty to prevent the genocide (para. 438). This demonstrates that, in the context of externalised border controls operations, the UK could be held responsible for its failure to uphold its positive obligations to prevent refoulement, arbitrary detention and other human rights abuses, where it was or should have been aware of the risk of such violations occurring.

The ECtHR has similarly adopted a preference for resolving complex state responsibility claims through the lens of positive obligations. In El-Masri v Macedonia, the court acknowledged the possibility that aiding and assisting may suffice for attribution, but ultimately relied on direct responsibility through the failure to prevent and facilitation by state agents (para. 206, 211). In Nasr and Ghali v Italy, the ECtHR relied on positive obligations to hold Italy responsible because it failed to take reasonable measures to prevent a risk of ill-treatment. The Court accepted that the state “should have known” of the danger, satisfying the due diligence standard (para. 294). This standard was also applied in Al-Nashiri v Poland and Husayn v Poland where the Court stated that Poland ‘ought to have known’ that the torture would have occurred, based upon an awareness of serious risk as opposed to virtual certainty (para. 512, 517). Interestingly, in the Al-Nashiri and Husayn judgments, the Court differentiated actions that go beyond a failure to prevent into the realms of complicity. The difference appears to be in the manner of the assistance, amounting to more than a failure to act but to actual active assistance (para. 503–19). As such, breaches of positive obligations are an avenue for establishing the UK’s responsibility, where it has minimised control such that it has not actively assisted in the human rights abuse at the asylum processing centre or return hub, but has failed to prevent it and demonstrably ought to have been aware of the risk.

The strategic and evidentiary advantages of relying on breaches of positive obligations are apparent. The standard of proof is lower, and states may be held responsible for their own omissions, rather than for conduct attributable to another state or non-state actor.(Liguori 2022) As established in the Corfu Channel case, for breaches of positive obligations, the Court does not need to take into account the actions of another State, as it is the state’s own failures that derive responsibility, such that it dispenses with the monetary gold principle (see Corfu Channel Case).

There is also evidence that positive obligations can apply extraterritorially (den Heijer and Lawson 2012, p. 108). The ECtHR has certainly shown a willingness to adopt a lower jurisdictional threshold for positive obligations, emphasising capacity to influence rather than territorial control.10 In Manoilescu, the court dismissed the idea that effective control is a necessary prerequisite for establishing jurisdiction.11 The duty to prevent thus arises from the influence a state exercises over a specific situation. In line with this, the Human Rights Committee has also applied an extraterritorial model to failures to protect at sea (A.S. and Others). Vandenhole has posited a similar due diligence-based jurisdiction that focuses on the effectiveness of protection and proximity of the risk.12 Conversely, Milanović argues that positive obligations depend on effective control, as this is necessary for their implementation (Milanović 2011, pp. 215–16). Stoyanova is also sceptical about the possibility of extraterritorial obligations applying extraterritorially, finding that the Court lacks the tools to thoroughly examine whether the rights can be applied in this context (Stoyanova 2023, chp. 8).

Relying solely on positive obligations also raises some normative concerns. Positive obligations generally pertain to omissions and failures to protect, which are less serious than acts of intentional facilitation. A finding that a state has failed to prevent harm carries less stigma and legal weight than a determination of complicity, which suggests active participation or intentional support (Moreno-Lax 2025, 2024; Ollino 2022). There is also a risk that states might evade accountability for active roles in serious human rights violations by reinterpreting their involvement as mere inaction or systemic failure. Further, a clear set of legal requirements for protective duties is yet to develop, particularly in the ECtHR’s inconsistent approach, and there is a lack of clarity on the extraterritorial application of positive obligations.13

Despite these challenges, the turn to breaches of positive obligations has proven to be a practical and pragmatic means of establishing state responsibility. It can facilitate accountability where jurisdictional hurdles, arising from distance-creation, preclude direct attribution. As such, when considering the UK’s responsibility for breaches of its positive obligations during externalised border controls, this avenue for litigation should ensure that the UK is held responsible for failing to prevent human rights violations that it ‘knew or should have known’ would occur. This will be the case where it maintains a ‘position of influence’ (per the Genocide Case, para. 434–438) through financial, political, or other support to the cooperating state, such that it has the capacity to influence the occurrence of the breach.

