1. Digital, mobile, and smart

The ‘digitalisation’ of border security, which increasingly influences migration management, is a rapidly growing global phenomenon, often justified and expedited by perceived crises surrounding human mobility. Consequently, the deployment of new technologies, devices, and systems designed to enhance surveillance, monitoring, and registration processes has become a defining feature of contemporary border and migration control.

This process unfolds globally, particularly along the divides between the Global North and the Global South. One prominent example is the USA southern border in Arizona, where Integrated Fixed Towers equipped with radars, thermal cameras, and night-vision technology monitor ‘suspicious’ activities from up to seven and a half miles away (Aizeki et al. 2021). In addition, since February 2022, the United States has also been trialling four-legged robotic patrol dogs along its border with Mexico (Holmes 2022). In Australia, the Border Risk Identification System, operated by the Department of Immigration and Citizenship, utilises big data analytics to identify correlations and patterns, targeting so-called ‘risky travellers’. In Europe, in 2019, the European Border and Coast Guard Agency (Frontex) supported a pilot project that deployed a surveillance blimp capable of remaining airborne for forty days to monitor the seas off Samos Island (Leese, Noori, and Scheel 2021; Monroy 2022). Not far, at the border between Greece and Turkey, Long Range Acoustic Devices (LRAD), or ‘sound cannons’, were tested in spring 2021. These devices emit sound at volumes reaching 150 decibels – equivalent to a jet engine – posing a serious risk of permanent hearing loss (Gatopoulos and Kantouris 2021). Similar systems operate along the Hungarian border, where loudspeakers broadcast warnings in Farsi, Arabic, Urdu, and Serbian, notifying would-be border crossers that they are trespassing on Hungarian territory: «I’m warning you to hold back from committing this crime» (see Cockerell 2021). Additionally, in October 2021, Poland approved the construction of a €350 million wall, intended to incorporate motion sensors as the latest frontier of border security (Joly and Sandford 2021).

The introduction of thermal cameras, motion sensors, interoperable registration systems, biometric technologies, and drones in border zones reveals more than a quantitative increase in data collection and exchange, or a shift in the performance of surveillance practices. At stake is a profound transformation in how State sovereignty is exercised at borders. Indeed, digitalisation reshapes the dynamics of international travel and mobility, altering the very concept of borders by fragmenting and dislocating them (Everuss 2021, 339). Moreover, in addition to the widespread enthusiasm for technological experimentation, these innovations drive a fundamental rethinking of one of the most tangible and on-the-ground aspects of border enforcement: patrolling.

Long regarded as a core expression of modern State sovereignty and a central function of border control, integral to the filtering and sorting processes carried out at the frontiers, the practice of patrolling is in fact undergoing significant transformation due to the proliferation of new technologies. Their diffusion often unfolds within a complex and opaque web of security practices, offering fertile ground for critical enquiry into the broader dynamics of border digitalisation.

Today, borders are increasingly both mobile and omnipresent as they are designed to sort out individuals not only at territorial frontiers but also far beyond them. As Étienne Balibar (2009) observes, we are experiencing a moment of border vacillation: borders are no longer at the border but increasingly outsourced or subcontracted. This shift involves complex frameworks of cooperation with third States, greater engagement of private actors, the proliferation of border agencies, and the expanding mandate of International Organisations (IOs) in border-related issues – trends further amplified by the integration of technologically advanced systems. Borders, as a result, take on new shapes. Leanne Weber’s (2006) notion of «mobile borders», which captures the decoupling of border control from physical borders, illustrates this evolving landscape and highlights the pivotal role of new digital technologies in this domain.

The digitalisation of borders, exemplified by both the convergence of biometric databases and high-tech barriers – often referred to as ‘smart borders’ – primarily aims to enhance the efficiency of mobility control and filtering. Overall, these technologies are designed to prevent the entry of ‘undesirable’ individuals while facilitating seamless passage for pre-cleared travellers. Operating within a pre-emptive and anticipatory logic, they are presented as a one-size-fits-all solution for managing future indeterminate threats, positioning security measures around mobility in an increasingly proactive mode (Suchman, Follis, and Weber 2017).

Clearly, digitalisation in this field does not unfold in a political or legal vacuum. Instead, it reproduces existing power asymmetries and social hierarchies, mirroring broader drifts in migration governance that encompass restrictive procedures, practices, and legislation, which de facto limit access to international protection (see Pannia et al. 2018). The use of new technologies in border control is, therefore, neither neutral nor incidental, nor is it the use of certain technology to perform certain functions in a certain context. Rather, as technology is also socially constructed, it reflects how digitalisation is shaped by and embedded within broader governance frameworks.

While border digitalisation is a global trend, often embedded in regional initiatives that transcend national boundaries, the level of cooperation and coordination achieved in the European Union (EU) has not been replicated anywhere so far (Topak and Vives 2018). This makes the EU and its Member States, which are central to this book, especially relevant. Although efforts to harmonise border policies in regions such as North America are notable, the EU’s integrated approach stands apart in both scope and impact. Thus, this study focuses on Europe, where long-standing dynamics of externalisation, criminalisation, and privatisation in border and migration management are deeply entangled with the relatively recent digitalisation process.

In the EU, the increasing interest (and investment) in next-generation border technologies, often enhanced by Artificial Intelligence (AI), warrants closer scrutiny at least for two reasons. First, the digitalisation of borders is shaped by socio-technical interactions embedded in a complex interplay between legal obligations, societal values, and technological constraints (Burgess and Kloza 2021)1. Especially from a human rights perspective, this makes the consequences of their integration far from straightforward and inherently multifaceted. Second, the push for smarter border security, driven by converging narratives and practices around technology deployment across various legal systems, extends beyond border zones. It systematically spills over into other areas of society, meaning that the technological experiments carried out at the borders demand particular attention due to their broader implications.

Building on these premises, the book aims to critically examine the digitalisation of borders through the lens of ‘digital patrolling’, contributing to the broader discourse on border surveillance and digitalisation. By adopting a comparative approach within the EU context and engaging with an emerging and dynamic body of literature, this study seeks to provide new insights into the evolving field of border digitalisation.

2. A look at the literature: digitalisation, borders, and emerging perspectives

Notably over the last decade, critical security scholars, science and technology studies (STS) researchers, and political geographers have significantly contributed to the growing body of literature that examines the impact of digitalisation on the border-crossing experience (see Amoore 2006; Bigo 2014; Dijstelbloem and Broeders 2016; Leese 2016). This vibrant and multidisciplinary debate has led to the development of new subfields, such as digital migration studies (see Candidatu, Leurs, and Ponzanesi 2019; Leurs and Smets 2018; Sandberg et al. 2022).

At the intersection of critical security studies and STS, this literature has illuminated how material allowances and constraints shape security practices2, while also considering the subjectivity and agency of both those who operate the border and those who cross it (see Amicelle, Aradau, and Jeandesboz 2015).

Borders digitalisation has been described as a shift towards «techno-securitisation» (Marin 2016), a process of «smartening» of border security (Jeandesboz 2016), shaped by the «datafication» of mobility and migration management (Broeders and Dijstelbloem 2016), the diffusion of «techno-borderscapes» (Godin and Donà 2021), and the emergence of new forms of «dataveillance» (Degli Esposti 2014; Peoples and Vaughan-Williams 2010, 162) – defined as the monitoring and mining of multiple data types to identify potentially risky groups and individuals.

As previously mentioned, digitalisation and high technologisation of borders also fit within – and cannot be adequately understood aside from – a widely documented shift towards the criminalisation and containment of non-pre-vetted or unauthorised travellers through risk-based taxonomies, the externalisation of border control, and the outsourcing of migration management to private companies, international organisations, and third countries (see Csernatoni 2018, 177; Giuffré and Moreno-Lax 2019; Ferraris 2020; Nagore Casas 2019).

These strategies intertwine with forms of contactless, remote, and de-territorialised control, often described as preventive «non-entrée» policies and cooperative deterrence mechanisms, which also contribute to the shrinking of access to asylum and international protection (Mountz 2020; Tholen 2010, 273). Under this frame, increasingly hard-line and intrusive technologies for policing the European Union’s periphery are being justified and encouraged. Scholars have termed this shift «the transformation of the EU into a technological fortress» (Marin 2011), describing it as a deliberate security strategy for controlling migration within «emergent, recursively performed, and mutable» digitalised borders (Glouftsios 2021a, 9).

Karolina Follis (2017) describes this approach as a project of trans-territorial expansion, enabled by technologies that facilitate surveillance and governance in previously ‘unseen’ spaces beyond territorial borders, in the so-called ‘pre-frontier’ area, encompassing seas and lands adjacent to, but outside of, EU Member State jurisdiction. In this context, digitalisation enables forms of strategic «borders manipulation» (Weber 2006, 22), allowing for the shifting of frontiers in line with evolving security priorities.

These technological innovations thus challenge the notion of borders while raising significant concerns about the safeguarding of fundamental rights of people on the move. The Special Rapporteur on Contemporary Forms of Racism, Racial Discrimination, Xenophobia, and Related Intolerance’s September 2021 report highlighted how the digitalisation of migration policies and border enforcement often occurs with little regard for human rights abuses, racially discriminatory structures, and «experimental risks» (Human Rights Council 2021, para. 35). This is even more alarming as this context diverges significantly from other domains of social life similarly affected by the expansion of technological infrastructures. Firstly, people on the move often possess fewer rights and enjoy limited legal protections. Secondly, States still retain broad discretionary authority in matters of border and immigration enforcement, powers that can be largely expansive and at least partially shielded from judicial review.

Echoing the Special Rapporteur’s solicitude, at the European level, a resolution adopted by the European Parliament on 6 October 2021 concerning AI in criminal law and its use by police and judicial authorities highlighted serious concerns regarding the use of biometric data and facial recognition technologies within law enforcement practices (European Parliament 2021). These developments have in fact spurred criticism, particularly regarding the risk of indiscriminate mass surveillance in public spaces. In light of these concerns – but mindful of the significant disparity between the rapidity of technological advancements and the slower pace of legislative and regulatory frameworks – the proposed Artificial Intelligence Act recognises the need to classify AI systems used in migration, asylum, and border control as ‘high-risk’ (European Commission 2021, 10).

The digitalisation of border security – along with its attendant risks and potentials – has thus begun to attract attention in both academic literature and among international bodies and institutions, with discussions unfolding from diverse perspectives and across different normative schemes. Scholars have engaged with various facets of the border network, exploring the «technopolitics of border control» (Dijstelbloem 2021), the changing role of security professionals (Bigo 2014), the «dronisation» of borders (Csernatoni 2018, 178), and the deployment of satellite technologies for external border surveillance (Słomczyńska and Frankowski 2016). Other areas of enquiry include the expansion of biometric databases (Scheel 2019), the datafication of migrant populations (Broeders and Dijstelbloem 2016), the introduction of risk analysis algorithms (Amoore 2006), and the emergence of new rationalities within «deep border» regimes through machine learning (Amoore 2021; 2023). Additionally, attention has been drawn to large-scale information systems (Glouftsios 2021b), digital mapping interfaces aimed at visualising border zones and transboundary flows (Tazzioli and Walters 2016), and the discriminatory outcomes engendered by the datafication in this field (Leurs and Shepherd 2017; Broeders and Dijstelbloem 2016). On the other hand, human rights bodies have underscored the ‘protection crisis’ that the digitalisation process risks exacerbating.

3. Methodological considerations: a comparative and multidisciplinary perspective on digital patrolling

While patrolling is not absent from these discussions on border digitalisation, its transformations have rarely been treated as a distinct and essential lens of analysis. This work seeks to address this gap expanding the research agenda in this field. The value of this focus lies in exposing the frictions and contradictions within State-led border patrol activities, where security logics and border violence intersect with humanitarian aims, such as coordinating search and rescue operations at sea.

Within the scope of this analysis, digital patrol systems refer to a wide range of technologies deployed at and before the borders, including unmanned aircraft, drones, satellites, offshore sensors, thermal cameras, radars, autonomous surveillance towers equipped with infra-red cameras, high-resolution imagery, smart walls, and geographic information systems. It should be noted that this conceptualisation of patrolling deliberately centres on State-led activities in border control, which overlap significantly with migration management. Patrols conducted by non-State actors, including humanitarian organisations, are thus beyond the scope of the study3. Focusing on the European Union, the book endeavours to define and preliminarily classify the forms of digital patrolling implemented at various stages of borderwork4. It is thus based on the analysis of diverse ‘bordering’ strategies at both European and national levels showcasing the digitalisation of patrolling activities, with a particular emphasis on drone deployment at external borders and the functioning of integrated surveillance systems5.

This research focuses on three central aspects. Firstly, it seeks to define and contextualise digital patrolling, tracing its proliferation across the EU’s external borders. Secondly, it examines how and under which conditions Member States are advancing digital patrolling strategies and technologies as part of their broader border digitalisation efforts, specifically through a comparative analysis of this phenomenon in Spain and Greece. Finally, the study investigates the implications of digital patrolling for the rights of people on the move, expanding the analysis from the case studies to both European and international levels. Thus, it is questioned whether legal frameworks in place – national, European, and international – provide sufficient safeguards against the risks posed by digital patrolling, or whether they are instead facilitating indiscriminate digitalisation at the borders while being largely devoid of necessary protection measures.

This research adopts a multidisciplinary approach, engaging with a broad range of literature and sources. It draws on an extensive review of both primary and secondary materials, encompassing national and EU legislation, case law, reports, and policy documents. The triangulation of diverse materials is crucial for uncovering patterns, paradoxes, and rationales that reveal far-reaching and persistent tendencies, often greater in scope than individual data points might suggest. While not claiming exhaustive coverage, this approach resembles the construction of a mosaic, where each component contributes to a more complex and nuanced picture.

Of course, examining specific technologies, national policies, and legal frameworks within a meaningful and coherent dialogue presents significant challenges. To address such complexities, the study adopts a comparative perspective that builds conceptual bridges across disciplines and spaces6. Indeed, the comparative perspective inherently favours the integration of multidisciplinary approaches by examining diverse legal systems, cultures, and practices, thereby supporting a dynamic understanding of how different disciplines interact within specific legal contexts. As socio-legal methodologies suggest, a comparative approach is especially valuable when addressing cross-border technological issues, as it integrates non-legal knowledge into legal research, fostering a form of «methodological pluralism» that surfaces the global challenges underlying local developments (see Guerra 2018; Scarciglia 2015). In this sense, this book not only aims to advance the literature on digital patrolling but also to offer insights into the broader processes of border digitalisation within the European Union, through the analysis of the key case studies of Greece and Spain.

Engaging with this complexity offers significant promise: if digital borders are increasingly invasive and pervasive compared to physical barriers like barbed wire, their full impact is often concealed behind the veil of secrecy surrounding surveillance technologies (see Pallister-Wilkins, Goede, and Bosma 2020). This holds especially true in the context of border control and international mobility. As Tazzioli (2021) further observes, digital forms of border violence frequently remain obscured, operating beneath the threshold of political visibility. This opacity presents a notable challenge for research in this area, making it difficult to fully grasp the implications of digital surveillance practices.

The limitations imposed by this opacity are compounded by the constraints of this study7. Nevertheless, these challenges will be critically addressed on a case-by-case basis, acknowledging the potential gaps while drawing insights from the available sources. In carefully navigating these limitations, this research aims to offer a nuanced understanding of how the digitalisation of border control is reshaping contemporary practices of surveillance and mobility management.

4. Structure of the book and case studies selection

The book is structured into four chapters. Chapter 1 aims to establish the theoretical framework through which the process of digitalisation of border patrols is examined. Specifically, it scrutinises the various capability areas underpinning border security functions, enabled or reconfigured by digitalisation in this field. These capabilities include situational awareness, detection and tracking, information management, and risk analysis, all constitutive to the digitalisation of patrolling. The chapter also offers an analysis of the European Border Surveillance System (EUROSUR) and the deployment of drones at the EU’s external borders, alongside a reflection on future developments driven by increased investment in automated patrolling systems.

Chapters 2 and 3 present a cartography of emerging patrolling systems, focusing on case studies at the EU level and in two Member States: Greece and Spain. These countries, which have been interested in the most significant investments in border digitalisation, are emblematic «surveillance sandbox[es] at the frontiers of Europe» (Molnar 2022, 54). Located along the Eastern and Western migratory routes, they both provide invaluable insights into the digitalisation of border patrols, allowing for an exploration of how this process intertwines with national and European policies.

In both case studies, the analysis traces the specific ways in which digitalisation unfolds along the most critical sections of their external borders, highlighting the peculiarities of different patrolling systems. A thorough examination of the legal framework underpinning this digitalisation follows, identifying regulatory gaps and discrepancies between law and practice.

Finally, chapter 4 offers a comparative analysis of the findings from the case studies, with a particular focus on the tensions between digital patrolling and the obligation to respect and protect the fundamental rights of those subjected to such surveillance. From this, the discussion examines the broader implications for human rights from both European and international legal perspectives. Outlooks around human dignity, the right to international protection, privacy and data protection, as well as the principles of equality and non-discrimination, are discussed. The chapter concludes with reflections on the challenges of holding States accountable in times of digitally enhanced border surveillance.

To conclude and move forward, a few remarks concerning the selection of case studies are necessary. The decision to focus on Greece and Spain stems from the need to explore how the Member States that are most exposed to migratory flows are progressively digitalising their border patrol strategies, thus intertwining border security and migration policies to an increasingly indistinguishable degree (Topak and Vives 2018). At present, there are three primary migratory routes towards Europe: the Eastern, Western, and Central routes8.

Although the Eastern route affects not only Greece but also Croatia and Bulgaria, Greece has seen the most notable investments in border digitalisation. The country has, in fact, increasingly relied on technological experimentation to secure its maritime and land borders. This trend is particularly evident in the growing use of drones for patrols.

On the Western route, Spain stands out as a crucial case study for several reasons. First, Spain remains the primary European destination for people on the move travelling via the Atlantic route. Its geographical position, coupled with the location of the Canary Islands and the enclaves of Ceuta and Melilla, has driven the development of increasingly sophisticated remote surveillance systems since the early 2000s. The Sistema Integrado de Vigilancia Exterior (SIVE) is particularly noteworthy, often regarded as the forerunner of EUROSUR. As will be discussed, the SIVE deeply exemplifies the process of digitalising border patrols, particularly through its capabilities in situational awareness, detection, and risk analysis.

