I. INTRODUCTION
Contemporary manifestations of State power and technological developments in external border management are increasingly being witnessed in relation to the phenomenon of irregular migration by sea in the Mediterranean.Footnote 1 By utilizing strategies like maritime drone surveillance to monitor people on the move, States can focus on detection and interception of migrant vessels while abstaining from search and rescue (SAR) activities.Footnote 2 In this way, European States, acting alone or through European Union (EU) coordination, appear to have found alternative ways to divest themselves from the humanitarian imperative of saving lives at sea, creating a vacuum in human rights protection.Footnote 3 This article will illustrate the increasing integration into migration processes of technologies that transform the way in which the EU, its Member States and third countries manage borders and mobility, and examine the potential harms that such strategies pose to migrants at sea.Footnote 4
Under the international law of the sea, the duty to render assistance at sea is not only a well-established moral but also legal norm under customary international law, now codified in the 1982 United Nations Convention on the Law of the Sea (UNCLOS),Footnote 5 and strengthened by the 1974 International Convention for the Safety of Life at Sea (SOLAS)Footnote 6 and the 1979 International Convention on Maritime Search and Rescue (SAR Convention).Footnote 7 The duty is of critical importance for preserving the safety of life when in danger of being lost at sea and also for safeguarding the right to life under international human rights law.Footnote 8 In complying with this duty, States should refrain from returning migrants to a place where their life or liberty could be at risk.Footnote 9 Accordingly, the prohibition of refoulement under international refugee and human rights law ties into the disembarkation of persons to a ‘place of safety’ in order for a rescue operation to be marked as completed.Footnote 10 However, it is crucial to recognize that the duty to rescue was not intended to act as a policy response to the challenges posed by mass migration movements.Footnote 11 As a result of the lack of a coherent policy and incompliance with the duty, distress situations have become regularized in the Mediterranean where thousands of migrants perish each year.
The year 2023 witnessed unprecedented numbers of migrant tragedies, marked by shipwrecks off Italy (eg off the coast of Crotone province near Cutro) and Greece (eg the Pylos shipwreck) underscoring the ramifications of the extraterritorial practices of the EU and its Member States.Footnote 12 Despite the use of border surveillance in external border management improving the conveyance of critical information and alerting the relevant coastal States about the strong likelihood of life in danger of being lost at sea, SAR responses have not seen a corresponding increase, raising questions about the failure to initiate (timely) rescue missions.Footnote 13
Beyond inactivity in the face of migrant drownings, States have towed migrants outside SAR regions and abandoned them at sea,Footnote 14 or have pushed or pulled migrant boats back to their country of origin.Footnote 15 The latter practice manifests a metamorphosis of what were direct push-backs, as evidenced in the Hirsi Jamaa and others v Italy (Hirsi)Footnote 16 where intervention involved transferring the migrants onto a State navy vessel and delivering them back to their country of origin, into aerial refoulement,Footnote 17 facilitated by the use of surveillance technology. Coastal State authorities increasingly rely on information gathered from aerial assets, such as unmanned aerial vehicles (UAVs, commonly known as drones), to identify the location of a migrant boat; this information is then shared with third-country authorities in order to facilitate their intervention (or, interception) and, ultimately, to deflect boats from reaching the European State's territory.
In this way, European States are no longer at the forefront of migrant interceptions and push-backs, since third countries have undertaken those tasks. This tactic of outsourcing responsibilityFootnote 18 has become prevalent as it allows States to avoid direct physical contact with migrant boats and the associated obligation to rescue together with disembarkation duties that would bring the individuals concerned under the State's de facto and de jure jurisdiction. It is imperative to highlight at this stage that border controls (and distress incidents) most often happen within the international space of the high seas, which is not subject to the sovereignty or jurisdiction of any State.Footnote 19 Nonetheless, the freedom of the high seas does not render this area a space ‘devoid of regulation’,Footnote 20 and human rights must still be respected and realized, though many violations do go unnoticed.Footnote 21 For this reason, the limited response facilitated by European States to blatant violations of fundamental rights has raised uncertainties on the applicability of human rights at sea in practice.Footnote 22
On land, ever more compellingly, sophisticated and enforceable norms safeguard human life, but protection of that same human life and its associated rights at sea remains a jurisdictional and juridical grey area.Footnote 23 The European Court of Human Rights (ECtHR or the Court) has interpreted the concept of jurisdiction along different and sometimes contradictory lines, leading to legal uncertainty and factual mistakes as to when a State's human rights obligations are triggered beyond its national territory. In the ECtHR's established case law on extraterritoriality, jurisdiction beyond a State's terra firma is contingent upon ‘exceptional’ circumstances, where the acts of the State party, performed outside its territory or producing effects there, constitute an extraterritorial exercise of jurisdiction within the meaning of Article 1 of the European Convention on Human Rights (ECHR).Footnote 24 What remains unclear is how jurisdiction can accommodate cases in the context of maritime migration, where ‘effective’ control over an individual's rights does not happen spatially or personally. Instead, State power is exerted through technologies of remote control to prevent (ultimately) access to territory (eg during a border surveillance operation at sea). An enquiry into the theory that encompasses these types of situations is therefore necessary to address securitization practices that impact human rights protection at sea negatively.Footnote 25 Challenging questions will be raised concerning the relationship between human rights and border technologies, thus providing a research agenda for scholars, practitioners and judges alike to build upon.
The structure of this article is as follows. Section II presents the gradual technologization of border control by examining extraterritorial State practices together with the evolving European policy approach in the Mediterranean region. The analysis shows that the transformation of European border controls has resulted in border violence and the preclusion of entry. Against this contextual background, Section III addresses the normative gap in the extraterritorial reach of international human rights by delving into the meaning of jurisdiction and analysing relevant jurisprudence on extraterritoriality. Section IV reconceptualizes jurisdiction in functional terms and sets the stage for a discussion on the ability of technologies to act as knowledge generators that could trigger the jurisdictional link between a duty-bearing State and the potential right-holding individual. The conceptualization of jurisdiction posited herein does not consider direct physical contact with the concerned individuals as a conditio sine qua non for the application of human rights law.Footnote 26 On the contrary, it proposes a paradigm that addresses instances of State power that, if and when exercised, can constitute effective control, triggering a State's human rights obligations.