The courts’ preference for addressing breaches of positive obligations is not a dismissal of Article 16 ARSIWA but demonstrates the lack of clarity on the thresholds for establishing a state has aided and assisted in a breach of human rights. This is a crucial legal mechanism for establishing responsibility for intentionally aiding and assisting internationally wrongful acts. A balanced approach that combines both frameworks is essential to ensure accountability and normative clarity in international law.

4. Derived Responsibility for Aiding and Assisting in the Breach

Where the level of control and cooperation over externalised border control is insufficient for human rights to apply extraterritorially, the State may deny the applicability of international legal obligations despite their indirect support of externalised border controls. However, when the UK cooperates with a third state through the provision of aid and assistance, it is often still sufficient to trigger its derivative responsibility, provided there is awareness that breaches of an individual’s human or refugee rights would result (ARSIWA Article 16; Gauci 2017). Derivative, or ancillary, responsibility arises in line with the level of cooperation and the contribution made towards the commission of that wrongful act. A State will be responsible for aid or assistance in the commission of a wrongful act to the extent that the aid and assistance given brought about the wrongful act (ARSIWA Commentary, Article 16, p. 66; Crawford 2013, p. 399). The responsibility of the assisting state is limited in three ways: there must be awareness of circumstances that make the conduct of the assisted state wrongful; aid must facilitate the wrongful act; and the completed act must be wrongful if committed by the assisting state itself (ARSIWA Commentary, Article 16, p. 66).

To incur derived responsibility, the State’s conduct need not amount to a breach of its human rights obligations and hence, does not need to exercise jurisdiction or have the acts in question be attributed to it (Pijnenburg 2021, p. 205). It assumes that the cooperating state commits the wrongful act. Thus, responsibility under Article 16 ARSIWA can address the complex causation and attribution that result from distance creation and attempts at irresponsibilisation (Pijnenburg and van der Pas 2025, p. 5; Gammeltoft-Hansen 2011; Gammeltoft-Hansen and Vedsted-Hansen 2016). In the context of third country processing, where a state aids or assists in breaches of international human rights or refugee law obligations, the supporting state would share responsibility to the extent that its own conduct caused or contributed to cause the breach, where it had ‘knowledge of the circumstances’ of the wrongful act and the act would be internationally wrongful if committed by that state (ARSIWA Commentary, Article 16, p. 66). This section examines how the thresholds for establishing aid and assistance can be applied to the UK’s externalised border controls, namely (1) the nature of the aid and assistance, (2) when aid or assistance is sufficiently ‘significant’, (3) when it will be opposable, and (4) the level of knowledge of the circumstances required for the State to be held accountable.

4.1. Nature of Aid and Assistance

The sponsoring state must provide some material aid to the cooperating State that is then used to commit an actual breach of international law.14 The exact form of the aid or assistance is not limited and so responsibility can arise from any aid or assistance provided.15 It is not possible to make a definitive list of what support or conduct could constitute ‘aiding and assisting’. It will be some act of assistance or facilitation that must have had a material effect in the commission of the wrongful act. The context of each case must be taken into account to assess the contribution the support made to the commission of the wrongful act. Examples could include providing financing, a facility or territory, weapons or other military assistance, logistics support, or intelligence (ARISWA Commentary, Article 16, para. 1). Hathaway and Gammeltoft-Hansen highlight that assistance in conducting cooperative deterrence can be characterised within the ambit of aiding and assisting (Hathaway and Gammeltoft-Hansen 2015, p. 279).

There would be a connection between the aid and assistance provided by the UK and countries that it sponsors to conduct asylum or return processing. This would be through finance and resourcing, sending equipment and training, building the administration of the cooperating state and training officials, and potentially providing ongoing strategic oversight. There would be material aid in the maintenance of asylum processing centres or return hubs, such that the UK would be aiding and assisting in any human or refugee rights that occurred there. The UK would acknowledge and intend that people would be detained in these centres awaiting processing of their claims or returns.