The Central Mediterranean route, which remains the deadliest and most heavily traversed route to Europe in 2022, is not the focus of a specific case study in this analysis for two reasons. First, the Member States along this route have made few recent advances in digital border patrolling, instead prioritising cooperation with third countries. While systems akin to Spain’s SIVE have been implemented, such as France’s Système Naval de Surveillance des Approches Maritimes et des Zones sous Jurisdiction Nationale (SPATIONAV)9, they are primarily deployed outside the Mediterranean. France’s digital patrolling initiatives, for example, focus more on internal borders with Italy and Spain, on the Calais region and the English Channel, in close collaboration with UK-led border control operations (Akkerman 2021, 154; Bonnevalle 2022)10. Such «internal externalisation» of border control (Barbero and Donadio 2019), gaining momentum since Brexit, lies beyond the scope of this study. Secondly, along the Central Mediterranean route, patrolling strategies and their digitalisation, are marked by a high level of cooperation, mainly coordinated by Frontex11. This dynamic highlights the role of «intermediary or hybrid agencies» (Bigo 2006, 391), operating between police and military functions, especially in response to perceived security threats at Europe’s borders. Given the predominance of Frontex-led operations – which, while integral to the broader context of this study, play a different role – the Central Mediterranean route is less relevant for a comparative analysis focused on the Member States’ initiatives in digitalising border patrols at the EU’s external borders.

1 The emphasis on the socio-technical dimensions of smart borders stems from efforts to reconcile the long-standing debate between essentialism and instrumentalism, revitalised by the recent “new materialist” turn in critical security studies. In brief – and with the caveat that this simplification may not fully capture the nuances of these positions – essentialist perspectives assert that technological development drives social change. According to this view, the emergence of new technologies directly triggers dynamics that shape political decisions made by States and social actors (see, for instance, Schweller 2014, 116). In contrast, instrumentalism challenges technological determinism by positing that technologies are passive and neutral tools, integrated into a socially constructed reality: it is social actors who ultimately harness technology to achieve specific objectives (see, among others, Herz 1976). Scholars such as Geoffrey Herrera (2006) and Stefan Fritsch (2011) have sought to bridge these two approaches, arguing that while technological innovations carry material consequences, those consequences are embedded within a social framework where human agency remains paramount. In this study, it is argued that such a perspective – acknowledging both the particularities of new technologies and the ensemble of social implications of border control – is especially fruitful for examining the digitalisation of borders.

2 On the ‘materiality turn’ in migration studies and STS, see Stephan Scheel, Evelyn Ruppert, and Funda Ustek-Spilda (2019).

3 For example, the work of AlarmPhone, particularly the Alarm Phone Sahara (2021) division, is noteworthy. This group periodically organises solidarity patrol operations in the Sahara Desert to locate and rescue migrant people in distress.

4 Chris Rumford (2008) effectively describes ‘borderwork’ as the process of «envisioning, constructing, maintaining and erasing borders».

5 On the relevance of the term ‘bordering’ to analytically capture the continuous processes of fixating and regulating mobility, see Nina Amelung, Rafaela Granja, Helena Machado (2021) and Martina Tazzioli (2018).

6 The relationship between social sciences and comparative public law, along with the fruitful intersections that arise from it, is greatly discussed by Ran Hirschl (2014).

7 This work has largely been written during the Covid-19 pandemic, which has restricted the ability to collect certain types of primary data, for instance through fieldwork. To gain a more comprehensive understanding of the phenomena under investigation as they unfold on the ground, I conducted eight in-depth interviews with a diverse range of stakeholders. These participants included researchers and academics, as well as journalists and activists from non-governmental organisations. The interviews, conducted online and lasting between 40 and 90 minutes, were designed with a purely exploratory purpose. As such, they are not directly relevant to the research design of the book and will not be explicitly referenced within its pages. Nonetheless, they have been invaluable in informing the factual reconstruction of recent developments related to the digitalisation of border patrols and in contextualising them.

8 From 2016 – the year after the so-called refugees’ crisis – until April 2022, 413.847 arrivals have been recorded from the route Eastern, 208.030 from the Western, and 450.501 from the Central route.

9 The SPATIONAV is a sea surveillance integrated system which produces a real-time operational picture fusing data from over 10.000 sources and connecting over 50.000 additional European and international tracks via a cyber-secured gateway. Launched in 2002, it is mainly based upon information collected by the French Navy and the Maritime Gendarmerie through a network of radars, cameras and infrared. The Sistème has a regional architecture based on three main hubs: SPATIONAV Channel/Atlantic, SPATIONAV Mediterranean, and SPATIONAV Antilles-Guyana. Moreover, France has recently started deploying drones to control its borders, and the data thus collected are very likely to be merged into the SPATIONAV.

10 In addition to the surveillance systems supporting patrolling activities in the Mediterranean, mention should be made of the Finnish Merivalvonnan tietojärjestelmä (Maritime surveillance system), MEVAT. SIVE, SPATIONAV and MEVAT are to date the most advanced integrated and remote surveillance systems in Europe. Similar systems, but with a lower level of technical development and a smaller surveillance area, are also widespread among other EU Member States. In Italy, for example, this surveillance role is covered by the Vessel Traffic Service (VTS).

11 Remarkably, since Frontex took over the Central Mediterranean surveillance with Joint Operation Hermes in 2011, its operations have been defined by broad mandates, prioritising border control over life-saving interventions at sea. Currently, Operation Themis, which replaced Operation Triton started in 2014, represents a key shift in digital border patrols, particularly with the transition from maritime to aerial surveillance marked by the 2020 launch of Operation Irini. As will be discussed, this change has significant implications for SAR capabilities. Moreover, the new aerial surveillance systems, including Multipurpose Aerial Surveillance (MAS) and satellites, are now used to detect migrant boats and assist third-country authorities, such as the Libyan Coast Guard, in conducting interception and return operations. These practices expose migrant people to severe human rights abuses today thoroughly documented, including arbitrary detention in inhumane conditions, where people on the move intercepted are denied basic rights.

1.1. Digitalising the EU borders

«Leonardo’s Falco EVO drone is used to monitor irregular migration during Frontex operation» headlines a Leonardo press release, published in July 2019. The drone, it reads, helped identify a «‘mothership’ trawler as 81 illegal migrants were transferred to smaller boats». Without dwelling on the fact that the persons on board are a priori identified as ‘illegal migrants’ and not, for instance, as potential asylum seekers, similar reports on the use of Unmanned Aerial Vehicles (UAVs) or Remotely Piloted Aerial Vehicles (RPAS) – commonly called drones – and other surveillance tools for the control of the external borders of the European Union seem bound to become more and more present. According to an interview released by an official from Frontex Press Office on a review specialised in unmanned systems, in 2022, Frontex aerial surveillance planes and drones detected over 35.000 migrant people attempting to cross the Mediterranean Sea and heading to Europe (Gurierrez 2022).

The digitalisation of the European Union’s borders is frequently depicted by actors engaged in borderwork through its substantial irreversibility. This idea is underpinned by a combination of path dependency, which propels the further dissemination of border and migration digitalisation as new technologies are incorporated into bordering practices, and a «technological solutionist» approach (Oliveira Martins and Jumbert 2020) or «techno-solutionism» (Morozov 2013), wherein every technological challenge is met with increased digitalisation, and border issues are framed as requiring technological solutions. Noteworthy in this context are the two contract award notices for aerial surveillance services issued in August 2021, directed at Frontex, amounting to €84.5 million1. On the one hand, Frontex designates drones as state-of-the-art instruments to enhance the effectiveness of border patrols and search and rescue operations. On the other, various nongovernmental organisations (NGOs) and humanitarian groups contend that these measures predominantly reflect a calculated strategy to perpetuate illegal pushbacks (Alarm Phone et al. 2020; Mazzeo 2021), thereby violating the principle of non-refoulement enshrining the right not to be removed, expelled or extradited to a State where there is a severe risk of being subjected to death penalty, torture, and other inhuman or degrading treatments.

Investments promoting border digitalisation and relying on technological solutions to address migration challenges are often spurred by situations framed in public discourse through the lenses of crisis and emergency. At the EU level, the debate on the digitalisation of borders commenced in 2002 with the development of the EU Integrated Border Management (IBM), framed as an operational supplement to Schengen cooperation and aimed at enhancing collaboration among Member States at the external borders (see Hanke and Vitiello 2019). The premise was that the freedom of movement within the Schengen zone necessitated integrated efforts in border governance to prevent external threats from penetrating the area.

In 2007, the EU Integrated Maritime Surveillance was adopted to promote coordination in monitoring Europe’s coasts, followed by the launch of the construction of a Common Information Sharing Environment (CISE) in 2014 (European Commission 2014). In February 2008, shortly after the so-called ‘cayuco crisis’ in the Canary Islands, the European Commission (2016a) presented a New Borders Package aimed at enhancing the governance of the external borders through increased digitalisation. However, it was the so-called 2015 ‘migration crisis’ that marked the pivotal moment for the adoption of an increasingly integrated smart borders system. In April 2016, a revised legislative proposal for the Smart Borders Package was approved, comprising measures to establish an Entry/Exit System (EES)2 and to implement significant modifications to the Schengen Borders Code3, aimed at fighting irregular migration and facilitating border crossings for ‘pre-vetted’ and ‘trusted’ non-EU travellers (European Commission 2016a; 2016c; 2016b).

Since then, information systems have been deployed both at and before European borders. At the border, the EU Agency for the operational management of large-scale IT systems in the area of freedom, security and justice (EU-LISA) manages three major information systems4: the European Dactyloscopy (EURODAC)5, the Schengen Information System (SIS II)6, and the Visa Information System (VIS)7. As argued by Sandro Mezzadra and Brett Neilson (2013), these large-scale IT information systems should be fully understood as bordering practices.

Before the border, information is exchanged and stored through the European Border Surveillance System (EUROSUR)8, which aims to make visible irregular movements across land and maritime borders. Today, new systems such as the Passenger Name Record (PNR)9 and the European Travel Information and Authorisation System (ETIAS)10 are also being implemented. The significance and scope of these databases have increased dramatically with the approval of Regulation 2019/818, which establishes a framework for interoperability between EU information systems in the fields of police and judicial cooperation, asylum, and migration11. Once again in an emergent context, recent facts associated with the Covid-19 pandemic have significantly accelerated the digitalisation of border enforcement and migration management. This transformation has coincided with a surge in border violence, largely justified under the pretext of what Maurice Stierl and Deanna Dadusc (2021) have termed the «Covid excuse».

Building on these developments, this chapter aims to define and investigate the modes of digital patrolling carried out at various stages of borderwork. First, it provides a critical mapping of the operations conducted through smart borders under the overarching concept of digital patrolling, thereby offering insights into the rationale behind the digitalisation of borders. The objective is to underscore the specificities and distinctions that the smartening of particular segments of borderwork entails or might entail. Second, the chapter endeavours to problematise the articulation of operations associated with digital patrolling by examining different tools and systems currently in use at European borders, including devices and systems currently under trial. To do so, the focus shifts to the capability areas enabled or (re)assembled by the digitalisation of patrolling. According to the terminology used in the latest Frontex report on Artificial Intelligence, capability areas denote selected skills employed to execute the corresponding border security functions, encompassing the work facilitated by various technologies (Frontex 2021).

The chapter is structured as follows. The first section proposes a definition of border patrolling, emphasising its positioning within the framework of border digitalisation. Engaging with the literature, it offers an overview of the capability areas shaping digital patrolling: situational awareness, detection and tracking, information management, and risk analysis. Each area is examined and articulated in relation to the deployment of drones at external borders and the functioning of EUROSUR. The aim is to elucidate the logic behind the progressive digitalisation of patrolling, showing how this shift can profoundly transform borderwork. Lastly, in light of the growing emphasis on achieving higher levels of automation, a brief discussion on potential future developments in digital patrolling is proposed.

1.2. Defining ‘digital patrolling’ along smart borderwork

Border patrol can be understood as the array of control and surveillance practices carried out at the external borders and adjacent zones to safeguard border areas, primarily by preventing irregular crossings. At the EU level, this definition aligns with the primary aim of border surveillance as delineated by the Schengen Borders Code, i.e. to prevent unauthorised border crossings, to counter cross-border criminality (for instance, related to terrorism, smuggling, or other illicit traffic), and to take measures against persons who have illegally crossed the border12. Border patrolling, therefore, includes measures to counter such activities also through intelligence, surveillance, and reconnaissance missions. At the same time, however, patrolling can cover search and rescue operations (SAR) as well as specific environmental missions to address issues such as oil spills, deemed particularly dangerous for border areas. Patrolling, therefore, necessarily appears multipurpose and – consequently – ambiguous, or at least non-neutral, about the diverse objectives it pursues.

As noted in the introduction, this definition is partial as it lies on an understanding of patrolling limited to the expression of State sovereignty in border areas. This approach acknowledges that such sovereignty is fragmented and redistributed among multiple actors (see Bigo 2022), but it leaves aside other forms of patrolling whose existence and significance should not be overlooked. For instance, it does not encompass patrol activities carried out within and across State territory for purposes similar to those justifying border patrols. Louise Amoore argues in this regard that border spaces are becoming «feature spaces», with the result that border policing activities potentially enter every space – «the city street, the university campus, the clinic» (Amoore 2021, 4). Similarly, this definition does not cover patrolling activities conducted by organisations or civil society associations under rather opposite rationales, notably to provide support along the migratory routes.

If this approach does not fully engage with the ambivalence of «digital passages and borders» (Latonero and Kift 2018), exploring digital border patrol within a narrowly defined angle can contribute significantly to critically unravelling the implications of digitalisation. Patrolling, in this context, represents in fact a mechanism through which terrains and spaces are actively reorganised through the exercise of sovereignty manifested in border control. The underlying rationales of this reorganisation can tangibly impact pre-border zones, redefine migratory routes, and attempt to reshape the rules and practices governing human mobility (see Scheel, Ruppert, and Ustek-Spilda 2019, 584).

Rogier van Reekum (2019, 629) argues that the practice of patrolling in migration and border governance is intrinsically linked to processes of visibility: it centres on rendering both the movements of people and the inherent violence of border enforcement visible. Patrolling does not merely aim at «enacting the ‘really real’ border»; rather, it involves «encountering life in webs of terrains and tactics», intersecting the vision afforded by the tactical domain with the specificities of individual border areas. This encounter, increasingly digitally mediated, can take different shapes or even be circumvented altogether through various externalised forms of expulsion or abandonment, exemplified by the instance of boats left to die. Therefore, patrolling becomes a matter of managing, maintaining, and scrutinising, increasingly reliant on data collection. This process breaks down space into digital fragments produced by and through digital patrolling, which are presented as neutral and self-evident.

In summary, digital patrolling extends to the intersections between the proposed understanding of border patrolling and the capabilities along which smart borders are articulated. It emphasises the complex interplay between digital borders and their physical and geographical counterparts, resulting in something emergent and new.

1.3. Situational awareness, detection, and tracking: conflicting rationales and multipurpose technologies

After discussing the proposed taxonomy, the analysis now shifts to the different modes of digital patrolling, focusing on exemplar technologies and functionalities. In particular, the transition from a ‘patrolling-driven’ to an ‘intelligence-driven’ strategy in border control (Dijstelbloem 2021, 105), resulting from increased technological mediation for surveillance purposes, is problematised. Attention is given to the technical and political justifications for this shift. As previously mentioned, the capability areas of situational awareness, detection, and tracking are interrelated and complementary within various devices and systems that facilitate digital patrolling and should thus be studied together. To illustrate how these capability areas are articulated throughout the digitalisation process, the functioning of EUROSUR and the deployment of drones at the borders are here discussed.

1.4. Information management and risk analysis: beyond filtering and pre-empting mobility

Two more capability areas – information management and risk analysis – enable border patrolling through increased technological mediation. The interconnection and interdependence of different segments of borderwork, particularly in collecting information via drones and tools that integrate into EUROSUR, have been partially addressed in previous sections. However, the mechanisms of data exchange and fusion that underlie digital patrolling, as well as the growing role of risk analysis and rating processes, should be further explored.

Firstly, data exchange and circulation are central to the increasingly externalised and outsourced management of borders and human mobility. This trend highlights the role of third-countries, agencies, and private companies also in patrolling schemes, having an impact on the use of drones and EMSA’s vessel detection service within the EFS framework, the routine satellite surveillance of maritime flows in pre-frontier areas and third countries, and the use of EUROSUR data. Notably, several EUROSUR contact points with Frontex headquarters in Warsaw exist not only within Member States but also in North African countries such as Morocco, Algeria, and Libya. Here, data collected by satellite stations, coastal surveillance stations, and remote surveillance platforms converge (Glouftsios and Panagiotis 2022). Remarkably, some information shared by third countries flows directly into National Situational Pictures thanks to bilateral agreements. While these arrangements are often shrouded in secrecy and generally lack substantial public oversight, their scope is deemed growing as the volume of data-sharing practices is likely to escalate rapidly.

Moreover, the exchange of surveillance information with third-country authorities, including satellite information for example under the EU’s Copernicus space programme (see Słomczyńska and Frankowski 2016)14, demonstrates how data sharing fragments patrolling, refining mechanisms for filtering ‘undesirable’ mobility and strengthening the surveillance of key areas of interest. For Frontex’s activities, cooperation with third countries typically involves working arrangements where information exchange and risk analysis are crucial. These arrangements allow data collected by drones and other digital patrolling tools in the pre-frontier area to be accessible to neighbouring third-country authorities15.

Additionally, through the Maritime Simulation Module Service (MSMS), images captured via digital patrolling also converge on EUROSUR. The MSMS works to facilitate predictions on suspicious or abnormal vessel movements through ship reporting systems like the Automated Identification System (AIS) and EMSA-developed algorithmic analysis. Once again, these systems aim to identify and make visible (and potentially governable) events occurring before the border (Glouftsios and Panagiotis 2022; Monroy 2021).

The digitalisation of patrolling through information sharing with third countries is particularly problematic, and has faced criticism. EUROSUR’s Regulation in Article 20(5) states that «[a]ny exchange of personal data with third countries in the framework of EUROSUR shall be strictly limited to what is absolutely necessary for the purposes of this Regulation» and must comply with data protection provisions. This is crucial especially if the data could allow to identify individuals or groups seeking international protection or at serious risk of fundamental rights violations. However, information exchange – though limited and conditional – remains possible and is not immune to potential creep. Monitoring the use of shared information by third authorities is challenging, and there is sporadic reporting of how the data is used in practice, especially concerning migrant people and asylum seekers’ rights in third countries (Marin 2020).

1.5. The future of digital patrolling: towards a new role for automation?

In mapping the landscape of digital patrolling, the rapid progression of border digitalisation has emerged quite clearly. As new technologies and increasing automation gain prominence in European policy documents and discourses, it is pertinent to indulge in some preliminary reflections on the future trajectories of digital patrolling by having a look at the current research and development initiatives across Europe.

In discussions on digital border patrolling, a central focus lies on the increased efficiency that new patrolling systems may offer, alongside an advocacy for expanding reliance on advanced technological expertise and state-of-the-art surveillance technologies. This approach positions technological ‘solutions’ as responses to perceived security ‘problems’ associated with managing migratory flows at the EU’s periphery. According to Raluca Csernatoni (2018, 191), studies and projects aimed at enhancing digital patrolling often promote a narrative focused on effectiveness and cost-efficiency, thereby normalising the use of emerging surveillance technologies while advancing militarization and military rhetoric in border management. This perspective is underpinned by a belief in the infallibility of technology, termed the «EU’s border technology fix» by Panagiotis Loukinas (2021, 4), which drives a push towards digitalisation.