II. MOBILITY AND THE TECHNOLOGIZATION OF BORDERS
By definition, migration is a source of human mobility that is best described as a geographical phenomenon conceptualized by the movement of people across State borders and spaces.Footnote 27 In Europe, third-country nationals who do not have the appropriate visa documents to secure legal entry are often forced to seek unauthorized access via dangerous migration routes—the most well known being the sea. Migration policies to control or set entry requirements in a State or group of States, like the EU, equally have a geographical dimension.Footnote 28 The EU per se does not carry out border controls, but these are undertaken by European States individually or coordinated by the EU's Border and Coast Guard Agency (Frontex), usually in the form of joint naval operations or under bilateral cooperation agreements between States (eg the Italy–Libya Memorandum of UnderstandingFootnote 29).
Thermal imaging cameras, night-vision goggles, ground sensors, aerial video surveillance, maritime simulations, UAVs and even experimental robotic technology are some of the technologies employed to monitor and control movement before people reach Europe's physical borders.Footnote 30 Whilst scholars like Mattelart have rightly pointed out that the use of surveillance shifts control ‘from visible to invisible’, and that it is this invisibility that engenders the efficiency and uniqueness of such technologies of control,Footnote 31 it is argued here, and by others,Footnote 32 that the use of such surveillance results in an invisible phenomenon—migration by sea—becoming visible, or, ‘knowable’ and hence ‘governable’. It is against this background that sophisticated technologies play an instrumental role within border management as they generate spatial knowledge and, therefore, afford significant power to State authorities to control remotely the passage and entry of irregular migrants.Footnote 33 This is important to comprehend at the outset, as the increased visibility of migration routes and the timely knowledge of potential distress situations affords States the opportunity to affect an outcome or course of events. Such exercise of power is capable not only of triggering positive obligations to render assistance at sea under the law of the sea framework but also extraterritorial human rights obligations, as will be argued in Section IV.
A. ‘Smart Borders’ versus Lives at Sea
Underpinned by the aim of building ‘smart borders’ to tackle migration, the EU established, inter alia, the European Border Surveillance System (EUROSUR) in 2013, a programme operated by Frontex that uses big data technologies (including satellite imagery and ship recording services) ‘to predict, control and monitor traffic across European Union borders’ and ultimately to block the passage of migrants.Footnote 34 Notably, one of the aims of EUROSUR was to ‘contribute to ensuring the protection and saving the lives of migrants’.Footnote 35 Although the number of migrant crossings has fallen significantly since 2016 as a result of the EU's securitized approach to border governance—which the EU heralds—this decrease has been accompanied by a dramatic increase in the mortality rates in the Mediterranean, which is particularly unsettling given that the area is heavily surveilled.Footnote 36 The International Organization for Migration (IOM) recorded more than 3,041 missing migrants in the Mediterranean region in 2023 compared to 2,411 in 2022. This translates to an 11 per cent increase in mortality rates in 2023, while in 2022 there was an 18 per cent increase from 2021.Footnote 37
These migrant drownings have been described in the literature as ‘border deaths’ as it is apparent that they result from border violence and are the consequence of structural and political conditions rather than the result of individual acts.Footnote 38 When States become aware of persons in distress by virtue of data transmitted from surveillance technologies, such as thermal images indicating an emergency, the international law of the sea mandates immediate assistance to those in danger of being lost at sea. However, as Goodwin-Gill notes, quoting the International Tribunal for the Law of the Sea (ITLOS), this obligation is ‘not a counsel of perfection … but rather “an obligation to deploy adequate means, to exercise best possible efforts, to do the utmost, to obtain this result”’.Footnote 39 Put simply, in legal terms, it is an obligation of conduct and not of result.
It could legitimately be expected that increased situational awareness of sea migration routes would enhance SAR responses by providing early warnings of distress and sustained vigilance over the danger to life at sea. However, the spiralling loss of life at sea along with the abuse of migrants’ human rights when being unlawfully intercepted and returned to their countries of origin have exposed the systematic obstruction of this humanitarian duty, potentially jeopardizing multiple rights, including the right to life, the prohibition of torture and the right to an effective remedy. Determining the exact number of deaths resulting from State authorities neglecting their duty to rescue is challenging, but information from various sources, including investigative journalists and academics, provides an indication of the human toll.Footnote 40 An earlier study conducted along the United States–Mexico (land) border revealed that border control policies employing new surveillance technologies doubled migrant deaths and redirected migration routes towards more perilous paths.Footnote 41 An assumption was made that Europe's ‘watery graves’ were similarly the result of the increased use of technology to facilitate interceptions and returns.Footnote 42 The expansion of State sovereign power and amplification of the State's discretion in interpreting legal obligations is being both facilitated and exercised through technological tools at the expense of migrants’ rights and lives. According to Mitsilegas such externalization measures create a preventive (in)justice paradigm, as the legal and policy frameworks aim to control and prevent migratory movements.Footnote 43
At a time of substantial EU spending on research and development for security technology,Footnote 44 the European project ‘NESTOR’Footnote 45 provides one of the most recent examples epitomizing the technologization of the border and practices of exclusion.Footnote 46 This ‘next-generation’ border surveillance system aimed to provide pre-frontier situational awareness beyond maritime and land borders by employing mixed reality glasses, long-range electro-optical sensors, 360° cameras, three-dimensional radar, an unmanned underwater vehicle and a platform that integrated data from radio frequency analysers. The primary aim of NESTOR was the detection of irregular migration and human trafficking through strengthening the long-range surveillance capabilities of States. The lines between migration management on the one hand, and the fight against transborder crime on the other, are blurred through effectively equating migrants with criminals. This process of ‘othering’ fuels destructive attitudes and allows for stricter security measures to address the supposed ‘threat’.Footnote 47 Conceptualized in this manner, the border becomes a key instrument of migration management, with technology aggravating the practice of refoulement and allowing States seemingly to sidestep any associated obligation of rescue by shifting the responsibility to third countries.