4.2. Significant Facilitation

A nexus must be established between the aid and the commission of the wrongful act. This defines the relationship between the supporting states’ conduct and the principal wrong. The aid and assistance must make some contribution to the wrongful act. The International Law Commission (ILC) Commentary defines that the aid or assistance must have made the wrongful act ‘materially easier’, but it need not be indispensable to the commission of the wrongful conduct (ARISWA Commentary, Article 16, para. 5). The assistance must be linked with the wrongful act and make a significant contribution to it.16 The threshold for significant facilitation is important to clarify and corresponds to the responsibility and cooperation spectrums. The facilitation must not be too indirect, for example, diplomatic pressure or mere instigation, as this will be insufficient to trigger derivative responsibility. Moynihan clarifies that assistance in a generalised way will not result in a breach (Moynihan 2016).

On the other hand, if the levels of facilitation and cooperation are direct, essential, and reach a sufficiently high threshold, then the support may become such that joint responsibility is established (ARSIWA Commentary Article 16 para. 5, ARSIWA Article 47; Crawford 2013, p. 401). There are two dimensions to the nexus element, scale and remoteness (Moynihan 2016, pp. 8–9). Regarding scale, the contribution must be significant to the occurrence of the wrongful act. For example, providing diplomatic support to the setting up of asylum processing centres would be insufficient. The remoteness dimension concerns proximity. There must be a close enough connection between the aid and the wrongful act. For example, giving development aid that is diverted to another purpose would be too remote. Conversely, the UK’s financing, training and supervision of asylum processing or return hubs where there were human rights abuses would be more than incidental.

The contribution by the complicit State is generally held to need to be in the form of a positive act such that participation must be active and direct. Therefore, omissions are commonly held to be excluded as a form of aiding and assistance (Hathaway and Gammeltoft-Hansen 2015, p. 279; for an alternative argument see Brown 2016). For example, in the Bosnian Genocide case the International Court of Justice (ICJ) held that complicity ‘always’ requires some positive action (para. 222).17 As such, when the UK has actively aided and assisted another state in setting up and managing an asylum processing centre or return hub, through signing international agreements, providing resources, training, equipment, or strategic oversight, this will likely be considered significant facilitation. Some have even argued that where states provide aid to third countries and make it conditional on effective cooperation on migration control, yet without ensuring international protection obligations are adequately upheld, they could be liable for aiding and assisting this wrongful conduct (Davitti and La Chimia 2015).

The UK will be responsible for aiding and assisting a cooperating state in breaching the rights of transferred asylum seekers if it ‘significantly facilitates’ those breaches. It is clear that the aid and assistance the UK would be providing to the cooperating states would be substantial; such return hubs or asylum processing centres would not exist without the support of the UK, whether through financial, strategic, material or personnel support. For example, the Memorandum of Understanding (MoU) the UK signed with Rwanda outlined that the UK would pay Rwanda significant sums of money as part of the Partnership and would facilitate the transfers of asylum seekers. Whilst future plans may look different to Rwanda’s, the level of facilitation is unlikely to lessen, as no state would agree to host the UK’s asylum seekers or people awaiting return without significant support. The UK would materially facilitate the arrival of individuals in these centres by sending them there from the UK.

4.3. Opposability

The act of aid or assistance may not in itself be wrongful for the State (Pascale 2019, p. 45), for example, the provision of arms through a trade deal is lawful and the practice of providing training and supplies to a State is also common practice. The wrongfulness derives from the act of the cooperating State utilising the aid or assistance, for example, using weapons or training, to commit human rights abuses. In such situations, responsibility will only arise where the supporting State was itself bound by the obligation violated by the party to which assistance was rendered (ARSIWA Article 16(b), Crawford 2013, p. 401). The pacta tertiis rule ensures that States can only be held responsible for breaches of rules to which they are bound. However, when the breached rule is a peremptory norm, customary international law or an erga omnes duty, then this will not be relevant (Crawford 2013, p. 409). Furthermore, as held in Hirsi, States cannot evade responsibility by relying on obligations arising from a bilateral agreement with Libya (Hirsi 2012, para. 129). As such, the existence of an international agreement to the contrary does not override the primary obligations of the State to uphold its human rights obligations (Graefrath 1996).