In the aforementioned Frontex report on AI-based capabilities, alongside existing technologies such as surveillance towers, maritime domain awareness, and small unmanned aircraft systems, significant attention is given to technologies currently in development. Notably, these include new automated border control systems and diverse robotic systems (Frontex 2021, 23). These studies underscore how borders are increasingly viewed as prime sites for experimentation, to the extent that, as Claudia Aradau (2020) suggests, experimentality has become a primary rationale in the governance of border zones, leading to a «laboritazation of borders» (Bourne, Johnson, and Lisle 2015). The selection of border areas and vulnerable groups, such as migrant people and asylum seekers, for technological testing raises numerous ethical and legal concerns. This choice is particularly troubling given the stark power imbalances and the significant barriers these individuals face in asserting their rights or challenging violations arising from such experimental practices.

2.1. Setting the scene: digitalisation, militarization, and experimentation

Despite the scarcity of official documentation, evidence shows that drones are being utilised for border surveillance in several critical frontier areas in Greece (Molnar 2022). However, delineating the specifics of the deployment of Remotely Piloted Aircraft Systems – such as the technical attributes of these systems, the authorities responsible for their operation, and the modalities of their integration into border patrol operations – presents significant challenges. This difficulty is primarily due to the pervasive secrecy shrouding the digitalisation of patrols, both at the European level and, even more so, within national jurisdictions. As a result, a careful triangulation of fragmented information sourced from expert analyses, media reports, and procurement tenders is necessary.

Given this complexity, the analysis of the broader context within which these new surveillance technologies are being deployed becomes particularly valuable. The decision to experiment with new patrolling systems is, in fact, not merely a technical or operational matter; rather, it is deeply embedded within and shaped by the surrounding political, social, and legal contexts. A critical understanding of the situation at the Greek borders allows for a series of deductive inferences about the use of unmanned aerial vehicles, enabling a partial reconstruction of the phenomenon and suggesting paths to get around the limited access to primary sources. The opacity and inscrutability surrounding these developments are after all characteristic of new and evolving bordering practices that are taking place in the proximities of the military domain – areas that have been largely shielded from public scrutiny and judicial investigation.

In light of these challenges, to have a better understanding of this worn-out section of the mosaic, a comprehensive understanding of digital patrolling in Greece requires a broader perspective, a step back to take a look at the whole picture. This entails retracing the political and geopolitical role of border enforcement in Greece and examining the intersecting processes of technologisation and militarization that define this landscape.

The analysis thus begins by situating digitalisation within the wider legal and political frameworks that have shaped the militarization and experimentation along Europe’s external borders. Attention then turns to the strategic importance of these border regions, particularly concerning the allocation of funds aimed at advancing digital surveillance initiatives. Special focus is given to the deployment of drones by European agencies and national authorities for border monitoring. Following, the regulatory frameworks – positioned at the intersection of migration law and border control – governing the digitalisation of patrols and the use of drones are critically explored. Finally, these trends are considered against the backdrop of the endemic violence that has characterised the Greek borders in recent years.

2.2. Strategic frontiers: the Evros River, the Aegean Sea, and the Turkish side of the story

In examining the patrolling systems along the Greek-Turkish border, two key areas stand out: the Evros region, where the eponymous river demarcates the border, and the Aegean Sea, where Greek islands – frequently spotlighted due to the dire conditions in hotspots such as Chios, Samos, Leros, and Kos – stretch towards Türkiye. These borderzones are characterised by dense patrolling systems that, at the edges of these highly militarized regions, often give rise to extreme violence (Topak 2021, 6). The Evros border, which spans 206 kilometres, periodically draws national and European political attention, emphasising its geopolitical significance. This region is marked by the wide, fast-flowing Evros River, which branches into several streams, some of which are punctuated by small islands. Winter conditions are harsh in the area, with temperatures falling to ten degrees below zero. The Evros border is now a military exclusion zone, heavily monitored with cameras, searchlights, night-vision equipment, and sensors (Euro-Med Monitor 2021). Information from this zone is tightly controlled, with little available beyond what is officially released by the Ministry of Interior. Recently, several NGOs have raised concerns about an ‘information blackout’ from Evros, as access for researchers, human rights activists, and humanitarian workers has become increasingly restricted.

Since 2011, following the designation of the Evros route as a primary gateway to Europe for migrant people (Topak and Vives 2018), Greek authorities have fortified the border with a steel wall at key crossing points, built under the newly approved Integrated Border Management Programme for Combating Illegal Immigration (see Papatzani et al. 2020). The Programme, aimed at protecting both EU and national borders and reducing irregular migration, continues to have far-reaching implications. In May 2022, the wall extends over 38 kilometres, with plans to expand it by at least another 30 kilometres by the end of the year, incorporating more advanced surveillance technologies with substantial support from the European Union (Bathke 2021).

In the summer of 2012, Greek authorities supplemented the wall’s construction with Operation Aspida (Shield), which significantly increased the number of border officials and mainly resulted in the shifting of migratory routes from Evros to the Aegean Sea (Koca 2020; UN Office of the High Commissioner for Human Rights 2012). By September 2020, amidst heightened tensions with Türkiye and the instrumentalization of migrant people as political leverage, the Ministry of Civil Protection announced an escalation in the armament of the Evros border. This included the deployment of four drones for aerial surveillance, fifteen thermal cameras to detect nighttime migratory flows, ten armoured jeeps, and five inflatable boats to enhance patrols (e-evros.gr 2020). There is in fact evidence that drones, alongside other surveillance technologies such as night-vision goggles, thermal cameras, laser rangefinders, and pulse radars, are being tested and utilised in the Evros region (Molnar 2022). Moreover, since 2021, Frontex has introduced advanced aerostats in the area, capable of remaining airborne for up to forty days, marking a significant innovation in border surveillance (Monroy 2022b).

Similarly, surveillance in the Aegean Sea has intensified steadily over the years. Maritime borders are increasingly monitored by vessels and air units equipped with digital patrolling technologies, enhancing detection, tracking, and risk analysis capabilities, and fundamentally enlarging the scope of situational awareness in the region. Much like the Evros border, the Aegean Sea has seen a shift from intensive physical patrols to remote controls. The literature extensively discusses how this push for greater efficiency through technologically advanced and smart systems is largely driven by exclusionary rationales, aiming to effectively keep people out (Topak 2014).

Already in 2009, the Greek Coast Guard began utilising the Automatic Identification System (AIS), a tracking mechanism for ships and boats, instrumental also in identifying vessels used by people on the move heading to islands such as Lesbos, Chios, Samos, or Patmos, which lie just a few kilometres from Türkiye. The introduction of the Integrated Border Management System, comprising the Surveillance Operational Centre (SOC), has accelerated the shift towards digital and remote patrolling. This system allows multiple border sections to be monitored simultaneously and communicates with patrol units by analysing real-time data from various surveillance sources. The SOC is linked to the Greek National Coordination Centre (NCC), which, as discussed in the previous chapter, transmits data to EUROSUR, contributing to the European Situational Picture and the Common Pre-frontier Intelligence Picture. The rationale behind this development is rooted in a preventive approach to border control. As reported by Özgün E. Topak (2014, 826), the SOC is in fact particularly valuable for its ability to monitor border situations and direct patrol units precisely to locations where migrant people are approaching.

Moreover, it should be noted that the digitalisation of patrols extends beyond the Greek side of the border. On the Turkish side, border areas are also subject to intense surveillance. The increasing digitalisation and militarization of Turkish border regions sometimes align with Greek efforts to contain and limit arrivals, while at other times, they contribute to documenting Greek border operations, including pushbacks.

The use of advanced security technologies in Türkiye dates back to 2011, when the country became the world’s largest refugee host following the outbreak of war in Syria. The European Union played a significant role in modernising the Turkish-Greek border, supporting the establishment of additional patrols at both sea and land borders. In May 2012, Turkish authorities signed a memorandum of understanding with Frontex to prevent irregular migration and update border surveillance systems through cooperation in risk analysis, training, research, and development (Frontex 2012). Once again, on all sides, these operations were justified with references to both security threats and humanitarian concerns for the lives of people on the move in border areas (Koca 2020).

Of course, such technical agreements should be viewed within the broader framework of cooperation between Greece, the European Union, and Türkiye on migration management and control. In this context, the conclusion of the EU-Türkiye Statement stands out as a pivotal moment. Agreed upon in March 2016 during the height of the so-called 2015 refugee crisis, the Statement was signed by Türkiye and the Member States gathered within the Council – the latter acting on behalf and in the interest of the Union, despite not in the Council capacity – under a nebulous legal framework. The Statement remains central to the EU’s externalisation strategy. Through this controversial and debated agreement, Türkiye was declared a ‘safe third country’ under Article 38 of the Asylum Procedures Directive1. In essence, Türkiye committed to receiving and protecting approximately three million Syrian refugees in exchange for substantial funding and the initiation of visa liberalisation negotiations for Turkish citizens (see Favilli 2018). A ‘fast-track’ border procedure was introduced in Greece by Law 4375/2016-55 (Art. 60 para. 4)2, de facto establishing hotspots at external borders to facilitate the Statement’s implementation. The core of the Statement lies in a non-arrival policy, with Türkiye pledging to take all necessary measures to shut down maritime and land routes to Europe (Petracou et al. 2018). This commitment includes enhancing surveillance capabilities and adopting a preventive border control approach using advanced technologies. Consequently, there have been several reports of joint patrol operations between Greek and Turkish forces, where people on the move intercepted by Greek remote patrolling systems have been apprehended by Turkish patrol units.

Recently, Türkiye has also developed a sophisticated drone surveillance system over its border areas. According to a map published by a local newspaper in Evros, Turkish drone patrols have significantly increased since 2019, with extended flight hours along the Greek border (e-evros.gr 2021b). During periods of heightened tension with the EU, Türkiye has publicly highlighted numerous instances of violence and violations of the non-refoulement principle by the Greek Coast Guard in the Aegean Sea. Various videos circulated online by the Turkish Armed Forces, supported by drone footage, document pushback operations carried out by Greece3, illustrating a new ‘materiality’ and ‘visibility’ in digital patrolling that may lead to emerging forms of public scrutiny (İşleyen 2021). Not surprisingly, similar processes of drone surveillance, image collection, and public dissemination have also been employed by Greek authorities against Türkiye.

2.3. National and European priorities and funding

Investing heavily in state-of-the-art technologies for border patrolling is resource intensive. Over the years, the European Union has provided substantial support to address the border security needs expressed by Greek governments. The shift toward smart patrol systems emerges clearly when examining the projects, actions, and tools developed within the framework of European funds dedicated to border surveillance.

During the 2007-2014 financial period, the European Borders Fund played a significant role in supporting the purchase and use of coastal patrol vessels, particularly favouring very high-speed coastal patrol vessels (VHSCPVs) and high-speed boats for special operations (HSBSOs), which are mainly useful for rapid patrols at night (External Borders Fund 2007). Additionally, since 2007, several motion sensors have been installed in areas close to the borders, indicating a gradual shift toward technologically enhanced border control.

This trend continued under the Internal Security Fund (ISF) for the 2014-2020 period4, where significant emphasis was placed on developing and installing the Maritime Borders Surveillance System and extending the automated surveillance system at the Greek-Turkish border (Hellenic Ministry of Citizen Protection 2015, 36). These measures have led to increasingly remote management of surveillance, aimed at preventing unauthorised mobility, that include the installation of cameras and radars capable – according to media reports (e-evros.gr 2021a) – to monitor up to 15 kilometres into Turkish territory. The strategic objectives outlined by the Greek government under the ISF prioritised improving national situational awareness capacities and enhancing integrated border management within the EUROSUR framework, with a focus on promoting automated surveillance systems. The goal was to transition from a man-based to a technologically assisted surveillance system, with significant contributions from Frontex (Hellenic Ministry of Citizen Protection 2015).

A key funding priority under the ISF was the deployment of a National Integrated Maritime Surveillance System (NIMSS), a network of integrated surveillance stations connected to an operational control centre. Although the NIMSS was expected to be fully operational by mid-2021, delays in the tender process mean that the first surveillance stations are expected to be fully functional by 2023 (Hellenic Ministry of Citizen Protection 2015). According to national security advisors, the NIMSS will significantly enhance border surveillance while minimizing unnecessary involvement of patrol vessels (Dokos 2021). Additionally, the Greek authorities have procured unarmed drones for surveillance of the sea borders in the East Aegean, as acknowledged by the European Commission (Lagos 2019).

The ISF also supported the purchase of hardware and software for risk assessment analysis aimed at preventing illegal entries at the borders. The national program defined by the Greek Ministry of Citizen Protection explicitly mentions the deployment of new technological equipment, such as mobile scan units, mobile heartbeat detection devices, Closed-Circuit Television Cameras (CCTV) systems, and UAVs to detect illegal migrants (Hellenic Ministry of Citizen Protection 2015, 18). This is one of the few governmental sources that explicitly outlines the intention to systematically integrate drones into the patrol system, with policy objectives focused on strengthening preventive policing and containing migrant people before they reach the EU’s external borders (see also Human Rights Watch 2022).

Without taking into account the allocation of additional funds under the Borders Emergency Assistance (in particular in 2019), Greece benefited from almost €167 million under the ISF, second only to the €195 million allocated to Spain (European Commission 2020a). Recently, Greece has invested heavily in technology-led policing and border management, including advanced facial recognition and biometric processing software. Police officers have also been provided with smartphone-sized devices capable of collecting and storing such data (Chelioudakis 2020). A ‘Smart Policing Program’, aimed at increasing the efficiency of identification systems for third-country nationals, was also introduced by the Hellenic Police in April 2021 (European Union 2021). This program has faced criticism from various organizations for non-compliance with fundamental rights, in particular with the principle of non-discrimination, and the most basic privacy and data protection provisions (Human Rights Watch 2022).

Looking ahead, the initiatives funded under the 2021-2027 Integrated Border Management Fund indicate a persisting trend towards the use of increasingly advanced technologies to enhance the automation of patrolling activities. This is also supported by a recent draft decision by the EU Commission, which calls for Greece to further enhance its surveillance capabilities at land and sea borders by increasing reaction capabilities and deploying additional thermo-vision vehicles, cameras, helicopters, and drones (together with «a number of service dogs»)(European Commission 2022). Although the draft also emphasizes the need for Greece to strengthen the fundamental rights component of border management and to investigate pushback allegations, it highlights that the Greek authorities will need to further tighten controls at the borders.

2.4. Deploying drones: Frontex, EMSA, and national authorities

Following this overview of the advanced bordering practices in Greece, it is possible now to focus on the specific role and deployment of Unmanned Aerial Vehicles.

In Greece, the deployment of UAVs for border surveillance is mainly attributed to European agencies like Frontex and the European Maritime Safety Agency, although national authorities also operate drones in various border areas, particularly in the Evros region. In the European Union, competencies in the field of border security and control remain in fact closely linked to the exercise of State sovereignty. Thus, once advanced surveillance tools such as UAVs are introduced for security purposes by EU agencies, national authorities are likely to expand their use and claim more competencies over them. In this case, Frontex and EMSA have been pivotal in integrating drones into patrol systems.

Frontex’s involvement at the Greek borders became significant in 2010, when the Greek government requested specific assistance, leading to the deployment of the Agency’s Rapid Border Intervention Teams (RABITs)5. This operation, the first of its kind in Europe, ended in March 2011 but marked the beginning of Frontex’s sustained presence in Greece, particularly in the Evros region. The Agency contributed to the militarization of patrol systems and the increasing preference for remote surveillance tools in the area (Topak and Vives 2018). Frontex has also been instrumental in establishing a risk management approach to border control, focusing on future-oriented, remote interventions designed to «secure unknown futures» (Amoore 2013, 153).

Since 2010, several operations have taken place at the Greek-Turkish borders, with Joint Operation Poseidon being particularly significant. Initially aimed at controlling the maritime borders between Greece and Türkiye, this operation was reinforced in 2016 to assist Greek authorities with enhanced border surveillance capabilities and a higher number of officers (Ilias et al. 2019, 25). More recently, between March and October 2020, Frontex launched the Rapid Border Intervention Aegean 2020 and the Rapid Border Intervention Evros 2020 operations at the Greek-Turkish land and sea borders (Frontex 2020). Currently, Frontex operates under the multipurpose Operation Poseidon, covering both Greek sea borders with Türkiye and the Greek islands (Frontex 2021b). Despite the absence of official statements, there is evidence that UAVs have been deployed in these operations (see Monroy 2020). Moreover, there is evidence of collaboration between Frontex and Greek authorities aimed at indiscriminately reducing arrivals, even at the expense of potential international protection applicants. This was highlighted during a meeting in May 2021 between Greek Prime Minister Kyriakos Mitsotakis and Frontex head Fabrice Leggeri, where they complimented each other over cooperation resulting in an 80% reduction of arrivals to Greece in 2020 and a further 72% reduction registered until the spring of 2021 (European Council on Refugees and Exiles 2021).

EMSA’s public communication is more transparent regarding its use of drones to support Member States in enhancing maritime surveillance capabilities. For example, EMSA records indicate that, in 2019, Greece was provided with remote-controlled surveillance aircraft for a total of 155 flight hours, designated for use by both the Hellenic Coast Guard and Frontex (European Maritime Safety Agency 2021). EMSA’s 2022 Agency Outlook confirms the continued provision of Remotely Piloted Aircraft System services to Member States and EU agencies, with plans to develop multipurpose regional services for operational capability sharing among neighbouring coastal states (European Maritime Safety Agency 2022). While in 2022 these systems are primarily intended for monitoring oil spills, the same national and European agencies involved in maritime border surveillance are the final users of these drones, raising concerns about the potential extent of their use beyond EMSA’s stated objectives.

At the national level, the deployment of drones is closely tied to Greece’s security apparatus, including the Armed Forces, Police Services, and the National Intelligence Service. In 2017, the Greek Deputy Minister of Citizen Protection announced that the Hellenic Police had been equipped with drones for border surveillance, aiming to intensify border controls with Türkiye (Loukinas 2017). More recently, in 2020, the Hellenic Police also procured two drones for border patrols under the HEFESTOS project (Hellenic anti-Fraud Equipment and relevant training for Strengthening the Operability against Smuggling) (Hellenic Republic Ministry of Civil Protection 2020; Homo Digitalis 2020b). The Western Greece Region as well procured UAVs for security forces, intended for various purposes, ranging from environmental and civil offence detection to smuggling prosecutions, and area surveillance (Region of Western Greece 2020). In November 2021, the Greek Navy purchased five unmanned helicopters for maritime security and surveillance, specifically the New Alpha 900, known for their redundancy in critical systems, making them suitable in particular for target acquisition and reconnaissance operations (McNabb 2021). Additionally, media reports suggest that Greece will soon lease two more Heron-1 UAVs from an Israeli company to enhance surveillance in the Aegean, explicitly meant to manage a potential new wave of migrant people and asylum seekers (see Monroy 2022a). These developments underscore the increasing, albeit partially opaque to public scrutiny, presence of drones in Greek border surveillance.