B. European Practices in the Mediterranean
There has been an ever-growing technologization of border controls in the Mediterranean.Footnote 48 In 2015 as part of its response to the ‘migration crisis’, the EU instigated its Common Security and Defence Policy military operation in the Mediterranean, EUNAVFOR MED Operation Sophia (Operation Sophia).Footnote 49 This operation marked a clear operational shift in the EU as the aim of the mission was to prevent departures, disrupt smuggling networks and trafficking operations and, ultimately, ‘to better contain the growing flows of illegal migration’.Footnote 50 In 2016 the Council reinforced Operation Sophia's mandate and added two supporting tasks: the training of the Libyan Coast Guard and Navy (LCGN), and the use of aerial, satellite and maritime assets to contribute to the enforcement of the United Nations (UN) arms embargo on the high seas off the coast of Libya.Footnote 51 In 2020 EUNAVFOR MED Operation Irini (Operation Irini) superseded Operation Sophia, which had fulfilled mainly training of the LCGN and surveillance roles in the Central Mediterranean due to a lack of naval assets to conduct physical inspections.Footnote 52 Operation Irini's mandate is focused on providing direct engagement in support of the identification and interdiction of arms transfers, representing a shift from Sophia's anti-migration focus. However, its remit retains the controversial capacity building and training of the LCGN in law enforcement tasks at sea and has an expanded scope that allows for the use of aerial surveillance within Libyan airspace.Footnote 53
Despite the gradual shift in immediate focus, the long-term goal behind these operational moves was to buttress Libya's capacity to stop migrants from leaving its territory —‘for pre-emptive take-backs to replace SAR, shifting the responsibility for refugee and migrant flows to Libya’.Footnote 54 The EU's external border policy is geared towards the reinforcement of third countries’ capacity to guard their borders and intercept migrant boats, distracting attention from the more pressing challenges presented by irregular migration and the widening gaps in refugee and human rights protection.Footnote 55 Given the extensively documented deterioration of the human rights situation in Libya and the widely documented abuses suffered by migrants under the control of the LCGN,Footnote 56 preventing migrants from departing from Libyan territory or returning individuals to Libya can amount to complicity with arbitrary detention, torture and even violations of the right to life.Footnote 57
In addition to Operation Irini, the EU has funded other external border management initiatives, including, inter alia, Frontex's ‘Joint Operation Poseidon’ in the Eastern Mediterranean which carries out border surveillance mostly along the Greek sea borders with Turkey,Footnote 58 Frontex's ‘Joint Operation Indalo’ in the Western Mediterranean, that carries out border surveillance and SAR primarily from North-West Africa to Spain,Footnote 59 and Frontex's ‘Operation Themis’ which replaced the ‘Joint Operation Triton’ in February 2018 and has as its primary mandate border control and surveillance supporting Italy in the Central Mediterranean region.Footnote 60 Operation Themis represents an extension of the geographic coverage of these missions, extending surveillance activities to the waters of Algeria, Tunisia, Libya and Egypt as key parts of the migration routes leading to the Mediterranean.Footnote 61 These missions have one operational aspect in common: the use of air-naval assets to collect intelligence in the Mediterranean region.
Technology plays a leading role in Frontex's capacity to geolocate migrants near to borders and take strategic decisions about who to alert.Footnote 62 Rather than launching a SAR response, Frontex has principally conducted aerial sightings of distress incidents via chartered aircrafts operated by private companies through its multipurpose aerial surveillance scheme in the Central Mediterranean. These aircraft transmit video footage and other information to a situation centre in Frontex headquarters in Warsaw,Footnote 63 which then makes operational decisions about who to alert, communicating information to various actors, including the LCGN. In so doing, it does not attempt to ensure that pull-backs, which would result in increased harm to the lives of migrants, will not occur.Footnote 64 This is particularly alarming given that Frontex has been repeatedly scrutinized and accused of being complicit in push-back operations by a number of EU bodies, human rights organizations and UN organs, and has faced legal action against its practices and claims that it has overstepped the limits of its powers.Footnote 65 The European Ombudsman's decision on the fundamental rights obligations of Frontex with regard to SAR in the context of its maritime surveillance activities has highlighted significant shortcomings in how Frontex handles maritime incidents and showed a failure to ensure fundamental rights monitoring in its decisions.Footnote 66
The analysis of available data has demonstrated a significant correlation between Frontex surveillance flights and the number of interceptions performed by the LCGN, pointing to the conclusion that on days when the aerial assets fly more hours over their dedicated area of operation, the LCGN tends to intercept more boats.Footnote 67 These interventions have escalated in recent years, indicating a departure from established international obligations in this space. In 2022 alone the LCGN intercepted 24,684 people at sea, whilst that year marked the highest death toll recorded since 2016, with 6,876 documented deaths.Footnote 68 It is clear that the reliance on aerial surveillance by Member States operating under Frontex jointly with the LCGN indicates a progressive abstention from SAR activities, but also an intensification of systematic remote interceptions and returns of migrants to Libya. Frontex's ‘airborne complicity’ has resulted in systematic refoulement operations that could be regarded as crimes against humanity.Footnote 69
The use of security concerns as a justification to prevent the entry of migrants has created a bulwark to accessing international protection.Footnote 70 As FitzGerald notes, States’ adherence to the formal legal principle of non-refoulement encourages them to create policies, like the interception of migrant vessels at sea and border externalization, that, in practice, undermine the essence of the international refugee regime by increasing the risk of refoulement.Footnote 71 Considering that more than 70 per cent of the EU's external borders are maritime, the surveillance monitoring of wider areas has created a new generation of human rights violations where an individual's rights become conditional on their geographical location. However, to date there has been no real judicial test of the idea of refoulement described above,Footnote 72 or of the aerial cooperative actions of States witnessed as part of joint border operations at sea.