The UK is a signatory to the International Covenant on Civil and Political Rights (ICCPR), the ECHR, the Convention against Torture (CAT), and the 1951 Refugee Convention, as are most States in the world. In addition, it is unlikely, given the response to the Rwanda scheme, that the UK would be able to partner with a State that is not a signatory to these agreements, as it would be unable to claim it was a ‘safe third country’. Further, many of the breaches of international laws that could occur because of offshore processing are customary, i.e., prohibition of torture. As such, whilst the terms of any agreements to externalise the UK’s border controls and the support rendered by the UK may not inherently breach international obligations, should this aid and assist conduct by the cooperating states’ authorities that do breach both States’ international obligations, the UK may be complicit in these.

4.4. Knowledge

Establishing responsibility for aid and assistance also requires a mental, or subjective, element. The level of knowledge is the source of disagreement, partly because the ARSIWA and their commentary seem to conflict. The controversy revolves around whether the ARSIWA require that the sponsoring State intends to facilitate the wrongful act, or whether ‘constructive’ knowledge of the circumstances and the potential for the wrongful act to occur is enough. What is clear is that aid and assistance must be rendered with at least knowledge of the circumstances (Crawford 2013, p. 399). If the sponsoring State was utterly unaware of the cooperating State’s intentions regarding the wrongful act, then it cannot be found to have aided and assisted in the act. Liability should not follow when aid and assistance are given in good faith but are misused by the cooperating State, i.e., if development aid was used to refoule individuals. However, there are situations in which the State cannot claim not to have known, given the level of general awareness of breaches occurring (M.S.S. 2011).

A broad interpretation of the knowledge element establishes responsibility where States should have reasonably known of the likelihood of a breach occurring as a result of their aid and assistance. Thus, knowledge is required of the circumstances surrounding the provision of aid and assistance, to the extent that the assisting States can foreseeably see the wrongful act occurring because of their aid (Pascale 2019, p. 49). Moreno-Lax and Giuffre hold that Article 16 requires States to be aware that their assistance may foreseeably be used to perform wrongful conduct (Moreno-Lax and Giuffré 2017, pp. 102–3). Gammeltoft-Hansen and Hathaway argue that if a State has constructive knowledge—meaning it knew or should have known about the potential rights breaches—and still provides aid, this fulfils the knowledge requirement (Hathaway and Gammeltoft-Hansen 2015, p. 280; Nolte and Aust 2009). In Hirsi, Italy argued that it considered Libya a ‘safe third country’ and had no reason to believe Libya would evade human rights commitments. The Court rejected this, stating that reliable sources indicated the risks of abuses in Libya (Hirsi 2012, para. 128–31). In such circumstances, for a State to deny responsibility, despite their knowledge of the likelihood of the wrongful act occurring, would amount to ‘wilful blindness’, amounting to a deliberate effort by States to escape clear knowledge of the wrongful acts. There is a general refusal of courts to countenance ‘wilful blindness’ (Aust 2011, pp. 244–49). As such, it appears likely that the element of knowledge should be proved when the sponsoring state ‘should have reasonably known’ that its lawful act was facilitating the wrongful act of the assisted state.

The narrower approach is that the knowledge requirement is a form of intent. For the sponsoring State to be responsible, it must be aware not only of the circumstances but also of the cooperating State’s intentions to commit the wrongful act, including knowledge of the ‘specific purpose’ for which the aid is given or used (Crawford 2013, p. 406). In the Bosnian Genocide case, the ICJ held that there cannot be responsibility unless ‘the organ acted knowingly… and at least was aware of the specific intent of the principal perpetrator’ (para. 218, 421). The Court here required a high degree of particularity, of full knowledge, for the crime to be committed (Dominice 2010, p. 286). This aligns with Aust and Nolte, who find that more than mere knowledge is required; Aust proposes a level of knowledge ‘approaching intent’ (Aust 2011; Nolte 2005). It has been suggested that the sponsoring State must know and share the cooperating State’s intent (Graefrath 1996, p. 378). These arguments centre on the ILC commentary, which outlines that the aid must have been given with ‘a view to facilitating the commission and must actually do so’ (ARSIWA Commentary Article 16 para. 5). The intent here is to commit the wrongful act via the assistance, not merely to assist the cooperating state. This not only entirely eclipses the requirement of knowledge but also shifts the intent to dolus directus. Arguably, this crosses the line from aid and assistance to direct responsibility.