From an organizational perspective, the introduction of new surveillance systems has created a significant divide between Greek agencies and officials working remotely from operational centres and those conducting physical patrols on the ground. This divide, as highlighted by Dijstelbloem et al. (2017), exacerbates the risk of a divergence in the perception of situations ‘remotely’ versus ‘on the ground’, potentially having serious consequences from a fundamental rights perspective. The growing reliance on digital patrolling reduces direct encounters between patrols and migrants, which could, on the one hand, limit episodes of violence and arbitrary conduct but, on the other, contribute to the dehumanization of border enforcement, a concern that will be further explored in chapter 4.

2.5. Delineating the use of drones for border patrols: a rhizomatic and vanishing regulatory framework

Just as the analysis of the integration of drones in border patrols, reconstructing the regulatory framework governing the use of drones for border patrols is fraught with challenges. The complexities arise from several factors: the limited and fragmented information available on these regulations, the operational and technical nature of the existing framework, and the ongoing development of new legal structures, which make the regulatory landscape appear rhizomatic – an interconnected and somewhat elusive system that is difficult to analyse comprehensively (Marin 2017). Furthermore, the novelty of digital patrolling means that there is a lack of case law directly addressing the use of drones in border surveillance.

Under these circumstances, to understand the regulatory context in which drone use for border patrols is situated, as for the previous paragraphs, it is useful to first explore the broader trends in Greek migration law and border control policies. This exploration highlights once again the increasing securitization and militarization of borders and the corresponding reduction in protection for people on the move.

2.6. The use of drones in endemic border violence: the case of pushbacks

The deployment of drones for border surveillance in Greece must thus be critically examined within a context characterised by a lack of transparency, particularly in militarized zones, and within a regulatory framework that remains vanishing. This practice also aligns with a broader strategy aimed at limiting access to third-country nationals under the guise of migration prevention. Moreover, in recent years Greece has gained notoriety for the widespread use of indiscriminate violence and refoulement at its external borders. While the fourth chapter will provide a more detailed examination of the rights at risk in digital patrolling, it is wise to first outline some of the most notable instances of pushbacks that have precluded the possibility of lodging claims for international protection. In this context, the digitalisation of border patrols through the use of drones raises significant concerns, particularly regarding the facilitation of these illegal expulsions.

Over time, the evidence of systematic and violent pushbacks has become increasingly irrefutable. To focus on recent developments, in February 2021, the Greek National Commission for Human Rights (GNHCR) submitted a report to the UN Special Rapporteur on the Human Rights of Migrants, highlighting the growing number of recorded pushbacks (Greek National Commission for Human Rights 2021). The GNHCR noted that, despite numerous complaints lodged by organisations such as the Hellenic League for Human Rights and the Greek Council for Refugees since 2017, no Greek court has yet had the opportunity to rule on the issue. However, several complaints regarding border violence and pushbacks in contravention of the non-refoulement principle are currently under examination by the European Court of Human Rights31.

In April 2021, the Greek Ombudsman released an interim report as part of an investigation initiated in 2017, condemning the relevant authorities for their failure to adequately investigate allegations of pushbacks at the Greek-Turkish land border (Ombudsman 2021). The Greek Helsinki Monitor (GHM) in May 2021 filed a criminal complaint concerning 147 pushbacks involving over 7.000 individuals between March and December 2020. This complaint was subsequently forwarded by the Supreme Court prosecutor to first-instance prosecutors for further investigation. The gravity of these incidents has prompted scrutiny and calls for investigations from both the Council of Europe’s Commissioner for Human Rights, Dunja Mijatović, and the EU Commissioner for Home Affairs, Ylva Johansson (Commissioner for Human Rights 2021b). However, despite persistent calls from the United Nations High Commissioner for Refugees (UNHCR) and the European Commission for a rights-monitoring mechanism at Greek borders, the Greek Migration Minister formally rejected such a mechanism in October 2021, citing concerns over national sovereignty (European Council on Refugees and Exiles 2021).

These violations are not limited to Greek authorities alone; Frontex has also been implicated. In October 2020, an investigation by a consortium of media outlets including Lighthouse Reports, Bellingcat, Der Spiegel, ARD, and Asahi TV claimed that Frontex was complicit in and aware of several illegal pushbacks and collective expulsions of asylum seekers in the Aegean Sea (Bellingcat 2020). This revelation prompted an internal investigation by the Frontex Management Board, which could not definitively rule out the agency’s involvement in pushbacks (Frontex 2021a). By July 2021, the European Parliament published a report following a fact-finding investigation into Frontex’s alleged fundamental rights violations, concluding that while there was no decisive evidence of direct pushback operations by Frontex, the agency failed to address substantial evidence of fundamental rights violations at Greek borders (European Parliament 2021). Today, Frontex continues to be at the centre of different investigations and legal actions in the country32.

Against this backdrop, the digitalisation of border patrols, including the use of drones by both Greek authorities and Frontex, raises profound concerns. While instances where drones have facilitated rescue operations at sea are occasionally publicised in government and agency press releases, the use of drones in supporting illegal pushback operations remains shrouded in secrecy. To fully understand the purposes behind drone deployments, it would be essential to have a comprehensive view of other assets contributing to border patrols. For example, if it could be established that drones are primarily used to enhance situational awareness within existing surveillance frameworks to support other units already on the ground or at sea, it would suggest that they are an integral part of broader border control mechanisms. Conversely, if drones are found to be replacing conventional patrols, it would indicate a shift in priorities away from rescue operations at sea. However, accurately reconstructing the operational landscape remains challenging, highlighting the need for further investigation.

In a context marked by widespread and often violent expulsions, it is reasonable to surmise that the deployment of advanced surveillance technologies, including drones, will reinforce these established practices. Nonetheless, there is currently no documented case of a pushback operation being directly facilitated by drone-collected data. Given the covert nature of such operations, it is highly unlikely that official sources would acknowledge this dynamic, and no judicial proceeding has addressed the issue to date.

Nevertheless, testimonies collected by organisations such as the Border Violence Monitoring Network (BVMN) provide a glimpse into the potential implications of drone surveillance in border control. Although BVMN does not specialise in new technologies, the network has in fact documented episodes that substantiate the fears of human rights activists on the use of drones for border control.

According to BVMN, for instance, on 6 October 2021, a group of approximately 200 people, including minors, attempting to cross the Evros River from Türkiye to Greece reported encountering a drone overhead (BVMN 2020). Shortly after, they were intercepted by Greek officers, their belongings were confiscated, and they were detained without food or water. The day after, they were forcibly returned to Türkiye across the Evros River, where Turkish military personnel were waiting. Similarly, on 27 August 2020, a drone reportedly facilitated the pushback of nearly 80 individuals in the same region, who were stripped of their personal belongings, including valid UNHCR-issued Khartias (temporary residence permits), before being expelled (BVMN 2020). Comparable incidents have also been reported at the Croatia-Bosnia border (BVMN 2019). Moreover, evidence suggests that when political conditions permit, drones and radars are also employed by Turkish authorities to carry out pullbacks, intercepting people on the move before they reach Greek territory, thereby systematically curtailing their chances of seeking asylum (İşleyen 2021, 1094).

Many scholars and organisations argue that these cases are not isolated incidents but reflect a broader strategy in which drones are used to intercept, apprehend, and push back people on the move, either directly or indirectly (Topak 2014, 824). What is indisputable is that the digitalisation of border patrols has not reduced recorded deaths during border crossings. Furthermore, the narrative that advanced surveillance technology enhances rescue operations is contradicted by numerous documented cases where Greek authorities have allegedly used distress signals to locate and push back migrant boats, rather than rescue them (ProAsyl 2013, 28). Reports also suggest that Frontex’s surveillance technology is utilised to oversee illegal pushbacks, suggesting that further examination of the intersection between technology and human rights violations in the activities conducted by the agency is necessary (Habib 2021). These observations gain further significance when considered alongside recent ethnographic research by Covadonga Bachiller López (2022), which accurately describes Frontex’s role in early detection tactics aimed at deterrence and externalisation, systematically performed by Frontex in collaboration with the Hellenic Coast Guard since early 2020.

In conclusion, while the use of drones in pushback operations is widely suspected, it remains today difficult to definitively prove. The deployment of these technologies in such operations could be described as a ‘common secret’, understood by many but officially acknowledged by none. The increasing reliance on drones for border surveillance, within a context of endemic border violence, thus raises profound ethical and legal concerns that require urgent attention and further investigation.

1 Directive no. 2013/32 on common procedures for granting and withdrawing international protection (Asylum Procedures Directive), 26 June 2013.

2 Law no. 4375/2016 on the organization and operation of the Asylum Service, the Appeals Authority, the Reception and Identification Service, the establishment of the General Secretariat for Reception, the transposition into Greek legislation of the provisions of Directive 2013/32/EC, 3 April 2016.

3 See, e.g., “Drone footage shows Greece pushing back asylum seekers in Aegean – 05.04.2021”, available at: https://www.yenisafak.com/en/video-gallery/news/drone-footage-shows-greece-pushing-back-asylum-seekers-in-aegean-2206349

4 For a compendium of key documents on national legislation for the use of ISF funds, see (only in Greek): https://www-ydeap-gr.translate.goog/isf-b-v-tameio-esoterikis-asfaleias-synora-kai-theoriseis/nomiko-plesio1/?_x_tr_sl=el&_x_tr_tl=it&_x_tr_hl=it&_x_tr_pto=sc

5 At EU level, Member States can initiate joint operation and rapid interventions with the possibility for Frontex to intervene at the external border of a Member State, as foreseen in Regulation no. 2016/1624 on the European Border and Coast Guard, replaced in 2019 by Regulation no. 2019/1896.

6 Law no. 4636/2019, on international protection and other provisions, 1 November 2019 and Law no. 4686/2020, improvement of the migration legislation, amendment of Law 4636/2019, 4375/2016, 4251/2014 and other provisions, 12 May 2020.

7 Law no. 4825/2021, reform of deportation and return procedures of third country nationals, attracting investors and digital nomads, issues of residence permits and procedures for granting international protection, provisions within the competence of the Ministry of Migration and Asylum and the Ministry of Citizen Protection and other emergency provisions, 4 September 2021.

8 See Ilias et al. (2019, 22) for a comprehensive reconstruction of the Greek national legal framework on border surveillance.

9 Law no. 3838/2010, on Current Provisions related to Greek Nationality and the Political Participation of Expatriates and Legally Residing Immigrants, 24 March 2010.

10 Council Regulation no. 2007/2004 establishing a European Agency for the Management of Operational Cooperation at the External Borders of the Member States of the European Union, 26 October 2004 (no longer in force) and Regulation no. 1168/2011 amending Council Regulation (EC) no. 2007/2004 establishing a European Agency for the Management of Operational Cooperation at the External Borders of the Member States of the European Union, 25 October 2011 (no longer in force).

11 Law no. 4249/2014, Reorganization of Greek Police, Fire Brigade and the General Secretariat for Civil Protection, upgrade of the Services of the Ministry of Public Order and Citizen Protection and regulation of other matters concerning the Ministry of Public Order and Citizen Protection, 24 March 2014.

12 Law no. 4332/2015, Amendment of the provisions of the Greek Nationality Code – Amendment of Law 4521/2014, 9 July 2015.

13 Law no. 4650/2019, Regulations of issues of the Ministry of National Defence and other provisions, 17 December 2019.

14 Law no. 3917/2011, Retention of data generated or processed in connection with the provision of publicly available electronic communications services or of public communications networks, use of surveillance systems with the obtaining or recording of sound or image at public areas and relative provisions, 21 February 2011.

15 Directive no. 2006/24 on the retention of data generated or processed in connection with the provision of publicly available electronic communications services or of public communications (Data Retention Directive), 15 March 2006 (no longer in force).

16 CJEU, Joined Cases Digital Rights Ireland Ltd v. Minister for Communications, Marine and Natural Resources and Others and Kärntner Landesregierung and Others, C-293/12 and C-594/12, 8 April 2014.

17 An in-depth commentary on these cases can be found in Podkowik, Rybski, and Zubik (2021).

18 Presidential Decree no. 98/2019, Organisation and structure of the Drone Service, 25 October 2019. The Decree introduces significant changes to Article 25 of Law 2800/2000 regulating the operating provisions regarding air means deployed by Security Forces and covering in particular the suppression of fires and the rescue of victims in case of disasters.

19 Presidential Decree no. 98/2019, Organisation and structure of the Drone Service, Establishment of Procurement Directorates and History of the Hellenic Police and Unmanned Aircraft Service, 21 March 2017.

20 Personal data, according to the definition provided by Article 4 para.1 of the Regulation No. 2016/679 (General Data Protection Regulation), means «any information relating to an identified or identifiable natural person (‘data subject’); an identifiable natural person is one who can be identified, directly or indirectly, in particular by reference to an identifier such as a name, an identification number, location data, an online identifier or to one or more factors specific to the physical, physiological, genetic, mental, economic, cultural or social identity of that natural person».

21 Law no. 4624/2019, On the Hellenic Data Protection Authority, the implementation of Regulation 2016/679 and the transposition of Directive 2016/680, 29 August 2019.

22 Directive no. 2016/680 on the protection of natural persons with regard to the processing of personal data by competent authorities for the purposes of the prevention, investigation, detection or prosecution of criminal offences or the execution of criminal penalties, and on the free movement of such data (Law Enforcement Directive), 27 April 2016.

23 Homo Digitalis (2020a, 23) offers an in-depth analysis of the specificities of ‘new technologies’, proceeding by analogy from the List of types of processing operations subject to the requirement to carry out a data protection impact assessment (2018) published by the Hellenic Data Protection Authorities and addressing the provisions of the GDPR. The list is considered a relevant guide in interpreting the provisions defined by Directive 2016/680 apply. In fact, the Authority explicitly states that a relevant example of systematic and large-scale processing is given by the processing of data through drones.

24 On impact assessment and new technologies in border control, see also Burgess and Kolza (2021).

25 Notably, the explanatory memorandum accompanying draft law 3917/2011 clarifies that the provisions established by the act do not apply in situations where the processing of personal data is unfeasible, such as in cases involving systems that capture remote images without the capability to enlarge and identify individuals or other information pertaining to natural persons.

26 It is also worth noting that Homo Digitalis (2020e) submitted an official inquiry regarding the use of drones by the Hellenic Police. However, it appears that no response has been received to date.

27 Presidential Decree no. 75/2020, Use of surveillance systems obtaining or documenting sound and pictures in public places, 10 September 2020.

28 Although it falls beyond the scope of this analysis, it is worth mentioning that this Decree has been strongly criticised by several organisations, such as Amnesty International (2021), for infringing or potentially infringing on the right of assembly and expression.

29 Regulation no. 2016/679 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation, GDPR), 27 April 2016.

30 At the time of writing, the Hellenic Police Service has denied access to these operating decisions even after requests for access have been issued by directly interested citizens.

31 See in particular ECtHR, L.A. and Others against Greece and A.A. against Greece, Applications nos. 12237/20 and 12736/20, lodged on 5 March 2020 and 7 March 2020.

32 Consider for example the action for damages brought against Frontex before the Court of Justice, alleging the illegal deportation of a Syrian family to Türkiye and the violation of their fundamental rights. According to Front-LEX, the organisation promoting the legal action, «despite undisputed and overwhelming evidence for serious and persisting violations of fundamental rights, FRONTEX and its Executive Director, Fabrice Leggeri, have failed to terminate the Agency’s activities in the Aegean Sea, in flagrant infringement of the EU Charter of Fundamental Rights, the Treaty on the Functioning of the EU and Frontex Regulation» (Prakken d’Oliveira Human Rights Lawyers 2021). Furthermore, in February 2020, the GHM lodged a complaint against the violent expulsions and lawlessness occurring at the sea borders (Pagoudis 2022).

3.1. Surveillance at the Pillars of Hercules and beyond: securitisation and digitalisation of Spanish border control policies

Non plus ultra: the mythological Pillars of Hercules, representing the Calpe and Abila mountains that flank the Strait of Gibraltar, were once considered the limits of the known world – the outer boundary, impossible to cross. Today, this same strait is one of the most monitored and securitized borders in Europe and embodies the violence that often marks the lines dividing the global North and South. However, despite the sophisticated systems in place to deter migration, people continue seeking to cross, echoing the folle volo – the daring journey beyond humanity’s unsurpassable threshold, famously depicted in the Divine Comedy1.

People on the move attempting to enter Spain typically follow three main corridors: the autonomous cities of Ceuta and Melilla, the Strait of Gibraltar, and the Canary Islands. Interestingly, over time, enhanced surveillance and border enforcement in one corridor have often led to increased pressure on the others, resulting in shifts in migration strategies rather than a reduction in the overall number of attempts to cross the border (Godenau and López-Sala 2016, 7).

This chapter focuses on the digitalisation of border surveillance in Spain, specifically through the implementation of the Sistema Integrado de Vigilancia Exterior (SIVE). The analysis begins by examining the digitalisation process within the broader context of migration routes to Spain, highlighting the militarization of Ceuta and Melilla, and the cooperative networks involved in monitoring the Strait of Gibraltar and the Canary Islands. This enquiry provides an in-depth look at the SIVE’s functionality and its multifaceted impact, concluding with an exploration of the legal framework within which the system is operated.

3.2. Deployment and maintenance: the costs of one of Europe’s largest surveillance systems

Evaluating the operational functioning of the Sistema Integrado de Vigilancia Exterior also requires outlining the financial resources, predominantly borne by European funds, required for the maintenance and replacement of its equipment, which now has been in use for over two decades.

As discussed in the Greek case study, the European Borders Fund (EBF) – covering the 2007-2014 financial period – was established to support national actions under five strategic priorities. Among these, the development and implementation of national components for a European Surveillance System at the external borders, as well as the establishment of a permanent European Patrol Network for the southern maritime borders, were particularly salient (European Commission 2014). As articulated in Spain’s national report on the EBF, the efficacy of the SIVE is intrinsically linked to the achievement of these priorities, thus justifying continued financial support (Spanish Ministry of the Interior 2012). Out of the total €630 million allocated to Member States under the EBF during the 2007-2010 period, Spain, Italy, and Greece, received nearly half, with Madrid alone obtaining approximately €134.5 million. This funding enabled Spain to procure a substantial number of surveillance components, primarily for the expansion of the SIVE. According to a European Commission report, these investments facilitated the interception of 5.279 migrant people within the specified financial period (European Commission 2014).