At this juncture, it needs be underscored that Frontex's operations at sea are, in principle, required to abide by EU law, including the Charter of Fundamental Rights of the European Union (EU Charter) and relevant international law, including the provisions of UNCLOS.Footnote 73 The EU Charter, in particular, pertains to all actions taken by the EU institutions, bodies and agencies as well as its Member States whenever EU law is being implemented, whether territorially or extraterritorially.Footnote 74 There is thus no jurisdictional threshold requirement for the applicability of EU human rights law, as unlike the ECHR, the EU Charter does not contain a jurisdictional clause, and it can offer more extensive protection than the ECHR under Article 52(3).Footnote 75 Therefore, both the EU institutions and its Member States are bound by the EU Charter irrespective of the potential impact of their decisions,Footnote 76 or the location of the affected individuals in question. As Ryngaert notes, when the EU exercises its powers, ‘it owes human rights obligations to persons affected by such exercise of power, irrespective of the location of those persons’.Footnote 77 However, it is important to stress the challenge of providing definitive answers regarding the extraterritorial reach of the EU Charter, which remains an area surrounded by ambiguity.Footnote 78
III. JURISDICTION: A CONDITIO SINE QUA NON FOR SAFEGUARDING RIGHTS AT SEA
The ongoing situation in the Mediterranean accentuates the importance of clarifying extraterritorial human rights obligations during maritime crossings, as the risk to life is particularly amplified during migration by sea. The situational and structural vulnerability of migrants at sea relates not only to navigational hazards and migrant smuggling or human trafficking but extends also to exclusionary border control practices,Footnote 79 or ‘non-entrée’ policies,Footnote 80 applied from the moment an individual attempts to leave their country. The analysis below reflects on the notion of jurisdiction in human rights law, exploring its meaning and scope qua threshold criterion for safeguarding human rights. In turn, it explains its association, if any, with related yet dissimilar notions of jurisdiction in international law, before showcasing how the ECtHR has interpreted the notion of jurisdiction, both in general settings and in the maritime space. The quest for a unified understanding of extraterritorial human rights jurisdiction is an important one. Clarifying this normative issue allows for an examination of whether the use of border technologies in migration processes can serve as a trigger to jurisdiction for the application of human rights law.
A. Understanding Human Rights Jurisdiction
Apart from the positive obligations of assistance found in the international law of the sea framework, individual rights are protected under international human rights law, which States should guarantee for everyone under their jurisdiction.Footnote 81 When looking at the territorial scope of the application of a treaty under international law, Article 29 of the 1969 Vienna Convention on the Law of the Treaties (VCLT) provides that ‘unless a different intention appears from the treaty or is otherwise established, a treaty is binding upon each party in respect of its entire territory’.Footnote 82 As has been highlighted by the above discussion, border control practices, interception measures and distress incidents (SAR operations) most often occur beyond the territorial sea (which is considered as part of a State's territory and extends up to twelve nautical miles) and, in particular, take place in the Exclusive Economic Zone (EEZ), which for SAR purposes, is treated as if it is the high seas.Footnote 83 States ordinarily do not enjoy sovereign control on the high seas and thus individuals enjoy no protection of their rights, in the absence of a normative relationship with a duty-bearing State. The VCLT does not address the application of treaties beyond the territory of a State party. In addition, the main international human rights treaties on civil and political rights, the International Covenant on Civil and Political Rights (ICCPR), the American Convention on Human Rights and the ECHR and their Protocols, conceive State responsibility for securing the rights they contain only in terms of the State's ‘jurisdiction’. For example, the ECHR provides in Article 1 that its State parties ‘shall secure to everyone within their jurisdiction’ the rights and freedoms provided in the Convention. It is thus necessary to establish extraterritorial human rights jurisdiction in order to apply the relevant human rights treaties for the protection of migrants at sea.
It is well established that jurisdiction in international human rights law is fundamentally different from jurisdiction as it is primarily understood in public international law.Footnote 84 Jurisdiction in international human rights law is understood as responsibility giving rise to specific human rights obligations,Footnote 85 whereas, as Milanović puts it, jurisdiction in public international law concerns a State's right to regulate its own public order;Footnote 86 hence, it is concerned with whether or not a State's activity constitutes a lawful exercise of jurisdiction.Footnote 87 Accordingly, in the international human rights domain, the exercise of power by the State does not have to be necessarily within its legal competence,Footnote 88 but needs only ‘to flow, by definition, from a lawfully organized institutional and constitutional framework through which these state agents exercise some kind of normative power with a claim to legitimacy, even if that claim might prove to be unjustified’.Footnote 89
Additionally, jurisdiction in human rights law should also be demarcated from the law on State responsibility, especially the principles surrounding attribution of conduct. In the human rights field, jurisdiction is tied to the emergence of an obligation of the State to secure (or take responsibility for) the rights of individuals.Footnote 90 Conversely, with regard to the law of State responsibility, ‘[a]ttribution of certain acts to a public institution and State agents, and, more generally, responsibility, only come later once the State's duties have arisen in the first place and have been violated’.Footnote 91 Thus, in relation to this discussion, the focal ‘responsibility’ of the State is not the attribution of legal responsibility (after a breach has taken place), but the need to ensure compliance with human rights obligations beforehand.
B. The Extraterritorial Understanding of Jurisdiction: General Application
The ECtHR in its jurisprudence on extraterritoriality has so far employed two types of jurisdiction in order to determine the applicability of the ECHR outside a State's own terra firma: (1) jurisdiction defined spatially, the ‘spatial model’, also known as ‘control over an area’; and (2) jurisdiction defined personally, the ‘personal model’, also known as ‘State-agent authority or control’ or ‘power over an individual’. The first landmark case of the ECtHR examining the extraterritorial application of human rights treaties was Loizidou v Turkey, in which the Court developed the concept of spatial jurisdiction.Footnote 92 The Court emphasized that a State's responsibility might arise ‘when as a consequence of military action—whether lawful or unlawful—it exercised effective control of an area outside its national territory’.Footnote 93 However, this conception of jurisdiction cannot cover the many different ways and situations in which States might violate human rights abroad without exercising effective control over an area.