Key criticisms have been raised against this narrow approach (Quigley 1987, p. 111; Aust 2011, pp. 235–37; Ferro 2015, pp. 144–45). First, the requirement of intent is not found in the ARSIWA, which only requires knowledge of the circumstances. This discrepancy between the text of Article 16 and the Commentary has been attributed to a compromise between States (Pascale 2019, p. 51). Gaja holds that where there is a discrepancy between Articles and commentary, the former should be given greater attention (Gaja 2014). Second, this approach is not echoed in the case law. The ICJ implies that knowledge of the circumstances is sufficient, but a higher level of knowledge may be required, depending on the primary obligation that was breached as a result of the aid and assistance (Genocide Case, para 419–420). The high bar applied in the Genocide Case aligns with the primary obligation at issue, namely genocide, which itself has a high intent requirement. Third, is how a court should determine the state of mind of a State (Ferro 2015, p. 150). The Court must infer whether the State intended and knew the circumstances of the wrongful act. A State aiding in a wrongful act is unlikely to document this and may conceal its intent to avoid liability. Fourth, the State may not desire the harmful outcome but be motivated by other reasons. In such circumstances, the sponsoring state could knowingly aid in a wrongful act, motivated by economic gain, thus avoiding responsibility due to a lack of intent.

Moynihan and Chatham House suggest that there is no real distinction between intention and knowledge when it comes to State responsibility; the difference in practice is more superficial than real. Knowledge should be enough to imply a state’s intent to support these acts (Moynihan 2016, 2018). When one State assists another with knowledge, or virtual certainty, that its assistance will be used by the cooperating state unlawfully, this is sufficient to fulfil the intent requirement. In cases where there is no intent, the level of knowledge and material facts are the same as if they had intended the harm. With no sure way of identifying the intent of the State, it is arbitrary to require such a high threshold for a State to be held responsible.

As such, there is increasing agreement that the aid and assistance must be rendered with knowledge of the potential for a breach of international law to occur. Where a cooperating State sends a refugee to a country where they know their rights may be breached, it aids and assists in that breach (Goodwin-Gill 2007). When the UK offers aid or assistance to a cooperating State—whether through funding or agreements related to asylum processing or return hubs—it recognises the potential for wrongful acts to occur. The UK cooperates with these states specifically to manage and prevent migration-related issues (Davitti and La Chimia 2015). The UK intends to outsource border control despite the knowledge that doing so leads to human and refugee rights abuses. The UK is aware of the potential for human rights abuses and asylum procedure breaches through the use of externalised border control, as these have been extensively documented. The aid is conditional on the implementation of measures that lead to breaches of international law. Arguably, the UK also shares the intent to commit such abuses in order to deter people from entering the UK irregularly. Therefore, we see that holding the UK accountable for aiding and assisting in breaches that occur during externalised border control operations, where it maintains minimal control but continues to actively and knowingly support the operations, is a viable way to ensure responsibility. This middle ground on the spectrum of responsibility guarantees a level of accountability that correlates to the level of control over the breach, preventing States from avoiding responsibility for their active facilitation of human and refugee rights abuses merely because it creates distance between its territory and the breach.

Should the UK implement a policy of utilising return hubs or offshore asylum processing, and human and refugee rights abuses are reported, the UK could also be responsible for maintaining the situation and recognising it as lawful through the ongoing provision of aid and assistance (Pascale 2019). Under Article 41(2) ARSIWA no state should recognise as lawful a situation created by a serious breach, nor render aid and assistance. This article outlines two additional obligations that arise when a State breaches a peremptory norm of international law ARSIWA Article 40.18 As such, should the UK continue to aid and assist a State found to be placing people in inhumane or degrading conditions, conducting torture or arbitrarily detaining them for prolonged periods, then they will be in breach of Article 41(2) ARSIWA. There is no opposability or knowledge requirement under Article 41(2) as these are taken as a fact, given the nature of the breaches concerned (de Wet 2018). This would lead to responsibility for the UK’s conduct in aiding and assisting in the maintenance of a wrongful situation, and de facto recognising it as lawful through continued support.