The 2013 national annual programme on the EBF provides one of the most comprehensive insights into the technical and operational life of the SIVE. In the programme, the Spanish Ministry of Interior emphasised the critical role of European financial support in maximising the system’s technological development. Moreover, it outlined the strategic importance of enhancing SIVE’s capabilities to advance the automation of border controls, thereby improving security while reducing operational costs and reliance on human resources (Spanish Ministry of the Interior 2013, 13). This emphasis on automation and the integration of AI into security protocols highlights the shifting priorities in border management.

Under the Internal Security Fund for 2014-2020, the digitalisation of border patrols in Spain was further accelerated through additional financial support, including from the Borders Emergency Assistance. This emergency funding, totalling €52 million in 2018 and 2019, was allocated to Spain, Greece, Hungary, Croatia, and Belgium to bolster border control efforts. In Spain, it once again primarily supported initiatives aimed at contrasting ‘illegal’ migration (European Commission 2021). Specifically, in 2018, over €6 million was dedicated to the maintenance and expansion of the SIVE, including repairs to its radars and sensors (Spanish State Secretariat for Security 2018). The investments continued over the following years, with approximately €3.6 million in 2019 and over €8.5 million in 2020 being allocated for similar purposes (Spanish State Secretariat for Security 2019). Furthermore, in 2020, an additional €1.4 million was earmarked to promote the digitalisation of borders and the development of automated smart borders (Spanish State Secretariat for Security 2020).

Tracking the precise allocation of funds within security-related projects is inherently challenging due to the opacity surrounding such expenditures. Nonetheless, a study by the Fundación PorCausa – a Spanish investigative foundation working on migration – offers revealing insights into the financial scale of migration control between 2007 and 2017. During this period, Spain spent approximately €896 million across 943 public contracts, with 97% of these funds aimed at enhancing border protection, surveillance, detention, and the expulsion of irregular migrants (PorCausa 2020a). A later analysis covering the 2014-2019 period shows that at least €660.4 million were awarded to companies involved in the contrast of migration through 1.677 public contracts, many of which raised transparency concerns7. Most recently, in March 2022, the Spanish Council of Ministers authorised the Secretary of State for Security to conclude a contract valued at €25.7 million for further upgrading the SIVE’s capabilities in Cadiz, Algeciras, and Ceuta (Carrasco 2022).

These figures illustrate a sustained and growing interest in advancing the digitalisation of border surveillance and patrolling systems, facilitated also by the allocation of emergency funds and contracts that often escape public scrutiny. Moreover, as evidenced in the case of Greece, EU funding mechanisms not only shape national policy orientations but also promote a convergence of approaches among Member States on border technologies.

3.3. The Sistema Integrado de Vigilancia Exterior in practice: situational awareness, detection, and risk analysis

Together with the investments it involves, the Sistema Integrado de Vigilancia Exterior carries significant expectations. Remarkably, the SIVE represents a prime and textbook example of digital patrolling systems as discussed in the first chapter, designed to enhance situational awareness and improve targeted detection along and beyond maritime borders. Essentially, the operational scope of the SIVE is closely aligned with the rationale underpinning the deployment of EUROSUR, which was indeed partially modelled on it (Ellebrecht 2020, 217).

Technically, the SIVE functions as an operational system that facilitates the surveillance of sea borders and surrounding areas. It shares real-time information with control centres through a network of fixed stations and mobile units equipped with still cameras, heartbeat detectors, CCTV cameras, night vision devices, infrared optics, long-range radar systems, and thermal cameras positioned along Spain’s coastal areas (Jumbert 2018, 16; Spanish Ministry of the Interior 2012). When weather conditions are favourable, the SIVE detects migrant boats within a range of 10 to 25 kilometres from the shore, capturing high-quality photographs and videos. Once a target is detected, an alert is sent to the Centro de Mando y Control in Algeciras, which monitors the situation remotely. Smaller control centres are also located in Cadiz, Malaga, and Ceuta, though the exact locations of the SIVE stations remain classified (González 2018).

As a boat approaches within 5 kilometres of the coast, the control centre can estimate the number of passengers, the vessel’s course, and its expected time of arrival. This information, gathered from multiple sources, is fused, processed based on risk criteria, and then relayed to patrol units (such as helicopters, boats, or other vehicles) or other entities, such as the Maritime Rescue, the Red Cross, or the National Police Corps. As previously mentioned, the entire system is operated by the Guardia Civil, a joint military and civilian police force responsible for the Coast and Border Service and the Maritime Service (Catalán 2014).

The primary objective of the SIVE is to efficiently detect and apprehend individuals attempting to enter Spain, blending a deterrence-based approach with a discourse that encompasses both trafficking and humanitarian concerns (Carling 2007). More precisely, according to the Spanish Ministry of the Interior (2013, 11), the purpose of the SIVE is to ensure coverage of the European Union’s southern border, thereby enhancing the Guardia Civil’s effectiveness in performing its duties.

Despite the focus on irregular immigration and drug trafficking being its primary objectives, the SIVE is often praised for its versatility. The system is in fact also aimed at combating terrorism, intelligence gathering, countering illegal fishing and piracy, protecting marine and land resources, conducting search and rescue operations, and ensuring port security (Fernández Jurado and Sabariego Rivero 2006). Over time, the system’s operational scope has thus expanded considerably.

Also in geographic terms, SIVE’s coverage has grown since its initial deployment in 2002 around the Strait of Gibraltar. It now spans the entire Spanish Mediterranean coast, the Balearic Islands, and the Canary Islands. The system’s expansion began with three fixed detection stations in Fuerteventura and later extended also to Lanzarote, Gran Canarias, and the Atlantic coasts (European Union Agency for Fundamental Rights 2013, 59). Beyond Spain, just like the EUROSUR, SIVE’s operational reach has extended also to non-European countries, reflecting externalisation ambitions. Since 2006, coordination centres have been established in Mauritania, Senegal, Guinea, and Cape Verde to foster interoperability and information exchange in pre-frontier areas, aligning with broader European policy objectives often supported by Frontex (Markard 2016, 612).

Interestingly, the continuous expansion of the SIVE has often led to the deflection (and not suppression) of migration routes, resulting in a ‘cat-and-mouse’ game that does not offer a definitive solution to migration governance challenges8. Migrant people have increasingly resorted to longer, more hazardous routes, away from both border surveillance and rescue systems (Spanish Refugee Aid Commission 2017).

3.4. Drones buzzing around the SIVE?

Determining whether the fixed radar and camera systems that form the backbone of the SIVE are currently complemented by the deployment of drones remains challenging, as has been the case studying the Greek context. However, evidence suggests that Spain is increasingly using UAVs for various purposes along its borders, and it is reasonable to assume that such operations might be integrated into the SIVE to enhance its effectiveness.

In fact, Spain has extensively used the drones provided by the European Maritime Safety Agency primarily for pollution control, search and rescue operations, and maritime surveillance purposes. These functions partially overlap with the objectives of the SIVE, suggesting a potential convergence of efforts (European Maritime Safety Agency 2019; 2022). Additionally, an infographic produced by the Guardia Civil and made public by Maribel Casas-Cortes, Sebastian Cobarrubias, and John Pickles (2016, 7) indicates that UAVs were deployed under the SIVE framework as early as 2008. However, specific details about the number of drones, their areas of deployment, and how they were integrated into the system remain unclear.

Spain’s involvement in several EU-funded drone projects under the 7th Framework Programme for Research further highlights the influence of EU funding and research schemes on national border policies. Notable here is the CLOSEYE project, which ended in 2017, aimed at fostering cooperation between Guardia Civil and Frontex personnel in this area9. The project involved extensive UAV testing to monitor European maritime borders, strengthen the SIVE, and improve its integration into the European Border Surveillance System (Marin 2017; Statewatch 2013).

Attempts have also been made to deploy drones over the Spanish enclaves of Ceuta and Melilla. Already in 2017, drones were in fact integrated into an artificial intelligence surveillance system called Surveiron. However, this initiative was quickly abandoned due to environmental factors, particularly the strong Levante wind, which made drone operations in the area impractical (Andres 2021; Testa 2017). Despite this setback, the Guardia Civil expressed renewed interest in integrating drones into the SIVE network in 2020. In February 2022, the Guardia Civil made an emergency purchase of four UAVs to enhance border surveillance in Ceuta and Melilla, aiming to detect crossing attempts at an earlier stage while mitigating risks for human personnel in cases of violent «border assaults» (Infodron.es 2022). Interestingly, this purchase was made under an emergency tendering model used during the Covid-19 pandemic, regulated by Article 120 of Law 9/2017 on Public Sector Contracts, which allows for rapid procurement in situations affecting national defence10.

At approximately the same time, Morocco strengthened its border monitoring capabilities in Ceuta and Melilla with the acquisition of 12 military drones from Türkiye (Drusila Castro 2022). Not surprisingly, the Spanish government responded positively, emphasizing that border enforcement decisions are a matter of State sovereignty on which no interferences should take place, and reaffirming its commitment to cooperating with Morocco for mutual benefit in this area.

These developments reflect a broader trend towards the integration of dual-use unmanned aircraft in border surveillance, with Spain taking a leading role. Indeed, Madrid is also spearheading the Next Generation Small RPAS project under the EU’s Permanent Structured Cooperation (PESCO), which aims to develop Remotely Piloted Aircraft Systems to enhance military cooperation among EU Member States (PESCO 2021). These multi-purpose RPAS are expected to support not only military operations but also law enforcement and disaster or emergency response agencies. The possibility of equipping these drones with less-than-lethal effectors for police use indicates potential implications for future border patrol strategies (Statewatch 2021). While it is premature to draw definitive conclusions about the impact of these developments on border patrols, Spain’s growing leadership in drone technology should not be overlooked.

3.5. The legal framework: between pushbacks, the SIVE, and drones

Once again, analysing the legal framework surrounding the Sistema Integrado de Vigilancia Exterior and the deployment of drones under its umbrella is a complex challenge. This complexity arises primarily from the lack of a specific legal basis for the SIVE, which can only be traced back to the general mandate of the Guardia Civil (Ranger Project 2018). Thus, it is crucial to broadly explore the main concerns related to both access to international protection and border control in Spain, focusing on large-scale pushbacks as a particularly alarming and compelling phenomenon. Pushbacks and interceptions, just as the attempts to give them legal legitimacy, can in fact be facilitated by the SIVE, marking a point of concern in the overall digitalisation of border patrols.

To set the scene, over the past decade, Spain has transitioned from being one of the European countries with the lowest number of international protection applications to one of the Member States hosting the largest number of asylum seekers, partly due to the Dublin Regulation (Regulation 604/2013) (Oso, López-Sala, and Muñoz-Comet 2021)11. However, this increase has not translated into easier access to international protection for people on the move arriving via Spain’s southern border. Scholars have here identified a phenomenon of «territorial deviation», whereby the majority of asylum applications are submitted ‘within’ Spain’s territory rather than ‘at the border’ (López-Sala and Moreno-Amador 2020). Furthermore, while the securitisation and militarization of borders are significant factors, political narratives have also played a role in framing people on the move arriving at the southern border primarily as economic migrants. This racialised and largely neocolonial discourse casts them as ‘bogus’ asylum seekers who should be deterred or repelled. Consequently, people on the move from African countries face substantial, systemic obstacles in reaching Spanish borders and accessing protection (BVMN 2021, 5).

In this context, a landmark development in Spanish migration and border policies occurred during the 2008 economic crisis, described as a pivotal moment in consolidating privatisation, outsourcing, remote control, and technologisation of border control patterns (Godenau and López-Sala 2016; López-Sala and Godenau 2020; Oso, López-Sala, and Muñoz-Comet 2021). These processes have been coupled with overt ‘migratory containment’ strategies followed, in 2009, by the introduction of accelerated asylum procedures at the southern border. These procedures, which take place in closed centres, have been criticised for leading to selective immobilisation of people, thus violating the principle of equal treatment while lowering procedural and substantial guarantees, therefore producing an «infrastructural vacuum of access to refugee status» (López-Sala and Moreno-Amador 2020, 12). Similarly, the institutionalisation of Temporary Stay Centres for Foreigners (Centros de Atención Temporal de Extranjeros) – fenced buildings or secured areas within ports where migrant people are held for up to 72 hours while police procedures are carried out – further jeopardises access to a fair examination of individual claims. Overall, these centres have, in fact, proven efficient in acting as barriers to protection more than in expediting related procedures (Barbero 2021)12.

Further barriers to international protection result from the harsh pushback operations documented at the borders, particularly around the fences of Ceuta and Melilla. One of the most tragic incidents that significantly shaped Spain’s approach to border security was the 2005 ‘Asalto Masivo’, where thousands of migrant people attempted to enter the enclaves. The ensuing violence, with police forces (Spanish or Moroccan, never conclusively determined) opening fire, resulted in at least 14 deaths (Carlotti 2022). This incident marked a critical moment in a ‘defensive’ turn in Spain’s border policy, further reinforcing a narrative heavily skewed towards securitization and the externalization of migration management.

A particularly striking episode occurred in May 2021, when numerous sub-Saharan migrant people scaled the border fence in protest against the total closure of the borders, justified under the pandemic emergency. Spanish Guardia Civil agents, equipped with riot gear, quickly made their efforts futile. The threat of immediate repatriation was compounded by the use of tear gas and plastic bullets, aimed specifically at the hands and neck of people attempting to cross13. Notably, no one was allowed to claim protection.

According to monitoring organisations, border operations of this kind have increased both in violence and frequency over the past two years. These large-scale incidents, involving hundreds of people, are marked by their indiscriminate nature, affecting even unaccompanied minors (BVMN 2021, 19). A notable instance in this regard occurred in August 2021, when the Spanish government started returning children to Morocco under a contentious readmission agreement. This action constituted a blatant violation of Article 35 of Organic Law 4/2000 on the Rights and Freedoms of Foreigners in Spain, as well as European and international obligations14. Despite the Spanish human rights Ombudsperson’s (Defensor del Pueblo 2021) objections, the repatriation of unaccompanied minors continued for months, until it was finally halted by a local court’s intervention (Testa and Sánchez 2021).

Moreover, allegations of severe mistreatment by Spanish border guards within the enclaves have been extensively documented, with some cases making their way before the European Court of Human Rights. In April 2021, a letter from four UN Special Rapporteurs addressed to the Spanish government expressed concern over pushback practices towards Morocco, notably in the case of indiscriminate and summary expulsions at the border (UN Office of the High Commissioner for Human Rights 2021). Additionally, there are also reports of ‘chain refoulement’, where individuals expelled from Spain are further deported from Morocco to Mauritania and beyond (BVMN 2021, 12). These practices, often amounting to ‘hot returns’ (devoluciones en caliente), are predominantly registered in Ceuta and Melilla (Costa Traba 2021).

Within the enclaves, foreign nationals intercepted by the Guardia Civil are in fact immediately handed over to Moroccan authorities, without any formal procedure. The Spanish government has attempted to legally justify these actions through the concept of the ‘operational border’ (frontera operativa), which suggests that the border is defined by the position of the Guardia Civil agents and not by the ‘real’ border, almost regardless of geographical considerations. This concept was first articulated in a report by the Deputy Operational Directorate of the Guardia Civil, submitted to the Ministry of the Interior and presented to the Congress of Deputies on 7 March 2014. The report asserts that «the internal fence materialises the line with which the State, in a free and sovereign decision, delimits, for the sole purposes of the aliens’ regime, the national territory». According to this logic, migrant people only enter Spanish territory – and thus become subject to Spanish immigration laws – once they cross the internal fence (Martínez Escamilla and Sánchez Tomás 2019, 31). This interpretation effectively denies foreign nationals handed over to Moroccan authorities any protection afforded by Spanish law (Defensor del Pueblo 2005, 292). Of course, this rather imaginative approach is legally questionable, as the external fence separating Ceuta and Melilla from Morocco is largely situated on Spanish territory.

In 2015, the Spanish government further entrenched these practices by amending Organic Law no. 4/2000 (Ley de Extranjeria) with the introduction of the Tenth Additional Provision under the Law on Citizen Security (Ley de Seguridad Ciudadana). This provision allows for the routine expulsion of foreign nationals intercepted at the Ceuta and Melilla borders back to Morocco, citing a special regime referring to the substantial exceptionality of these cities. Furthermore, this provision mandates that applications for international protection must be submitted at designated border crossing points, where the Asylum Offices (Oficinas de Asilo) are located. While people of Moroccan, Algerian, and Syrian nationality can generally access these points, people on the move of sub-Saharan origin face significant barriers. The process involves in fact a dual triage by both Moroccan and Spanish authorities, based on nationality and deeply racializing procedures. In 2018, only six people from Burkina Faso, Guinea, and Mauritania were able to request asylum at the Melilla border crossing point, compared to the 290 who applied once inside Spanish territory (Costa Traba 2021). Consequently, sub-Saharan migrant people are often compelled to take extreme risks, such as scaling the border fence, to access Spanish territory (Amnesty International Spain 2016, 9–10; Commissioner for Human Rights 2015).

The adoption of the Tenth Additional Provision thus raised significant concerns, voiced also by the Council of Europe’s Commissioner for Human Rights (Muižnieks 2014). Although this provision stipulates that ‘hot returns’ must adhere to international human rights standards and protection laws, this requirement is fraught with controversy and contradictions, and no specific procedures are defined to ensure compliance.

Regrettably, the European Court of Human Rights’ jurisprudence on this matter also remains contentious. Notably, the Court’s ruling on the case of N.D. and N.T. v. Spain – initially decided on 3 October 2017 and later overturned by the Grand Chamber on 13 February 2020 – illustrates the deep tensions surrounding this issue15. In its 2017 decision, the Court found that the applicants, who had attempted to scale the fences in Melilla, were forcibly removed and returned to Morocco while under the exclusive and constant control of Spanish authorities. However, the Grand Chamber’s 2020 ruling concluded that Spain had not violated the prohibition of collective expulsions, attributing the outcome of the operation to the applicants’ «culpable conduct». This ruling provoked sharp criticism, as it appears to legitimise the indiscriminate pushback of migrant people if regular border crossing at specific points is made possible. However, the decision does not undertake a throughout examination of the real accessibility of such crossing points (Markard 2020). Concurrently, this approach risks endorsing non-entrée policies and effectively legitimising the establishment of a ‘no man’s land’ around the enclaves (Sardo 2021).

These developments illuminate how political choices and legal interpretations are converging to mask the violence inherent in border enforcement. This convergence is also evident in the increasing digitalisation of border patrols, as exemplified by the SIVE, which can enhance pushbacks while simultaneously reducing their visibility.