Subsequently, in interpreting jurisdiction in the Banković case, the ECtHR ruled that the people killed in the airstrike were not within the jurisdiction of the North Atlantic Treaty Organization (NATO) States, explaining that airstrikes over a territory do not establish control over the actual area. Disappointingly, it held that jurisdiction is primarily territorial and did not agree with the applicants’ submission that ‘anyone adversely affected by an act imputable to a Contracting State, wherever in the world that act may have been committed or its consequences felt, is thereby brought within the jurisdiction of that State’, finding that the text of Article 1 did not support such an approach to jurisdiction.Footnote 94
Since then, there have been several remarkable developments attempting to alleviate the stringency of the Banković rule. In Ilaşcu and others v Moldova and Russia, the ECtHR held that jurisdiction may persist even in parts of the State's own territory that are not under its effective control.Footnote 95 The State still has a positive obligation under Article 1 of the ECHR to take all appropriate measures within its power, in accordance with international law, to secure to the applicants the rights guaranteed by the Convention, even in the absence of effective control.Footnote 96 Importantly, in the ground-breaking case of Al-Skeini v UK,Footnote 97 the ECtHR, by affirming the ‘personal model’ of jurisdiction, stressed that ‘whenever the State through its agents exercises control and authority over an individual and thus jurisdiction’ the State is obligated to secure that individual the rights relevant to the situation flowing from the ECHR.Footnote 98 It further noted that the decisive element is ‘the exercise of physical power or control over the person in question’.Footnote 99 The reasoning highlights the conceptual ambiguities surrounding the use of ‘power’ and ‘control’ with regard to establishing jurisdiction. The ECtHR either intended to equate the ‘exercise of physical power’ with ‘control’, or it meant that power and control are alternatives in this context.Footnote 100 In his compelling concurring opinion, Judge Bonello argued that the failure of judicial institutions to establish a coherent and axiomatic regime of jurisdiction signifies a failure to recognize the universal status of personhood, which is the essence of human rights.Footnote 101
More recent cases seem to be indicating a return to the infamous narrow stance in Banković in interpreting the notion of jurisdiction, as scholars like Milanović have noted in the context of armed conflict cases that have arisen.Footnote 102 Despite debates on the extent of ECHR protection, these discussions do not easily translate to the migration context, especially at sea, where the principles are applied differently. This is because the widely acknowledged two-model approach, commonly recognized as the basis of extraterritorial jurisdiction, does not seem to apply to factual circumstances where a State exercises remote control over persons in distress at sea by way of information (and visual knowledge) acquired from technological devices used during border surveillance operations at sea. Neither the spatial model nor the personal model appears to fit appropriately in this context. The scope and content of pre-existing international human rights law norms are thus faced with the need for reformulation and adaptation to address effectively the intricacies of not only maritime migration but also the integration of technology in border management.Footnote 103
C. Jurisdiction in the Maritime Space
Human rights bodies, alongside law-of-the-sea specialized bodies and tribunals,Footnote 104 have recognized that human rights extend and apply at sea in an analogous manner to on land, as long as control is exercised by State parties’ authorities.Footnote 105 Accordingly, there is ample scholarship on the application of human rights treaties at sea,Footnote 106 pointing out that the law of the sea pursues the protection of human rights indirectly.
Xhavara is the only case heard by the ECtHR thus far invoking a violation of the right to life under Article 2 of the ECHR in the context of extraterritorial migration policies, and concerned an incident which resulted in the drowning of fifty-eight migrants.Footnote 107 The Italian authorities had intercepted a boat carrying Albanian nationals in an attempt to prevent the migrants from entering Italy, pursuant to an existing bilateral agreement between Italy and Albania.Footnote 108 Importantly, the ECtHR found that the incident was, in fact, caused by the Italian military vessel in the course of carrying out the activities under their agreement with Albania and thus, based on the factual circumstances of the collision, the victims were within Italy's jurisdiction.Footnote 109 While the case was inadmissible because the applicants did not exhaust all domestic remedies,Footnote 110 important insights can be gained from the ECtHR's reasoning in determining that there was jurisdiction. It implicitly confirmed that migration controls through the construction of bilateral agreements could give rise to the exercise of jurisdiction as they confer enforcement powers on a State.Footnote 111 Applying this reasoning of the ECtHR by analogy, it is apposite to posit that technological advancements in border activities transform the way that States exercise control over persons at sea and thus may be capable of giving rise to jurisdiction. Thus, even though territorial jurisdiction is the predominant source of jurisdiction, the development of a technologized border evokes a different dynamic that is capable of altering the traditional framework of international human rights law and the rules on jurisdiction.
The ECtHR addressed the specific issue of jurisdiction at sea in Medvedyev and others v France Footnote 112 in the following way:
[A]s this was a case of France having exercised full and exclusive control over the Winner and its crew, at least de facto, from the time of its interception, in a continuous and uninterrupted manner until they were tried in France, the applicants were effectively within France's jurisdiction for the purposes of article 1 of the Convention.Footnote 113
This approach was confirmed in the renowned Hirsi case, when the ECtHR determined that a State exercises jurisdiction on intercepting vessels on the high seas, when migrants are ‘under the continuous and exclusive de jure and de facto control’ of that State.Footnote 114 This is the most important ECtHR case to date in relation to extraterritorial human rights obligations in rescue operations as the ECtHR confirmed that jurisdiction extends across borders and applies on the high seas where a State's border control operations threaten the rights of migrants.Footnote 115 Notably, the ECtHR has recently affirmed this ruling in a case involving similar circumstances, M.A. and Z.R. v Cyprus,Footnote 116 and did not accept that the bilateral agreement between Cyprus and Lebanon of readmission ensures protection against human rights violations. The Court found Cyprus to have performed maritime push-backs in a similar manner to Hirsi, and reiterated that States ‘cannot evade their own responsibility by relying on obligations arising out of bilateral agreements with other countries, in this case Lebanon’.Footnote 117 It can be concluded that border control policies are primarily an expression of State authority and all forms of this control should result in the exercise of the State's jurisdiction.Footnote 118 Therefore, it can be posited that technologies fall in the context of such operations, as they can allow States to abstain from direct physical contact between the authorities and individuals in distress, who they are nonetheless in a position to impact remotely through their decisions.
With the exception of these few cases, the jurisprudence of the ECtHR on migration increasingly shows its ambivalence towards extending the notion of jurisdiction beyond borders, leading to more restrictive outcomes and leaving significant gaps in human rights protection. In the face of this apparent lack of political will or consensus, it will require a bold judicial intervention to develop the normative contours of jurisdiction to safeguard the rights of individuals on the high seas.