4.5. Joint Responsibility

The final point to clarify is how derivative responsibility can be established in the multi-actor context of externalisation. As discussed, this complex causal situation is utilised to attempt to diffuse and deny responsibility by creating distance between the sponsoring State and the breach (Hathaway and Gammeltoft-Hansen 2015; den Heijer 2013). The principle of independent responsibility dominates international law, which focuses on the responsibility of a single state (Pijnenburg 2021, pp. 219–20). This can lead to only the cooperating state being held responsible, as it has the primary obligations and is likely acting through its organ on its territory, even though it may be acting with the support and guidance of the sponsoring State. This gives rise to exclusive responsibility for the cooperating State, and the sponsoring states avoid responsibility (Pijnenburg 2021, p. 222). However, under the umbrella of ‘shared responsibility19’, notions of joint20 and concurrent responsibility21 are developing. The ARSIWA provides room for joint responsibility when multiple actors work together to breach an international obligation. Consequently, where multiple States have engaged in a wrongful act, their contribution can be attributed to them, and responsibility can be assigned accordingly.

Under the ARSIWA rules, situations in which a single course of conduct can be attributed to several States is internationally wrongful for each of them. International law recognises that two or more states can be responsible for the same wrongful act, each liable based on their conduct and international obligations (Nollkaemper 2011). As such, under Chapter II of the ARSIWA, the same conduct can be attributable to several states simultaneously (ARSIWA Commentary, 34 para. 6). Where several States have contributed to the wrongful act, each State may have responsibility invoked against them for the act as outlined by Article 47 ARSIWA (Crawford 2002, p. 124). Each State will be separately responsible for the conduct attributable to it (Crawford 2002, p. 125). Orakhelashvili refers to the principle under which a state’s responsibility is not reduced even if another state is involved in the perpetration of a wrongful act (Orakhelashvili 2010). Consequently, responsibility cannot be diminished because more than one State is involved.22 Joint responsibility finds each State involved independently liable for its contribution to the wrongful act (Nauru v Australia 1992, para. 4). Concerning the complex causation of externalised border controls that may have been caused cumulatively by acts of multiple States, including the UK, Article 47 ensures that each State can be held responsible insofar as it caused the breach. International responsibility is not reduced or precluded because of the concurrent responsibility of a third State.23

4.6. Challenges to Establishing Complicity

The practical application of Article 16 remains complex. First, international courts rarely possess jurisdiction over both the principal and complicit states involved in such cases. Invoking Article 16 necessitates a finding that the assisted state committed an internationally wrongful act. The Monetary Gold principle bars a court from making determinations that effectively judge the conduct of a third state not before the court.24 In the Corfu Channel case, the United Kingdom invoked complicity, but this claim was dismissed due to jurisdictional issues related to the Monetary Gold principle. Instead, the ICJ applied the principle of due diligence, holding Albania responsible for its failure to prevent the wrongful act (Corfu Channel case, p. 4). The ECtHR often refrains from determining the complicity of a supporting state if such an assessment would necessitate an evaluation of the conduct of the primarily responsible state, over which the Court may lack jurisdiction.25 Jurisdictional issues thus hinder a Court’s capacity to assess the lawfulness of a third state’s actions. A claim against a state, such as the UK, for assisting another state in violating human rights and the rights of refugees requires a judgment on the conduct of the cooperating state. However, some scholars argue that establishing non-refoulement obligations already assess the conduct of a third party, such that in certain circumstances, the responsibility of a supporting state could be assessed without implicating the responsibility of the cooperating state (Jackson 2016; Hathaway and Gammeltoft-Hansen 2015).

Although the conduct of a complicit state could give rise to international responsibility under Article 16 of ARSIWA, this provision is generally structured for interstate invocation, which is rare in practice. Within the ECHR system—designed for individual petition rather than state-to-state complaints—the ECtHR remains the most plausible venue for addressing such claims. Fatima has argued that pursuing accountability is far more effective through primary norms within international human rights law (Fatima 2016). In contrast, Article 16 of ARSIWA functions as a general “roving” rule applicable only to inter-state claims, and it has limited practical utility at the domestic level. In the European context, the application of ARSIWA before the ECtHR is challenging in the context of externalisation, as the Court predominantly relies on its own primary rules (Hirsi 2012). Nevertheless, in Banković v. Belgium, the Court recognised that the law on state responsibility should be taken into account in its interpretation of the ECHR (para. 57). Similarly, in Hirsi, Judge De Albuquerque emphasised the relevance of general international law in the Court’s assessment of responsibility for aiding and assisting human rights violations (Hirsi, p. 80). Therefore, there is evidence the Court may be willing to pursue Article 16 cases.