As previously mentioned, the lack of a clear legal basis for the SIVE complicates any in-depth legal analysis. Additionally, according to data collected by the European Union Agency for Fundamental Rights in 2013, Spanish authorities claimed that the SIVE does not involve the collection or processing of personal data. However, as for the Greek case, in the absence of further evidence, this is not entirely reassuring. The Fundamental Rights Agency noted in fact that the ability of the SIVE to identify individuals depends on factors such as distance, lighting, and weather conditions, indicating that the processing of personal data remains technically feasible (European Union Agency for Fundamental Rights 2013). Thus, despite the absence of updated information regarding the data collected through radars, cameras, and data-sharing networks, it remains challenging to definitively determine whether personal data is processed by the Spanish surveillance system.

Furthermore, assuming that the SIVE can incorporate data from drones, it is essential to have a look at the national legislation governing the use of UAVs. Notably, Spain was one of the first European countries to introduce drone regulations, with the passage of Law 18/2014, which sought to address a significant gap in the legal framework (Pauner Chulvi 2016)16. However, this law provided only partial and incomplete regulations, later addressed by the Royal Decree 1036/2017, which specifically covers drones within the broader framework of the Air Traffic Regulations (Reglamento de Circulación Aérea)17. The national regime was later replaced on 1 January 2021 by the provisions of EU Regulation 2019/94718.

In terms of the professional and commercial use of drones, the 2014 Law permits a wide range of activities, including research and development, firefighting, aerial observation and surveillance, and search and rescue operations. Article 50 of Law 18/2014 stipulates that drone operators are responsible for all activities conducted with UAVs, including compliance with personal data protection obligations as outlined in Organic Law 1/1982 on the right to personal and family privacy and one’s own image19, Organic Law 15/1999 on data protection20, and Organic Law 3/2019 on the protection of personal data with regards to digital rights21. Law 15/1999 defines personal data in accordance with the relevant European Directive, applying this definition to any images, sounds, or voices captured by drones.

Concerning the use of drones in public spaces by law enforcement agencies (Fuerzas y Cuerpos de Seguridad), Organic Law 4/1997 provides provisions for mobile video surveillance systems, although not explicitly mention drones22. It permits the use of these systems to prevent crimes and maintain public security, thus potentially applying in border settings. However, any use of drones by police forces must adhere to the principle of proportionality, both in terms of appropriateness and minimal intervention (Article 6, Law 18/2014) (Pauner Chulvi 2016, 97). Moreover, the Spanish Data Protection Agency (Agencia Española de Protección de Datos, AEPD) (2019b) has repeatedly emphasized the importance of conducting risk and data protection assessments before deploying drones, although there is no binding obligation in this sense. Notably, there is still no evidence suggesting that risk analyses are routinely performed before digital patrolling operations are initiated.

Ultimately, the use of drones by State surveillance forces must not subordinate individual rights to the principle of security, as this could exert a deterrent effect that undermines the full enjoyment of these rights – a concern highlighted in chapter 1 regarding the disciplinary effects of surveillance. However, this deterrent outcome is precisely the aim of the deployment of drones and systems such as the SIVE for border control.

Regarding the collection of data via drones, the AEPD labels drone operations into three categories: those not involving personal data processing (typically recreational or domestic use), those with a risk of inadvertent data processing, and those specifically intended for data processing (Spanish Data Protection Agency 2019a). Border surveillance activities likely fall within the two latter groups, which are thus subject to both the General Data Protection Regulation and Organic Law 15/1999. The AEPD also advocates for the implementation of privacy-by-design features and adhering to the guidelines for video surveillance, regardless of whether the systems are fixed, as with the SIVE, or mobile, as with drones (Spanish Data Protection Agency 2019c). This is particularly pertinent concerning the southern frontera inteligente, where the use of facial recognition technologies is explicitly considered. The AEPD has also underscored the necessity of applying existing legislation not only to data collection but throughout all stages of data processing. Any drone-related activity must thus comply with relevant laws, including video surveillance regulations and fundamental rights legislation (Spanish Data Protection Agency 2019d). Of course, these rights encompass not only data protection but also the right to claim asylum and the prohibition of refoulement.

In conclusion, the digitalisation of border patrols in Spain is being integrated into a regulatory and political framework that is frequently hostile to people on the move, often leading to discriminatory outcomes. Furthermore, the rapid evolution and integration of border surveillance technologies – facilitated by cooperation with third countries – has not been paralleled by a corresponding development of protection frameworks. The current legal framework governing the use of drones remains ambiguous and underdeveloped, particularly in the context of border enforcement, also due to the scarcity of information on the data thus collected.

Despite the high maintenance costs, the SIVE has not replaced traditional patrolling methods but rather seems to facilitate them, also in cases of externalisation agreements and violent pushback operations. Moreover, the limited contributions of the SIVE to sea rescue operations do not mitigate the System’s role in endangering migratory routes and enabling push and pullbacks in coordination with Moroccan authorities, often marked by discriminatory stances. Today, the journey across the ‘inviolable sea’ of the Pillars of Hercules thus remains particularly perilous.

1 Dante Alighieri, ‘La Divina Commedia’, Inferno, Canto XXVI, v.125.

2 In Spain, the per capita income adjusted for the cost of living is approximately five times higher than that estimated in Morocco. Although this parameter has clear limitations, as it does not account for factors such as the welfare state, the rule of law, or guarantees of individual freedom when assessing the quality of life, the stark economic disparity remains relevant.

3 It should also be noted that until 1991 citizens of North African countries entering Spain were not subjected to visa requirements.

4 More recent aggregate data could not be retrieved.

5 Law no. 02/03, Entry and stay of foreigners into the Kingdom of Morocco, irregular emigration and immigration, 20 November 2003.

6 Association Council, UE-MA Action Plan 2702/1/05, 27 July 2005, para. 48.

7 All these contracts are collected and published by porCausa at https://docs.google.com/spreadsheets/d/1mEOHIKwFyGfiha5GJ0HkTpc08mDZc-I6hGpUFf2oDno/edit#gid=1367307146

8 To provide an example, according to the Asociación Pro Derechos Humanos de Andalucía (APDHA) (2022, 52), the decrease in the arrival of migrants on the coast of Cádiz registered between 2002 and 2005 was almost totally compensated by the increase in the arrivals that took place in the years 2005-2007 in the Canary Islands.

9 Collaborative evaluation of border surveillance technologies in maritime environment by pre-operational validation of innovative solutions (CLOSEYE), Grant agreement ID: 313184. 1 April 2013 – 28 February 2017.

10 Law no. 9/2017, Contracts of the Public Sector, 8 November 2017.

11 Regulation no. 604/2013 establishing the criteria and mechanisms for determining the Member State responsible for examining an application for international protection lodged in one of the Member States by a third-country national or a stateless person (Dublin Regulation), 26 June 2013.

12 See the Asylum Information Database (2021) for more information on these facilities and the related procedures in the centres.

13 A visual documentation of this violent episode has been published by the Border Violence Monitoring Network: https://www.borderviolence.eu/wp-content/uploads/CEUTA1-general-and-shootings.mp4

14 Organic Law no. 4/2000, On Rights and Freedoms of Foreigners in Spain and their social integration, 11 January.

15 ECtHR, N.D. and N.T. v. Spain, Applications nos. 8675/15 and 8697/15, 13 February 2020.

16 Law no. 18/2014, On the approval of urgent measures for growth, competitiveness and efficiency, 15 October 2014.

17 Royal Decree no. 1036/2017, Governing the civil use of remotely piloted aircraft, 15 December 2017.

18 Commission Implementing Regulation no. 2019/947 on the rules and procedures for the operation of unmanned aircraft, 24 May 2019.

19 Organic Law no. 1/1982, On the Civil Protection of the Right to Honour, Personal Privacy and Self-Image, 5 May 1982.

20 Organic Law no. 15/1999, On Protection of Personal Data, 13 December 1999.

21 Organic Law no. 3/2018, On Protection of Personal Data and Guarantee of Digital Rights, 5 December 2018.

22 Organic Law no. 4/1997, Regulating the Use of Video Cameras by Security Forces and Units in Public Spaces, 4 August 1997.

4.1. From the eastern to the western route: a comparative analysis of digital patrolling

By focusing on the case studies of Greece and Spain, this research has attempted to undertake a comprehensive analysis of the factual realities and legal frameworks surrounding the digitalisation of border patrols. These cases are significant due to their political, geographical, and legal dimensions, where remarkable technological experimentation in border areas is embedded, thereby enabling a comparative analysis of the strategies employed by EU Member States along the eastern and western migratory routes to Europe. Such a comparative approach is particularly effective in revealing processes, patterns, and concerns that extend beyond the specific cases of Greece and Spain, potentially applying – at least in part – to other settings and sites around the EU.

This chapter extends the comparative analysis to explore the wider ramifications of digital border patrolling, from local to national, EU, and international levels. Specifically, it critically examines issues surrounding human dignity and the right to international protection, particularly concerning the principle of non-refoulement, alongside the right to privacy and data protection. These discussions are framed within the context of recent developments on the regulation of interoperability and Artificial Intelligence. Additionally, the chapter addresses how the principle of non-discrimination is increasingly challenged by the digitalisation of patrolling, eventually raising further questions on accountability and jurisdictional matters.

To navigate these multiple layers, a ‘topographical approach’ is employed in order to read through and interrogate the most significant fallouts of digital patrolling in Greece and Spain. As articulated by Nikolas Feith Tan and Thomas Gammeltoft-Hansen (2020), this perspective advocates for a bird’s-eye view across various legal regimes, overlapping frameworks of liability, and geographical contexts. The overarching aim is thus to draw on the different border segments analysed to derive interpretive insights that shed light on the pervasive and rapidly evolving dynamics around digital patrolling. By comparing the two case studies, this section reflects on the significance of the specific contexts in which digitalisation processes unfold, the consequences of the general lack of transparency in border management, and the risks posed to people on the move by the inadequacy of current safeguards in digital border patrolling practices.

4.2. Technology and the law: rights at stake in digital patrolling operations

The digitalisation of patrolling profoundly affects the rights of people in patrolled areas and challenges the law’s ability to provide an adequate bulwark of protection against potential violations pouring from legal black holes, secrecy imperatives, and the difficulties in regulating rapidly evolving systems. Attempting to assess the implications of these developments in border surveillance on people’s lives and fundamental rights is a complex endeavour, but one of significant interest and importance.

As discussed, new technologies can introduce new risks, inequalities, and unforeseen consequences that turn oversight attempts concerning both their deployment and the use of the data thus collected into a complex endeavour (Dijstelbloem, Meijer, and Besters 2011, 15). This directly impacts human dignity and raises significant concerns regarding equality, data protection, and access to justice, especially when dealing with systems based on AI and with varying degrees of automation (European Union Agency for Fundamental Rights 2020). Currently, there is a notable absence or inadequacy of regulatory frameworks specifically addressing automation in the context of borders and migration management, notably concerning accountability schemes and protection mechanisms.

Given these challenges, it is helpful to embrace a framework that integrates and patches various legal regimes from a comparative (and bird’s-eye) perspective, moving from a fundamental rights-centred approach. The goal is thus to go beyond the specific considerations introduced in earlier chapters on Greece and Spain, adopting a more holistic human rights protection strategy when addressing the rights of people on the move at the EU’s external borders (see Beduschi 2022).

To this end, zooming out from national case studies to examine International Human Rights Law (IHRL) perspectives, European fundamental rights regimes, and migration law insights can be promising. This analysis will focus on four ‘pillars’ – both human rights and core principles – of the fundamental rights system: human dignity, access to protection and asylum, privacy and data protection, and equality and non-discrimination. As emerged from the Greek and Spanish case studies, these areas are particularly ductile to digital patrolling.

4.3. Human dignity in times of de-humanised surveillance

According to Özgün E. Topak (2021, 796), border violence should be broadly understood as «the entire set of processes whereby migrants’ somatic and mental capacities are repressed or destroyed at and beyond the territorial border». Building on a similar understanding, Claudia Aradau and Lucrezia Canzutti (2022) introduce the concept of «technologies of cruelty» to describe the processes of objectification and dehumanization crystallized in asylum politics when driven by the assumption that people seeking protection are inherently ‘bogus’. These governing strategies are mirrored also in digital border spaces, in a shift that threatens the full respect of human dignity, a principle that can be described as underpinning all considerations on fundamental rights.

In recent years, legal scholars have engaged in various debates about the definition and scope of the concept of human dignity. This renewed interest stems from the recognition that, due to its interpretative versatility, human dignity can be a solid yet flexible framework for addressing emerging and complex socio-legal challenges (see Fernández Burgueño 2016). Without delving into the details of this often thorny debate, it should be noted that many scholars view human dignity as a «mother-right» (Barak 2015) that serves as a foundational ‘framework right’.

In International Law, human dignity has been embedded in key human rights treaties since the 1948 Universal Declaration of Human Rights (UDHR)7. However, the interpretation and application of human dignity vary significantly across different international jurisdictions, making it challenging (and not particularly desirable) to unify these diverse approaches8. Accordingly, the notion of human dignity has been integrated into national and regional legislation in a highly differentiated way.

In Europe, Article 1 of the European Convention on Human Rights (ECHR) enshrines human dignity as the inviolable foundation of fundamental rights. The Court of Justice of the European Union has also repeatedly affirmed that human dignity is an integral part and parcel of EU law9.

Hans Jörg Sandkühler (2015) argues that the notion of human dignity becomes particularly influential in times marked by processes of societal dehumanization. This perspective seems indeed relevant in the context of digital patrolling, as it re-centres the debate on people on the move. These considerations pave the way to a deeper understanding of the implications of digital patrolling so far discussed, suggesting that a focus on human dignity could help to address some of the regulatory gaps that emerge with the smartening of borders.

As discussed, one of the most concerning aspects of digital patrolling is the shift in how surveillance is conducted. Patrolling personnel are increasingly removed from borderlines, operating from screens-cluttered coordination centres, while digital patrolling equipment is deployed closer to the people being monitored, also in international waters and third countries’ territory.

Through the deployment of UAVs and the use of advanced surveillance systems, the avoidance and the impossibility of a direct ‘encounter’ between migrant people and actors engaged in patrol activities is often used to justify evading international obligations, notably in cases of distress at sea. Moreover, by preventing people on the move from reaching borders, States distance themselves from the responsibility of assessing asylum and international protection claims (Laursen 2022). The Commissioner for Human Rights of the Council of Europe (2021), Dunja Mijatović, recently pointed out the shift to aerial surveillance, especially in the Central Mediterranean, as evidence of European States’ unwillingness to establish adequate protection systems.

Moreover, the digitalisation of patrolling shifts the discourse around border security towards a more impersonal and abstract direction, characterised by one-way interactions that reduce migrant people to ‘security objects’ to be ‘managed’ through advanced technological tools. This produces a deliberate distancing of people attempting to reach the external borders, leading to a dehumanisation of border security that closely parallels the dynamics observed in the deployment of automated systems in warfare (Wall and Monahan 2011).

While it is sometimes argued that digital patrolling could make border control more ‘humane’ and less prone to arbitrary violence, this claim is contestable for several reasons. First, advanced surveillance systems rarely replace physical barriers but often complement them. This is evident, for example, both in the Evros region, along the wall remotely surveilled, and in Ceuta, where the triple fence is reinforced by the frontera inteligente. Secondly, in the crevices between the dehumanisation of patrolling and the objectification of people attempting to reach European borders, violence has been shown to even be more likely (Molnar 2020).

Moreover, as observed in both Greece and Spain, digital patrolling often compels people to take more dangerous routes. On this note, a particularly striking example comes from the other side of the Atlantic, where the Biden Administration has opted for new technologically advanced (and ostensibly more ‘humane’) monitoring strategies at the border with Mexico, aiming to take a distance from the brutality of the wall vehemently advocated during the Trump Presidency. However, it has been shown that this more politically palatable approach – which has led to the construction of Integrated Fixed Towers and the deployment of robotic patrol dogs – has driven migration routes through the Arizona desert, resulting in the tripling of deaths at the border (Chambers et al. 2021).

Overall, the coexistence of digital patrolling with increased direct and indirect border violence can be deemed well-established. Moreover, the use of digital tools in border patrolling is often justified by their effectiveness in performing “3D-tasks” – those considered ‘dull, dirty, and dangerous’ and thus particularly challenging and unpleasant. However, deploying drones for these purposes often leads to increased ‘recklessness’ in patrolling activities (Val Garijo 2020, 138), driven by the assumption that what is technically feasible is also legally, socially, and ethically acceptable.

In an influential article, Luisa Marin (2017) traced the dehumanization of border surveillance through drones and other digital patrolling systems to two main issues. The first concerns the ability of drones deployed in dull and dangerous settings to differentiate between scenarios requiring different legal responses. In theory, drones could in fact detect boats in distress, potentially triggering adequate interventions and reporting cases where medical assistance is needed. In this direction, Mark Coeckelbergh (2013) argues that advanced surveillance technologies might bridge the epistemic gap caused by the ‘remoteness’ of drone patrols, at least partially balancing the dehumanization of surveillance by enabling more targeted human interventions. In practice, however, it is rather unclear whether current surveillance systems, including those using UAVs, are designed to detect and collect data on distress circumstances or the conditions of vulnerability among people approaching the borders. Moreover, while the fascinating debate surrounding this issue lies well beyond the scope of this study, critical scholars working on data and AI have expressed doubts about the capacity of automated systems embedded in border patrols to fully (or better) ‘comprehend’ situations that prompt international obligations (see Hildebrandt 2020).

The second issue raised by Marin’s analysis – in line with the study of bordering practices in Greece and Spain – further shows that the narrative of digitalisation as a humanitarian measure, ostensibly aimed at saving lives, has little empirical support. Instead, the latter suggests that digital surveillance aligns with securitarian and preventive priorities rather than protection. This discrepancy underscores a key point here: increased situational awareness through digital means simply does not necessarily lead to better assistance for migrant people in distress.

As mentioned, another dynamic fostered by the digitalisation of patrolling is the growing standardization of decision-making processes across various operational levels, from risk assessment to interception operations, reducing the space for individual case analysis. As it has been discussed, while standardization might limit discretionary and arbitrary decisions10, it also challenges the international protection system, which is fundamentally based on the assessment of particular cases and individual conditions (Dijstelbloem, Meijer, and Besters 2011, 15). These circumstances raise an inevitable question: can human dignity, with all its complexities, be fully accounted for within an increasingly standardized framework?

In conclusion, the dynamics of digitalisation in patrolling – resulting in dehumanization, dilated distances, and the objectification of migrant people as «indistinct, pixelated, and vaguely humanoid shapes»11– risk creating conditions that severely limit the recognition of human dignity for those attempting to reach European borders (and pre-border areas). However, this is not definitive. Future developments, ideally also grounded in legal advancements rather than merely technological ones, could potentially address some of these challenges – at least partially – in the (now, quite unlikely) event that the EU border landscape undergoes significant change.