IV. RECONCEPTUALIZING HUMAN RIGHTS JURISDICTION
A. A Functional Understanding
In the absence of a specialized refugee treaty body, or a migration court, or other supranational supervisory mechanism dealing with these matters, adjudication in this field is primarily national in conjunction with international human rights bodies. Ultimately, the responsibility for administrating justice for gross human rights violations at sea that may go unnoticed or unpunished falls on human rights bodies.Footnote 119 As a result, the ECtHR, as a court of last resort, has an instrumental role to play in condemning extraterritorial practices that violate human rights but also in guiding the EU and its Member States on migration policies and ensuring that an ‘à la carte’ respect for human rights is no longer entertained.Footnote 120
In addition to the ECtHR, the UN Human Rights Committee (HRC) has also dealt with an abundance of cases concerning the extraterritorial application of human rights. In general terms, the HRC has been seen to take a pro homine, universalistic approach to the construction of extraterritorial jurisdiction that has led to broader legal constructions, whereas the ECtHR has been seen to adopt a more sovereignty-oriented approach that has often led to more restrictive outcomes.Footnote 121 In the context of migration cases, the ECtHR is frequently seen as taking a deferential attitude to States in upholding their sovereign right to control migration, resulting in significant gaps in human rights protection.Footnote 122
In contrast to the ECtHR's approach to extraterritoriality, the HRC has interpreted jurisdiction in functional terms and used ‘impact’ as a ground for applying the ICCPR when a State exercises power in extraterritorial settings over ‘persons located outside any territory effectively controlled by the State, whose [rights are] nonetheless impacted by its military or other activities’.Footnote 123 To this end, if a State's act produces effects outside its territory that impact the enjoyment of the rights of the concerned individuals, it could amount to an exercise of jurisdiction.Footnote 124 The key aspect of jurisdiction, as Shany argues, is not about the actual act or omission, but ‘about states having the potential (or functional capacity) to comply with or violate IHRL obligations’.Footnote 125
Consequently, having the potential to exercise effective control in a situation arguably carries the same degree of power as actual placement within a State's control. This justification finds its origins in Aristotle's distinction between the potentiality and actuality of positive obligations.Footnote 126 The former translates into power (‘dynamis’; ‘δύναμη’), while the latter translates into activity (‘energeia’; ‘ενέργεια’).Footnote 127 On the one hand, potentiality is the power to affect an outcome; on the other hand, actuality is the realm of events and facts. Ontologically, therefore, the two are not to be equated but carry the same degree of power, that is, a State has the capacity (whether it acts on it or not) to impact the relevant rights of individuals in a direct and reasonably foreseeable manner by its decisions. Hence, in line with Raible's argumentation, the power to do something is potential, and if and when it is exercised, results in control.Footnote 128
To support the functional model of jurisdiction, an analogy can be drawn from a land-rescue case, Furdík v Slovakia,Footnote 129 concerning the delayed rescue of a mountain climber who died whilst waiting for help to arrive.Footnote 130 The ECtHR grasped the opportunity to clarify that the obligation of States to protect the right to life under the ECHR extends to emergency services ‘where it has been brought to the notice of the authorities’ that the life or health of an individual is at risk.Footnote 131 This reasoning could equally be extended to the case of migrants in distress at sea, as visuality or real-time monitoring generated by technology is capable of drawing the attention of the State authorities to a possible emergency, thereby inducing a situation of power.
Considering border technology in the reconceptualization of jurisdiction, it is posited that pre-frontier detection of maritime movements empowers States to control remotely how a situation of distress at sea will unfold.Footnote 132 Whilst scholars like Jumbert have argued that surveillance does not equal actual control, it nonetheless provides States with the choice whether to exercise sovereign power at the border.Footnote 133 From this perspective, border technologies could be conceptualized as knowledge generators that have the ability to activate the jurisdictional nexus vis-à-vis a State's human rights obligations to exercise ex ante due diligence through its conduct in managing migration. As a result, a coastal State may need to refrain from adopting measures that could disrupt migration movements—such as intensified surveillance—especially if such actions risk impeding the efficacy of rescue operations or even increasing the likeliness of distress incidents, by encouraging migrants to take more dangerous, less surveilled routes.Footnote 134 The continuing large-scale loss of life at sea should prompt States to review the applicable regulatory system, as Lisa-Marie Komp suggests, ‘with an eye to the question whether it actually prevents the loss of life in practice’.Footnote 135
At a high level of generality, the complexities raised above can be disentangled by answering the following principal question: at what point in time do persons in distress at sea become subject to a State's jurisdiction? It is submitted that the jurisdictional link is activated at the point when the notified-State has the functional capacity (power) to impact through its decisions (acts and/or omissions, including instructions to a third party) the relevant rights of the concerned individuals in a direct and reasonably foreseeable manner.Footnote 136 What makes control over the rights of individuals at sea ‘effective’ is the State's ‘capacity to determine the material course of events’ and compliance with the duty of protection.Footnote 137 A causal link is thus created, triggering a State's positive human rights obligations to exercise due diligence in coordinating rescue efforts, in order to stimulate conduct that is in line with the SAR system and the international human rights framework. The justification for this interpretation is clear: it is only by requiring action at a stage when it can still have effect that the right to life is adequately safeguarded. Thereafter, it is clear that both flag and coastal States are obliged to adhere to the principle of non-refoulement and ensure that those rescued are taken to a place that qualifies as safe.Footnote 138 This functional understanding aligns with a teleological interpretation of jurisdiction,Footnote 139 which becomes pertinent for addressing the evolving patterns in migration control and external border management.
In the below analysis, the functional understanding of jurisdiction will be incorporated into the practices most commonly seen in the Mediterranean region: (1) delayed/non-assistance; (2) push-backs by a proxy third actor; and (3) privatized refoulement in connection with aerial refoulement.