Another approach involves acquiescence and connivance, which the ECtHR has utilised to address state assistance to third parties or non-state actors (Milanović 2022). Instead of assigning a third party’s conduct to the state, the Court considers the state responsible for its act of acquiescence—thus framing complicity as a separate ground for liability under the ECHR. In El-Masri v. Macedonia, the Court held that states may be found responsible for acts committed by foreign officials on their territory if their authorities acquiesced or connived in the wrongful conduct. In Al Nashiri v. Poland the Court reaffirmed that a respondent state could be held responsible for acts of foreign officials conducted on its territory with the knowledge and tolerance of state authorities (para. 517). These cases highlight the ECtHR’s potential to develop a human rights-specific complicity doctrine that aligns with, yet distinguishes itself from, the ILC’s framework. It may be that, should the UK pursue a policy of return hubs or third-country asylum processing, where human and refugee rights abuses occur, individuals could utilise this avenue to establish the UK’s accountability for knowingly acquiescing to these breaches. Given the extensive research and evidence on the human rights implications of externalised border controls, the UK in handing individuals over to such facilities in the territory of another State, can still be responsible for the breaches that occur as a result, even when control over the breach itself is minimal.

5. Concluding Thoughts

This paper has proposed that the responsibility spectrum provides a valuable analytical framework for assessing a state’s responsibility, in this instance, the UK’s efforts to externalise its border control mechanisms. This spectrum demonstrates that attempts to create geographical or legal distance from the site of human rights infringements do not absolve a state of responsibility. Rather, responsibility remains dependent on the degree of control and cooperation exercised by the state. Thus, the responsibility spectrum can be applied broadly to instances of externalised border controls to highlight that legal responsibility cannot be subverted through distance alone.

Utilising the spectrum of cooperation that aligns with the spectrum of responsibility, it is evident that where cooperation is strong, such that the sponsoring state maintains a level of effective control over the externalised border controls, a state could be held directly responsible for breaches of human and refugee rights that occur during the course of migration controls. At the other end of the spectrum, where cooperation is diffuse and minimal, the state may still be held directly responsible for failing to prevent the rights abuses where it was reasonably in a position to do so. This article has sought to promote the middle course, where States purposefully create diffusion and distance through cooperation to avoid responsibility, it has been argued that they can still be responsible for aiding and assisting in the breach.

It may be questioned whether expansive interpretations of existing international obligations are desirable in the migration space, especially given the backlash against the ECrtHR, and whether such an approach could ultimately undermine protection. However, the increasing resort to unaccountable and arbitrary power in the context of migration control undermines the rule of law and contradicts the foundational aims of international human rights law and international refugee law. Courts must continue to pursue accountability for the arbitrary exercise of power, as legal scholars; we are in a position to provide avenues to do this. The disappointing outcome of the SS case illustrates the urgency of highlighting and pursuing the legal implications of aiding and assisting in the context of migration control. It underscores the importance of interrogating the conditions under which support may give rise to derivative responsibility, particularly under principles articulated in the ARSIWA. In this regard, states should proactively seek to mitigate the risk that their genuine support for third countries will contribute to internationally wrongful acts (Moynihan 2019). This necessitates both ex ante due diligence—making inquiries and assessments before providing assistance—and ex post investigations into potential violations (Moorehead 2016).

The prevailing disorder and irregularity observed in the movement of people seeking entry into the United Kingdom is frequently interpreted as a consequence of these individuals’ failure to adhere to formal legal channels. However, for many people seeking international protection or family reunification, irregular entry often constitutes the only viable option. The genuine solution lies not in further securitisation, but in the reform of border controls in a manner consistent with the imperatives of human dignity. Addressing the root causes and long-term dynamics of migration does require collaborative international efforts, but rather than continuing to prioritise securitisation and exclusion, this cooperation must focus on genuine development partnerships, the creation of legal migration pathways, including visa facilitation schemes, and the reconfiguration of migration policy to reflect a just political economy of mobility (Megerisi and Stella Martini 2023).