4.4. The right to seek protection: non-refoulement between access to asylum and border surveillance

As discussed, the digitalisation of patrolling, when embedded in migration containment and deterrence strategies, often manifests in violent border control measures designed to prevent access to European borders. As explored in chapters 2 and 3, the convergence of advanced patrolling technologies, securitizing policies, and vague legal frameworks raises significant human rights concerns, notably regarding the obligations stemming from the principle of non-refoulement and the right to seek asylum (Val Garijo 2020). Pushback operations, a widespread and alarming practice at the EU’s external borders, exemplify these concerns12. Pushbacks not only have immediate and severe consequences for potential asylum seekers but also often result in the perpetration of extremely violent operations against people on the move, regardless of their intentions to seek protection. Although the primary concern here is the inability to access asylum, it is essential to avoid reinforcing the often biased and oppositional distinction that Rebecca Hamlin (2021) refers to as the «migrant/refugee binary». If the boundaries of this binary are frequently contested and marked by arbitrariness, yet at and before the border, such divisions tend to fade into irrelevance.

As already discussed, the indiscriminate prevention of access to protection in Europe is not exclusive to the digitalisation of patrolling, but a glaring outcome of border policies centred on externalization, containment, and pushback (or pullback) operations. Moving forward, this section presents a brief overview of the legal framework designed to counteract the barriers preventing the lodging of protection applications.

At the international level, the right to seek and enjoy asylum is enshrined in Article 14 of the UDHR and further regulated by the 1951 Refugee Convention and its 1967 Optional Protocol13. The right to seek asylum is obviously dependent on the principle of non-refoulement, which ensures that no individual is rejected without a thorough analysis of their situation. As anticipated in the previous chapters, non-refoulement, as established by the 1951 Convention (Article 33), prohibits States from removing, expelling, or extraditing individuals to a country where they risk facing the death penalty, torture, or other inhuman or degrading treatment. Thomas Gammeltoft-Hansen (2013, 44) describes it as the «strongest commitment» that the international community has made to protect those who are no longer able to avail themselves of the protection of their own State. Recognized as ius cogens and a peremptory norm of international law, non-refoulement is thus the primary obligation that States have to fulfil with regards to people in need of protection and the cornerstone of the whole international asylum regime – which would otherwise result in being empty and futile (Giuffré and Moreno-Lax 2019; Simeon 2019; Trevisanut 2014).

The non-refoulement principle is sanctioned by several other key international conventions, including the 1984 Convention against Torture (CAT, Article 3) and the 1966 International Covenant on Civil and Political Rights (ICCPR), deriving from the right to life (Article 6) and the prohibition of torture (Article 7)14. At the European Union level, the right to seek asylum is guaranteed inter alia by Article 18 of the EU Charter of Fundamental Rights (CFR)15, while non-refoulement and the prohibition of collective expulsions are codified in Article 19 of the CFR and Article 4 of Protocol No. 4 of the ECHR. Furthermore, several EU Directives and Regulations, such as the Qualification Directive16, the Asylum Procedure Directive17, the Return Directive18, and the Schengen Borders Code19, mirror these international human rights obligations.

Moreover, under the ECHR, non-refoulement is an absolute obligation20, directly linked to the prohibition of arbitrary deprivation of life (Article 2) and the prohibition of torture (Article 3). The European Court of Human Rights has upheld the applicability of non-refoulement in cases of pushbacks, non-admissions, and rejections both at sea and at territorial borders, as seen in landmark cases such as Soering v. UK21, Hirsi Jamaa v. Italy22, and N.D. and N.T. v. Spain23. These rulings emphasize the centrality of the rights of individuals in the protection system and reject the creation of «areas outside the law» in border zones24.

Despite this comprehensive legal framework, violations of non-refoulement continue to be a major concern at the EU’s borders, often involving violent pushbacks or forms of ‘cooperative deterrence’ with third States, described in the literature as cases of «neo-refoulement» (Hathaway and Gammeltoft-Hansen 2014; Hyndman and Mountz 2008). As discussed, agreements with Türkiye and Morocco have often resulted in practices that effectively deny access to protection. Additionally, the preventive approach of smart borders tends to facilitate instances of «data-banned populations», as coined by Didier Bigo (2014), where individuals are denied entry to a European State based on data-driven profiling and categorisation, rather than individual assessment. Information gathered through digital patrolling can in fact lead to targeted policing operations that further entrench these exclusionary practices.

This opens the way to data-sharing practices with third States or non-State actors resulting in the denial of access to protection or, more broadly, in the limitation of mobility possibilities. These forms of cooperation, which could be described as forms of «digital refoulement», are difficult to document and challenge, yet their consequences can be severe from a human rights perspective (Fill 2021).

In addition to these fundamental provisions on the right to seek protection and the non-refoulement obligation, several international and European norms closely connected to the right to life apply to border surveillance operations and digital patrolling activities25.

At the international level, particularly regarding maritime frontiers, relevant legal frameworks emerge from the law of the sea, as defined by the UN Convention on the Law of the Sea (UNCLOS)26, the Safety of Life at Sea Convention (SOLAS)27, and the Search and Rescue Convention (SAR)28. While a detailed discussion of these Conventions is beyond the scope of this analysis, it is essential to acknowledge that they form the fundamental regulatory background for border surveillance29. Notably, a recent report by the United Nations Special Rapporteur on Extrajudicial, Summary or Arbitrary Executions reaffirms that UNCLOS, SOLAS, and SAR intend to establish a system to rescue all vessels in distress, without discrimination based on nationality, status, or circumstances in which a person is found (General Assembly of the United Nations 2017, para. 61). These principles were further reinforced in the Views adopted by the Human Rights Committee (2021) regarding the A.S., D.I., O.I., and G.D. v. Italy and Malta case, which addressed responsibility for failure to rescue at sea also emphasizing the obligation of conducting rescue operations seriously, even in situations involving concurrent jurisdiction among States.

However, the analysis of the case studies raises concerns about the actual enforcement of these frameworks in the context of digital patrolling. As discussed, the enhanced surveillance capabilities at and before borders are not necessarily correlated with more effective rescue operations and easier access to protection: while the occurrence of undetected shipwrecks should become increasingly rare or even exceptional, tragedies at sea and in pre-frontier areas continue to occur. This seems indicative of the underlying priorities driving digital patrolling operations. Despite the claimed full situational awareness, the persistence of such incidents highlights a troubling disconnect between the technological capabilities of surveillance systems and human rights obligations.

4.5. Data protection and privacy rights: the grey area of border zones

As highlighted in the analysis of the border areas in Greece and Spain, the digitalisation of patrolling raises significant concerns regarding data protection and privacy rights for individuals subjected to such surveillance. In fact, data collection at external borders often resembles indiscriminate trawling rather than targeted interventions. Given the uncertainty surrounding the type and volume of data collected and processed, it is essential to explore the legal questions arising from international and European obligations, which could potentially limit indiscriminate data harvesting practices at the borders.

4.6. Moving forward: data rights between interoperability and AI

Before concluding the discussion on privacy and data protection, it is crucial to touch upon two areas currently at the forefront of debates and poised to become increasingly significant: the growing interoperability among EU IT systems and the regulation of Artificial Intelligence systems.

Considerations on the increasing role of interoperability between EU large-scale databases might seem peripheral to digital patrolling. However, recent developments clearly indicate a trend towards greater integration of patrolling technologies and data collection systems at the external borders, making a brief exploration of interoperability timely.

The interoperability framework, which aims to facilitate identity checks of non-EU nationals, is regulated by Regulation 2019/817 on information systems in migration and border control and by Regulation 2019/818 on police and judicial cooperation, asylum, and migration60. This framework allows national authorities to cross-check information on individuals across various EU databases, including VIS, SIS II, EURODAC (briefly introduced in chapter 1), and the European Criminal Records Information System-Third Country Nationals (ECRIS-TCN)61, which collects information on criminal convictions of non-EU citizens (see Brouwer 2020).

The main concerns surrounding the growing interoperability among different EU information systems, especially within the management of EU-Lisa, revolve around the semi-automated decisions triggered by ‘risk flags’, the overall inconsistency of interoperability with data protection standard62, and potentially discriminatory outcomes based on nationality or ethnic grounds (Giannakou 2021). According to Evelien Brouwer (2020, 92), the non-discrimination principle and the essence of data protection provisions have been compromised for non-EU citizens whose data is collected through these interoperable databases. Moreover, the results of cross-database research are often considered influenced by ‘automated digital bias’, i.e. the tendency of the human operators to trust alerts and other semi-decisions without questioning the suggested outcomes. Taken together, these circumstances could lead to further erosion of safeguards for vulnerable groups of third-country nationals.

The next phase of the interoperability process is being pursued through the «enhanced interoperability paradigm» (European Commission 2016; see also Hanke and Vitiello 2019). This initiative aims to improve existing databases while closing information gaps, de facto blurring regulatory boundaries among different branches of EU policies, such as the Common Security and Defence Policy (CSDP) and the Area of Freedom, Security and Justice (AFSJ). The initiative also expands the scope of different databases for law enforcement and border management purposes.

Practically, large-scale European databases are now merged and cross-checked thanks to four new components: the European Search Portal (ESP), the Shared Biometric Matching Service (sBMS), the Multiple-Identity Detector (MID), and the Common Identity Repository (CIR). The ESP enables users to search multiple information systems at the same time, the sBMS allows them to compare biometric data, and the MID helps in detecting multiple identities connected to the same biometric data. The CIR, storing biographical and biometric data, is expected to significantly impact digital patrolling operations by requiring the collection and storage of biometric data, such as fingerprints and facial images, of any person crossing the EU’s external borders63. This direct and indiscriminate integration of biometric data in border control activities suggests an even more explicit merging of border surveillance practices and data collection activities targeting people on the move. This process warrants close attention, as future developments here are likely to have significant long-term consequences on the digitalisation of patrols.

Moreover, scholars and human rights organizations have often argued that interoperability rarely meets the proportionality and necessity requirements, nor the provisions on purpose limitation, fairness, and transparency. They conclude that general considerations on public security should not suffice to limit fundamental rights related to privacy to this extent (Blasi Casagran 2021, 443; Statewatch and PICUM 2019). The New Pact on Migration and Asylum also emphasizes enhanced interoperability (European Commission 2020b), but leaves little room for respecting principles such as proportionality, data minimization, and purpose limitation. This raises further doubts about whether it should truly be considered a «fresh restart», as suggested by President Ursula von der Leyen (European Commission 2020a).

These considerations are particularly relevant in light of the ongoing debate around the Artificial Intelligence Act Proposal, presented by the European Commission on 21 April 2021 (European Commission 2020c). For the purposes of this analysis, suffice it to mention that the Proposal adopts a risk-based approach to AI regulation, aiming to distinguish between different kinds of AI systems64. The regulatory framework defines four levels of risk, corresponding to precise requirements and obligations: minimal or no risk, limited risk, high risk, and unacceptable risk.

In this classification, AI systems applied in migration, asylum, and border control are considered high-risk technologies (Article 60). The discussions surrounding the list of high-risk AI systems have been marked by intense lobbying and contentious negotiations. Critics argue that these discussions are hindered by a lack of attention to the specific contexts in which AI systems are deployed, particularly in sensitive and liminal areas like border zones. Scholars have pointed out that the distinction between AI systems considered ‘unacceptable risk’ (and thus banned, such as predictive policing in criminal justice) and those deemed ‘high risk’ (allowed under strict safeguards) is not based on clear, objective criteria but rather on political compromises and arbitrary considerations (Edwards 2022). Despite these criticisms, the identification of high-risk technologies does introduce some (limited) accountability measures, such as the obligation to inform individuals affected by high-risk AI systems and the introduction of specific registration duties for public authorities. However, these measures do not result in corresponding individual rights, such as mechanisms for complaint or redress against the use of AI by public authorities, that are not mentioned by the Proposal (EDRI et al. 2021; EDRI 2022). This significantly lowers the scope of the protection thus afforded.

Given the rapid development and deployment of new AI systems, especially in the surveillance domain, several civil society organizations are advocating for a more ‘future-proof’ approach to the categorisation of risks. They argue that the current procedures for updating risk lists are too slow and inflexible to be able to keep pace with technological innovation (see EDRI et al. 2021). On a more positive note, the draft report published in April 2022 by the European Parliament’s Committee on the Internal Market and Consumer Protection and the Committee on Civil Liberties, Justice and Home Affairs has been praised for shifting towards a more people-centred approach to risk-based regulation, also making the update process more participatory, ensuring greater involvement from civil society (European Parliament 2021).

However, at the time of writing, debates on the AI Act continue to grapple with unresolved issues, particularly the harmful use of AI in migration and asylum contexts. Despite evidence that AI systems can contribute to discriminatory practices at and before borders, the draft fails to address these concerns adequately (Bircan and Korkmaz 2021; EDRI 2022). In support of these arguments, a significant concern is the exclusion of large-scale EU IT systems from the AI Proposal, as specified in Article 83(1). This exclusion raises serious human rights concerns, especially within the interoperability framework.

Although the Proposal frequently references fundamental rights as a criterion for assessing AI uses, it does not mandate comprehensive ex-ante fundamental rights-based impact assessments. This gap means there are no robust safeguards integrated by design into the development of AI systems. While certification for ‘essential requirements’ is required for high-risk systems, many civil society organizations, human rights agencies, and also Member States argue that a more comprehensive assessment should be mandatory to ensure accountability for AI-related harm (EDRI et al. 2022; European Union Agency for Fundamental Rights 2020).

AI systems that handle biometric data require special attention. At this stage, the Proposal classifies these systems as high-risk when used for identifying and categorizing individuals in contexts such as migration, asylum, border control, law enforcement, and emotion recognition. The 2020 European Commission White Paper on Artificial Intelligence allows the use of AI for remote biometric identification if the operation is «duly justified, proportionate and subject to adequate safeguards» (European Commission 2020e, 20). Despite ongoing discussions, the use of remote biometric identification systems is not yet prohibited under the Proposal, although there are indications that the European Parliament may advocate for banning such systems in publicly accessible spaces (EDRI 2022). However, even if such a ban were implemented, it would likely not extend to militarized areas like the Evros region or the enclaves of Ceuta and Melilla.

Moreover, the deployment of AI systems in border areas raises specific concerns about public transparency. The ‘black box’ problem, which refers to the difficulty of understanding how AI systems generate their outputs, is particularly troubling when it involves risk-based alerts used in surveillance systems like EUROSUR or the SIVE65.

In conclusion, while the EU’s efforts to regulate AI «assert and reinvent its artificial intelligence as anchored in the checks and balances of the rule of law» (Hildebrandt 2020), many questions about the use of digital patrolling systems remain unresolved. These legal provisions, in fact, do not provide sufficient protection for individuals in border areas, where the risks of fundamental rights violations are well known. Once again, digital patrolling – straddling the line between border control and migration management – seems to evade the most protective legislative and jurisprudential developments, potentially sidelining also the key principles enshrined by the European AI framework such as transparency, non-discrimination, and fairness (European Commission 2020d).

4.7. The right to non-discrimination: differentiated treatment and protected grounds

After examining human dignity, rights related to access to international protection, and the contentious issues surrounding privacy and data protection, this section explores another crucial human rights concern in the context of digital patrolling: the right to non-discrimination.

Discrimination is in fact a central issue in the realm of remote surveillance and digitalized patrolling, as these technologies inherently involve the categorization, filtering, and classification of human mobility. As previously discussed, the primary function of smart borders is to discriminate in the etymological sense of the term – differentiating between those who may pass and those who may not. However, as noted by Tendayi Achiume, UN Special Rapporteur on Contemporary Forms of Racism, Racial Discrimination, Xenophobia, and Related Intolerance, modern borders often discriminate in a normatively prejudicial way. This form of discrimination allocates fundamental human rights differently based on race, gender, class, national origin, sexual orientation, and disability status, among other factors (Achiume 2021, 333). Thus, it is imperative to critically assess how the digitalisation of border patrols might exacerbate discrimination on various grounds, potentially leading to differential compliance with fundamental rights.

Internationally, the right to freedom from discrimination and the principle of equality are enshrined in a vast network of legal norms, deeply embedded in the international legal order. These rights are in fact protected by several key international instruments, including the UDHR (Articles 7 and 10), the International Covenant on Economic, Social and Cultural Rights (ICESCR, Articles 2 and 10), the ICCPR (Articles 4, 24, and 26), the Convention on the Elimination of Racial Discrimination (CERD), the Convention on the Elimination of All Forms of Discrimination against Women (CEDAW), the Convention on the Rights of Persons with Disabilities (CRPD, Articles 3 and 5), the United Nations Convention on the Rights of the Child (CRC, Article 2), and the Refugee Convention (Article 3). These provisions prohibit both direct and indirect discrimination based on various grounds, including race, sex, religion, political beliefs, and national or social origin. Indirect discrimination occurs when ostensibly neutral policies or practices disproportionately affect certain individuals or groups, leading to less favourable treatment66. As discussed earlier, the motif of neutrality can often mask exclusionary practices and perpetuate power imbalances, making it a crucial theoretical juncture in the study of digital patrolling. It is in fact plausible that digital patrolling technologies may produce discriminatory outcomes related to protected characteristics, such as race and gender (Molnar 2021).

The 2021 report by the Human Rights Council on racial and xenophobic discrimination in the context of digital technologies in border and immigration enforcement is particularly insightful on this issue (Human Rights Council 2021a)67. The report highlights the lack of a cohesive global governance framework to regulate the use of automated and digital technologies in border enforcement. It thus calls for a moratorium on surveillance technologies until robust human rights safeguards are established. These safeguards should include due diligence in line with IHRL standards, independent oversight, and full transparency regarding the use of surveillance systems that could lead to racial and other forms of discrimination (Human Rights Council 2021a, para. 66). Similarly, legal scholars have argued that considerations of non-discrimination should be integrated by design and from the outset in digitalisation processes to prevent fundamental rights violations (Beduschi 2019). However, as with broader human rights assessments, the tendency is to conduct these evaluations ex-post often undermining their role.

In the European context, the principle of non-discrimination is well-established in EU law, enshrined in the Treaty on European Union (Article 2), the Treaty on the Functioning of the European Union (Article 10), and the Charter of Fundamental Rights (Articles 20 and 21, which address equality before the law and non-discrimination). Additionally, the ECHR prohibits discrimination in Article 14, while Protocol 12 (Article 1) provides a general prohibition on discrimination, not limited to the rights protected by the Convention itself. Furthermore, the Schengen Border Code specifically requires border guards to perform their duties without discrimination based on gender, race, ethnic origin, religion, belief, disability, age, or sexuality68.

As emerged in this chapter, two main axes of discrimination emerge in the context of digital patrolling. The first involves the targeting of non-EU nationals by border surveillance systems. The European Court of Human Rights has acknowledged that the protected position of EU citizens does not justify a priori differential treatment of non-EU nationals (see Brouwer 2011, 156)69. Such targeted surveillance may be inherently discriminatory against migrant people, particularly when border surveillance is intertwined with security objectives. This approach effectively singles out a well-defined population of non-EU people on the move as a high-risk group, often accompanied by stigmatizing and punitive attitudes (see Bigo, Ewert, and Kuşkonmaz 2020; Molnar 2021, 134; Statewatch and PICUM 2019). Various organizations have criticized the bias inherent in many surveillance systems at external borders, which often operate on the flawed assumption that migration is closely linked to criminal justice, fostering an equivalence between migration and serious transnational crimes, such as terrorism. Alarmingly, the scope of those deemed ‘risky’ continues to expand, increasingly encompassing migrant people in international waters or transit countries.