B. Practical Application of the Functional Model
1. Delayed assistance and non-assistance
The functional view espoused in this article is largely influenced by the much-discussed case of A.S. and others v Italy, concerning a migrant shipwreck which resulted in an estimated 200 deaths.Footnote 140 In this case, the Italian and Maltese authorities did not cooperate effectively in agreeing who should undertake the rescue, and, as a result, help arrived seven hours after the first launch of the distress call. The issue before the HRC was not whether the shipwreck occurred within the State party's territory, as it clearly did not, but rather whether the alleged violations of the right to life could be considered to have been within the power or effective control of the State, despite taking place outside its territory. The HRC found that Italy exercised effective control over the individuals in distress as they were ‘directly affected by the decisions taken by the Italian authorities in a manner that was reasonably foreseeable’Footnote 141 even though they were also concurrently subject to the jurisdiction of Malta who had formally assumed primary responsibility to coordinate the rescue operation as it was within its SAR zone.Footnote 142
Consequently, the HRC effectively broadened the concept of control, since it was not required to be over the person, but rather over the enjoyment of human rights. This facilitates attribution of certain types of rights violations where the element of physical control is not present. An equally important development in this context is the ECtHR judgment in Safi and others v Greece, where the Court emphasized the considerable delays in rescue assistance which have become the norm in the Mediterranean. In this case rescue arrived over an hour after the boat had completely sunk. The judgment made it clear that time is of the essence in cases where drowning can be reasonably anticipated.Footnote 143 It is thus clear that the coastguard, as a de jure State organ,Footnote 144 has a pivotal role to play in exercising State power.
Such legal and factual complexities were exacerbated in the Cutro shipwreck of 26 February 2023, which provides a paradigmatic example of the kind of policy control that portrays a functional understanding of jurisdiction as described in this article. In brief, while no distress call was placed from the migrants’ boat to alert the Italian authorities of their need of assistance, the use of surveillance technology by Frontex alerted the relevant authorities of a strong likelihood of a distress situation that arguably should have been marked as a SAR event.Footnote 145 The Italian Maritime Rescue Coordination Centre had access to the live-streaming sensors that were shared by Frontex by virtue of the boat being in its SAR region, but still did not classify the incident as an ‘emergency’, and thus did not launch a SAR operation.Footnote 146 It therefore seems that the primary cause of this human tragedy was Italy's negligent failure to launch a SAR mission within its SAR zone, resulting in fatal consequences.
The law of the sea provides that coastal States have authority over distress incidents in their SAR zone and there is an obligation of due diligence whereby they must exercise best efforts to activate the available SAR services in that geographical area and employ all adequate measures to save lives.Footnote 147 It appears that the creation of a technological surveillance infrastructure is capable of serving as an accountability avoidance tool, allowing coastal States an interpretative discretion as to whether a situation amounts to ‘distress’ that would oblige them to ensure that the necessary assistance is provided. Arguably, unduly delaying or negligently handling a rescue operation, or not responding to information indicating a distress situation from coastal States concerning vessels within their SAR zones, amounts to an exercise of human rights jurisdiction as the authorities are acting in the knowledge (actual or putative) that the lives of persons are at risk.
2. Push-backs by a proxy third actor
In a joint third-party application submitted to the ECtHR in the pending S.S. and others v Italy case, it was argued, inter alia, that the jurisdictional link is activated ‘whenever a State's authority, due to its powers under international law—and on the high seas, under the international law of the sea—issues instructions to a third actor with extraterritorial effect’.Footnote 148 The incident unfolded in 2017 and concerned an interception by the LCGN, acting under Italian orders, of a vessel carrying around 150 migrants which began to capsize on the high seas soon after departing from Tripoli. It should be noted that the non-governmental organization (NGO) Sea Watch 3 (SW3) also arrived at the scene and assumed on-scene coordination, as the LCGN initially objected to taking that role (and did not have the necessary equipment to carry out the rescue). Nevertheless, the Italian authorities insisted the LCGN should take charge of the rescue. The NGO rescued some individuals, some were ‘pulled-back’ by the LCGN to a camp in Libya where they reportedly faced abuse, and others, including a child, drowned amidst the unravelling of the chaotic situation.
The case is a stark representation of the countless incidents of Italy assigning rescue duties to the LCGN to try to avoid any direct physical contact that would bring the persons in distress under its jurisdiction. Reflecting on the case, Papastavridis has expressed the view that the knowledge of the distress incident alongside control over the persons in the case are decisive for the ECtHR to establish its jurisdiction.Footnote 149 Indeed, the State's influence over the perpetrator was highlighted by the Court in the Ilaşcu case, where it concluded that Russia bore responsibility for the abuses committed by Moldovan separatists in Moldova. This conclusion was based on Russia's ‘decisive influence’ over the separatists.Footnote 150 However, the Court's reasoning in this instance is rather ambiguous as it does not clearly explain the significance of Russian influence—whether it serves as a basis for attribution of responsibility or triggers the obligation to protect.Footnote 151 Therefore, in considering this case, the ECtHR has a real opportunity to consolidate its jurisprudence on the extraterritorial application of human rights at sea and in particular, when a State issues instructions to a third actor which have an extraterritorial effect. However, it might reject a functional understanding of extraterritorial human rights jurisdiction, given that it rejected the ‘capacity’ or ‘impact’ approach argued by the applicants in Banković .Footnote 152 It went on to say that the mere capacity to interfere with the enjoyment of a right will not suffice to entail jurisdiction over the right-holder.Footnote 153 However, Banković was a different scenario from the present context of external border and migration management, albeit also featuring competing policy considerations.
3. Privatized refoulement in connection with aerial refoulement
The practice of interdiction and refoulement is also increasingly carried out by private merchant vessels acting on information obtained by States through aerial intelligence. The Nivin incident in 2018 is an example of such a situation, which led to the submission of an individual complaint, SDG v Italy, currently pending before the HRC.Footnote 154 In this incident, upon rescuing migrants in the Mediterranean, the commercial vessel the Nivin was instructed by the Italian authorities to return them to Libya. The LCGN boarded the vessel and allegedly injured a number of the rescued migrants and returned them all to detention centres.Footnote 155
The complaint relies on the impact model mentioned above in Section IV.A, arguing that in their dealings with the Nivin, the Italian authorities through their coordination with and on behalf of the LCGN had impacted the right to life of the individuals involved, in a direct and reasonably foreseeable manner. Of particular relevance is the fact that the Italian authorities acquired knowledge about the migrant boat in distress by virtue of the data transmitted by a Spanish surveillance aircraft operating as part of Operation Sophia.Footnote 156 Looking at the ECtHR's jurisprudence, the Court has recognized a ‘reasonable knowledge’ condition which provides that preventive positive obligations arise if the State's authorities knew or should have known of a real and immediate risk to the life of an identified individual from the criminal acts of a third party.Footnote 157
In this vein, the knowledge criterion rests on a factual assessment of the circumstances which will be triggered by a certain risk that is foreseeable for States. In the context of maritime migration and external border management, it can be analogously argued that the knowledge criterion will be satisfied when States receive information about a distress incident that alerts them of a probable resultFootnote 158—the foreseeable risk to the life of migrants who are in distress—and requires immediate action.
4. Concluding remarks on the functional model
Clearly, technology has been very effective and has played a key role in preventing migrants from reaching Europe and curtailing the right to seek asylum. Recognizing the significance of this development, the Special Rapporteur of the Human Rights Council on extrajudicial, summary or arbitrary executions has underlined the responsibilities of States exercising such surveillance, noting that since the:
European Union and its member States have put in place an extensive surveillance system focused on security and border patrol [and have] chosen to provide security in the Mediterranean, the States members of the European Union … … are exercising sufficient functional control to be subject to the one obligation inextricably linked to ocean surveillance: an adequate and effective system of rescue. This includes the implementation of the principle of non-refoulement, including to unsafe third countries, the protection of refugees and migrants, including against preventable and foreseeable loss of lives and support to ships operated by non-governmental organizations.Footnote 159
The EU and its Member States have progressively redefined their border policies in a manner that not only shows nearly complete disengagement with rescue obligations but also involves delegating SAR activities and border management to Libya, a volatile State where human rights abuses are a concern. While Libya now has its own SAR zone, numerous reports from NGOs and individuals strongly argue for a re-evaluation of the Libyan SAR zone classification until the LCGN demonstrates that it will conduct SAR operations in line with international obligations.Footnote 160 As noted above, rescue operations must end in a safe harbour or a ‘place of safety’, and Libya is not considered a safe place to disembark shipwrecked persons rescued at sea.Footnote 161
In conclusion, an observation is in order: the fact that cases like the Nivin and A.S. are being brought before the HRC is likely to be a strategic move by lawyers amidst fear that the case will be unsuccessful in the ECtHR, which is increasingly criticized for its ‘non-friendly’ approach to migrants’ human rights.Footnote 162 In this light, Dembour acknowledges that the ECtHR's jurisprudence in the area of migration has proven to be less helpful when it comes to protecting the rights of migrants than one ‘might have hoped, given that protecting human rights is the raison d’être of the Court’.Footnote 163 She wonders, ‘will it always be like this? It does not have to be’.Footnote 164
One can think of cases (albeit only a few) where the ECtHR adopted a pro homine approach to migration, such as M.S.S. v Belgium (M.S.S.) in which the transfer of an asylum seeker from Belgium to Greece, in accordance with EU legislation, was found to violate multiple provisions of the ECHR,Footnote 165 in a similar manner to the Hirsi and M.A. and Z.R. cases in which Italy and Cyprus were condemned for their practices and cooperation agreements with third countries that exposed migrants to serious risks. Whilst recognizing the difficulties faced by vanguard States due to their geographic location at the gates of Europe, the ECtHR in M.S.S. made it clear that this cannot pardon States of responsibility for their serious shortcomings.Footnote 166 However, in a more recent case, N.D. and N.T. v Spain, concerning the expulsion of irregular migrants and a State's sovereign right to manage its borders, the ECtHR failed to reflect on surveillance activities in the context of border practices that are employed without geographical or territorial limits and impact the rights of migrants negatively. Cases such as this displace the individual ‘from the heart towards the periphery of the ECHR system’.Footnote 167
V. CONCLUSION
It is readily apparent that under the veil of technological advancements, the extraterritorial practices of the EU and its Member States have been becoming increasingly cavalier towards international human rights law, revealing a lack of solidarity towards third-country nationals in need of protection. At the same time, while the ECtHR has been a source of hope for many, it has taken an increasingly complex stance in relation to issues on extraterritoriality that has undermined the protection landscape and empowered States to turn a blind eye to vulnerable people in need of protection. Amidst the escalating phenomena of abandonment at sea and contemporary push-back practices,Footnote 168 the ECtHR needs to continue adjusting the parameters of human rights jurisdiction to ensure that the ECHR remains a ‘living instrument’ that is ‘interpreted in the light of present-day conditions’.Footnote 169 As evident from the analysis of its jurisprudence in this article, the notion of territorial jurisdiction is no longer capable of being easily discerned.
Arguably, the label ‘technological advancements’ offers a useful lens for interpreting contemporary manifestations of State power and lends support to the argument for developing a functional model of jurisdiction. This is because technology increasingly provides the means to control remotely how a situation will unfold, often at the expense of migrants’ human rights—and lives. The words of Judge Pinto de Albuquerque in his concurring opinion in Hirsi highlight the present disdain for migrants and their plight: today's Europe is no longer ‘the cradle of human rights’.Footnote 170 For the cradle to be re-established, the human rights framework needs to be adapted to such societal changes, including the unprecedented impacts of technological developments.
ACKNOWLEDGEMENTS
The author is deeply grateful to Professor Siobhán Mullally for her in-depth comments and suggestions, and to Dr Efthymios Papastavridis and Professor Richard Collins for their valuable comments on earlier drafts. The author has also benefitted from helpful discussions with Professor Yuval Shany. An earlier version of this article was presented at: the Visiting Scholar/Visiting Researcher Colloquium 2024 at Harvard Law School; the European Human Rights Law Conference 2023 (Cambridge University); The Protection of Vulnerable People at Sea workshop (Swansea University); and the Society of Legal Scholars Annual Conference 2022 (King's College London). The author thanks the participants of these conferences and workshops as well as the two anonymous reviewers and editor-in-chief for their useful comments and feedback. The author is also indebted to the National University of Ireland for funding their Post-Doctoral research. As always, any errors or omissions remain the author's own.
Open access