The second axis of discrimination relates to protected grounds under international and European law, such as ethnicity, race, gender, and religion. As discussed, it is extremely challenging to fully understand how the algorithms processing the data collected by drones or other digital patrolling systems work in practice. A notable example is the situation at the borders of Ceuta and Melilla, where overt discrimination against sub-Saharan migrant people is likely supported by the digital surveillance systems employed in the frontera inteligente. On a more speculative level, it can be assumed that UAV patrols may allow for the identification of a person’s origin, gender, and, in the case of visible religious symbols, their beliefs. Whether this information is then used to implement discriminatory policies based on risk assessments is difficult to prove or disprove with the information publicly available.

4.8. Keeping States accountable? Sketched considerations on jurisdiction

The analysis of the digitalisation of patrolling and, more broadly, the integration of new technologies into border surveillance reveals that these developments occur within highly dense and articulated regulatory frameworks. However, these often fail to adequately address the loopholes opened up by digitalisation. This creates a serious challenge where violations of rights are difficult to establish due to the lack of clear rules governing the use of these new technologies.

At the national level, existing accountability mechanisms frequently prove ineffective, even when they involve independent authorities, a concern repeatedly raised by the EU Fundamental Rights Agency70. Consequently, there is an urgent need for oversight mechanisms tailored to the specific contexts in which these technologies are deployed. Such mechanisms are essential to prevent the unchecked experimental use of new technologies in border control. Here, the question of jurisdiction regarding fundamental rights is particularly relevant, as it may provide a framework for developing accountability mechanisms and a valuable angle to further develop this study. The deployment of digital patrolling systems extends in fact the scope of externalisation and outsourcing policies to extreme and otherwise unimaginable levels. This contributes to the dilution of liability, accountability, and responsibility among actors, deliberately dispersing and shifting them among third States, European agencies, and private actors. Such practices raise complex jurisdictional issues.

The core question – which is here left completely open – concerns whether the various functions of digital patrolling, such as situational awareness, detection, tracking, information management, and risk analysis, constitute an exercise of jurisdiction over people on the move. This issue inevitably intersects with the broader and highly complex debate on the extraterritorial exercise of jurisdiction, which extends well beyond the scope of this book.

From the perspective of IHRL, the debate on the extraterritorial application of human rights conventions is relatively recent and limited. IHRL itself is grounded in the recognition of States’ legal orders and features relatively weak enforcement mechanisms. As Nehal Bhuta (2016) vividly describes, the extraterritorial application of human rights mirrors the fracture lines of the Westphalian system, suggesting that as State activities extend beyond their borders and disconnect from traditional notions of territorial and political unity, human rights law must evolve to keep pace. The challenge of extraterritoriality is fundamental to the future of human rights in this area: while an increased focus on extraterritorial obligations could lead to higher standards of protection, a conservative interpretation of the territorial scope of State’s positive and negative obligations could significantly weaken IHRL norms.

To tackle this challenge, various human rights courts and treaty bodies have developed a ‘factual control’ test to expand the scope of territorial jurisdiction, applying it to both the control of territory and individuals (see Committee Against Torture 2008, para. 7; International Court of Justice 2004, paras 109–110; OHCHR 2004). The ECtHR has also made strides in this direction, increasingly adopting a ‘functional approach’ to the applicability of fundamental rights in cases involving extraterritorial policies and practices (Carrera and Cortinovis 2019). The Court’s jurisprudence recognizes in fact a jurisdictional link between individuals affected by external border control measures and the State responsible for those measures, whether through de jure or de facto control over the territory where the violation occurs or over the individual affected71.

Some scholars, such as Fernando Val Garijo (2020, 148), argue that applying this test could be sufficient to trigger jurisdiction in cases where pullback operations are supported by drones and other remote-controlled systems. However, extending this interpretation to fully cover digital patrolling remains a particularly ambitious and underexplored proposition.

Border controls that are increasingly digitalized, externalized, and privatized pose a significant concern in legal scholarship. Valsamis Mitsilegas (2015, 21) highlights the resulting issues as a critical challenge to the rule of law and human rights, emphasizing the urgent need for a paradigm shift that centres individuals – not States – in border control legislation and practice. Such a shift would be crucial to ensure that individual rights do not fade into the background, overshadowed by security considerations and other priorities.

The issue of jurisdiction, particularly its extraterritorial expansion, and its implications for digitalisation, thus remains central in the effort to strengthen an accountability framework that considers human dignity, access to international protection, privacy and data protection, and non-discrimination in the context of digital patrolling. In conclusion, the debate on jurisdiction and accountability remains open and critically important as we continue to grapple with the legal and political implications of digitalized border control.

1 Law no. 4636/2019, On international protection and other provisions, 1 November 2019.

2 Law no. 4686/2020, Improvement of the migration legislation, amendment of Law 4636/2019, 4375/2016, 4251/2014 and other provisions, 12 May 2020.

3 It is crucial to highlight that transparency issues surrounding high levels of surveillance – often difficult to reconcile with democratic principles – are becoming increasingly prevalent, not only in Greece and Spain but across many European countries. According to the European Digital Rights (EDRi) (2020), at least fifteen European countries have, in recent years, engaged in trials involving highly invasive facial and biometric recognition technologies aimed at mass surveillance.

4 Presidential Decree no. 98/2019, Organisation and structure of the Drone Service, Establishment of Procurement Directorates and History of the Hellenic Police and Unmanned Aircraft Service, 21 March 2017.

5 Presidential Decree no. 75/2020, Use of surveillance systems obtaining or documenting sound and pictures in public places, 10 September 2020.

6 Law no. 18/2014, On the approval of urgent measures for growth, competitiveness and efficiency, 15 October 2014.

7 UN General Assembly, Universal Declaration of Human Rights, 217 A (III), 10 December 1948. In particular, reference to human dignity is present in the International Convention on the Elimination of all Forms of Racial Discrimination (UN General Assembly, International Convention on the Elimination of All Forms of Racial Discrimination, United Nations, Treaty Series, vol. 660, 21 December 1965), the International Covenant on Economic, Social and Cultural Rights (UN General Assembly, International Covenant on Economic, Social and Cultural Rights, United Nations, Treaty Series, vol. 993, 16 December 1966), the International Covenant on Civil and Political Rights (UN General Assembly, International Covenant on Civil and Political Rights, United Nations, Treaty Series, vol. 999, 16 December 1966), the Convention on the Elimination of All Forms of Discrimination against Women (UN General Assembly, Convention on the Elimination of All Forms of Discrimination Against Women, United Nations, Treaty Series, vol. 1249, 18 December 1979), the Convention against Torture and other Cruel, Inhuman or Degrading Treatment or Punishment (UN General Assembly, Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, United Nations, Treaty Series, vol. 1465, 10 December 1984), the Convention on the Rights of the Child (UN General Assembly, Convention on the Rights of the Child, United Nations, Treaty Series, vol. 1577, 20 November 1989), the International Convention on the Protection of the Rights of All Migrant Workers and Members of Their Families (UN General Assembly, International Convention on the Protection of the Rights of All Migrant Workers and Members of their Families, A/RES/45/158, 18 December 1990), and the Convention on the Rights of Persons with Disabilities (UN General Assembly, Convention on the Rights of Persons with Disabilities, A/RES/61/106, 24 January 2007).

8 See Klein and Kretzmer (2002) for an in-depth study of the development of the concept of human dignity between International Human Rights Law and international jurisprudence.

9 See CJEU, Netherlands v. European Parliament and Council, C-377/98, 9 October 2001, §70-77.

10 From this perspective, it is worth noting that the UN Global Compact for Safe, Orderly, and Regular Migration (GCM) has contributed to advancing, or at least facilitating, the adoption of unified and standardised approaches to migration management. These approaches include, among other aspects, the establishment of frameworks for efficient border crossings, with a pronounced emphasis on the integration of information technologies, pre-screening measures, and data collection on individuals involved. This trend toward the standardisation of migration management has prompted various concerns, particularly surrounding the endorsement of extensive data collection practices. These concerns are well-founded, given the disparate levels of data protection among States, which are far from standardised. For a nuanced discussion of these issues, see Kuşkonmaz (2021).

11 See chapter 3.

12 Among the reports documenting violations at European borders, Human Rights 360° (2020) is particularly relevent.

13 UN General Assembly, Convention Relating to the Status of Refugees, United Nations, Treaty Series, vol. 189, 28 July 1951 and UN General Assembly, Protocol Relating to the Status of Refugees, United Nations, Treaty Series, vol. 606, 31 January 1967.

14 The ICCPR does not explicitly prohibit the removal of individuals to States that practice or tolerate human rights violations. However, the UN Human Rights Committee has interpreted Articles 2, 6, and 7 of the Covenant to prevent extradition, deportation, expulsion, or any form of removal when there are well-founded fears of a real risk of irreparable harm, both in the country of expulsion and in any other country to which the individual might subsequently be sent (chain refoulement). Additionally, the principle of non-refoulement is reinforced by other international and regional instruments, such as the 1967 United Nations Declaration on Territorial Asylum, the 1969 Organization of African Unity (OAU) Convention Governing the Specific Aspects of Refugee Problems in Africa, the 1984 Cartagena Declaration on Refugees, and the 1966 Bangkok Principles on the Status and Treatment of Refugees. For a deeper analysis of these instruments, see James C. Simeon (2019).

15 At the European Union level, the right to asylum is moreover enshrined in the Qualification Directive (2011/95/EU, Articles 2, 13, 18), the Asylum Procedures Directive (2013/32/EU, Preamble paras. 12, 15-18 and Articles 1, 2, 10), the Reception Conditions Directive (2013/33/EU, Preamble paras. 26, Articles 3, 6), the Schengen Borders Code (Regulation EU 2016/399), and the Dublin Regulation (604/2013).

16 Qualification Directive 2011/95/EU, Article 21.

17 Asylum Procedures Directive 213/32/EU, Articles 9, 28, 35, 38, 39, 41 and Annex I.

18 Return Directive 2008/115/EC, Articles 4, 5 and 9.

19 Schengen Borders Code Regulation (EU) No. 2016/399, Article 4.

20 See ECtHR, Saadi v. Italy, Application no. 37201/06, 28 February 2008.

21 In Soering v. United Kingdom, Application no. 14038/88, 7 July 1989, the European Court of Human Rights clarified that the principle of non-refoulement is supported by a par ricochet protection. The Court also established that, under certain circumstances, the responsibility of a State can arise due to the actions undertaken by another State.

22 This pivotal ruling has, in fact, marked the transition from pushback operations between Italy and Libya to pullback operations, in coordination between Italian and Libyan authorities. These operations, while characterised by a greater degree of externalisation, are no less severe in terms of the violations of the fundamental rights of the individuals involved.

23 ECtHR, Hirsi Jamaa and Others v. Italy, Application no. 27765/09, 23 February 2012; and ECtHR, N.D. and N.T. v. Spain, Applications nos. 8675/15 and 8697/15, 13 February 2020. At §185, the Court provides a definition of the term ‘expulsion’, described as any forcible removal of aliens from a State’s territory, irrespective of the lawfulness of the stay, the length of time they spent in the territory, the location in which they were apprehended, and their status as migrants or asylum-seekers.

24 See ECtHR, Hirsi Jamaa and Others v. Italy, Application no. 27765/09, 23 February 2012, §178.

25 Border surveillance provisions within the EU have already been discussed before. Suffices here to recall that the legal framework for border surveillance is established by Article 12 of the Schengen Borders Code and Article 12 of the EUROSUR Regulation. The former emphasises that the primary objective of border surveillance is to prevent unauthorised border crossings, counter cross-border criminality, and take action against those who have crossed the border illegally. The latter mandates the common application of surveillance tools, thus creating the context in which Frontex’s coordination and support role can be fully realised. Frontex, indeed, plays a central role in the border enforcement framework through operations and interventions that frequently raise accountability and transparency concerns. On this issue, see Lena Karamanidou and Bernd Kasparek (2020).

26 UN General Assembly, Convention on the Law of the Sea, 10 December 1982.

27 International Maritime Organization, International Convention for the Safety of Life at Sea, 1 November 1974.

28 International Maritime Organization, International Convention on Maritime Search and Rescue, 27 April 1979.

29 For a deeper analysis of border surveillance under International Law, see Luisa Marin (2017).

30 See HRC, Van Hulst v. Netherlands, Communication 903/1999, UN Doc. A/60/40, Vol. II, 1 November 2004 and HRC, NK v. Netherlands, Communication 2326/2013, UN Doc. CCPR/C/120/D/2326/2013/Rev.1, 18 July 2017

31 The UN Special Rapporteur on the Right to Privacy (OHCHR 2019, para. 18) elaborated on this concept by applying the legitimate aims outlined in the limitation clause of Article 22 of the ICCPR, which pertains to freedom of assembly, to the grounds on which interference under Article 17 may be justified.

32 See, for instance, IACHR, Escher et al. v. Brazil, Series C No. 193, 6 July 2009, §116, and ECHR, Big Brother Watch and Others v. United Kingdom, Applications nos. 58170/13, 62322/14 and 24960/15, 25 May 2021, §304.

33 On limitations, see European Court of Human Rights (2021, 25).

34 Reference should be made to the CJEU decision in the Joined Cases Digital Rights Ireland Ltd v. Minister for Communications, Marine and Natural Resources and Others and Kärntner Landesregierung and Others, C-293/12 and C-594/12, 8 April 2014. In this ruling, the Court underscored the importance of the principle of proportionality within the EU law-making process, emphasising that legislative measures must not be more intrusive than necessary in order to achieve a lawful and specific purpose. The Court also urged the adoption of clearer and more precise rules concerning data collection, retention, and processing.

35 The Article states that any limitation on fundamental rights must be prescribed by law, genuinely pursue objectives of general interest recognised by the Union or protect the rights and freedoms of others, must respect the essence of the right, and be necessary and proportionate.

36 See CJEU, Joined Cases Digital Rights Ireland Ltd v. Minister for Communications, Marine and Natural Resources and Others and Kärntner Landesregierung and Others, C-293/12 and C-594/12, 8 April 2014; and CJEU, Schrems v. Data Protection Commissioner, C-362/14, 6 October 2015. See also European Data Protection Supervisor (2017, 21).

37 See ECtHR, Dimitrov-Kazakov v. Bulgaria, Application no. 11379/03, 10 February 2011, § 70, and ECtHR, Shimovolos v. Russia, Application no. 30194/09, 21 June 2011, § 33.

38 See ECtHR, Catt v. United Kingdom, Application no. 43514/15, 24 January 2019, § 105.

39 See ECtHR, Centrum för Rättvisa v. Sweden, Application no. 35252/08, 19 June 2018 and ECtHR, Big Brother Watch and Others v. United Kingdom, Applications nos. 58170/13, 62322/14 and 24960/15, 25 May 2021.

40 Regulation no. 2016/679 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation, GDPR), 27 April 2016 and Directive no. 2016/680 on the protection of natural persons with regard to the processing of personal data by competent authorities for the purposes of the prevention, investigation, detection or prosecution of criminal offences or the execution of criminal penalties, and on the free movement of such data (Law Enforcement Directive), 27 April 2016.

41 Without delving into the specifics, the Entry/Exit System Regulation, referenced in chapter 1, should be recalled. In Recital 36, it stipulates that the EES Regulation is to be regarded as lex specialis in relation to the GDPR. Consequently, in the event of any conflict, the data protection provisions are not applicable. This relationship between the regulations is particularly revealing of the actual priorities that are intended to be safeguarded.

42 At the national level, a notable development occurred in 2016, when the United Kingdom enacted data protection legislation that applied to every individual within its jurisdiction. This legislation, therefore, also potentially covered migrant people at the external borders.

43 According to the ECtHR, to be considered necessary, an interference must happen in light of a «genuine, objective, and sufficiently important need». See ECtHR, Handyside v. United Kingdom, Application no. 5493/72, 7 December 1976, § 48 and ECtHR, Sunday Times v. United Kingdom, Application no. 13166/87, 26 November 1991, § 62.

44 See ECtHR, Segerstedt-Wiberg and Others v. Sweden, Application no. 62332/00, 6 June 2009; and ECtHR, Dalea v. France (dec.), Application no. 964/07, 2 February 2010.

45 See ECtHR, McVeigh, O’Neill and Evans v. United Kingdom, Applications nos. 8022/77, 8025/77 and 8027/77, 18 March 1981; ECtHR, Kinnunen v. Finland, Application no. 24950/94, 15 May 1996; ECtHR, S. and Marper v. United Kingdom, Application no. 30562/04 and 30566/04, 4 December 2008; ECtHR, Dimitrov-Kazakov v. Bulgaria, Application no. 11379/03, 10 February 2011; ECtHR, M.K. v. France, Application no. 19522/09, 18 April 2013; ECtHR, Suprunenko v. Russia, Application no. 8630/11, 19 June 2018; ECtHR, Gaughran v. United Kingdom, Application no. 45245/15, 13 February 2020; ECtHR, P.N. v. Germany, Application no. 74440/17, 11 June 2020; and ECtHR, Willems v. Netherlands, Application no. 57294/16, 9 November 2021.

46 ECtHR, P.N. v. Germany, Application no. 74440/17, 11 June 2020.

47 ECtHR, P.G. and J.H. v. United Kingdom, Application no. 44787/98, 25 September 2001; ECtHR, Allan v. United Kingdom, Application no. 48539/99, 5 November 2002; ECtHR, Doerga v. Netherlands, Application no. 50210/99, 27 April 2004; ECtHR, Vetter v. France, Application no. 59842/00, 31 May 2005; ECtHR, Wisse v. France, Application no 71611/01, 20 December 2005.

48 For a deeper analysis of the access to justice on AI issues, see European Union Agency for Fundamental Rights (2020).

49 Directive no. 95/46 on the protection of individuals with regard to the processing of personal data and on the free movement of such data, 24 October 1995.

50 On the minimization of data quantities, see ECtHR, L.L. v. France, Application no. 7509/02, 11 April 2006, §§ 45-46; ECtHR, Vicent Del Campo v. Spain, Application no. 25527/13, 6 November 2018, § 51; ECtHR, Khadija Ismayilova v. Azerbaijan, Applications nos. 65286/13 and 57270/14, 10 January 2019, § 147; ECtHR, Kruglov and Others v. Russia, Applications nos. 11264/04 and 15 others, 4 February 2020, § 132. On the limitation of collection purposes, see for instance ECtHR, Karabeyoglu v. Turkey, Application no. 30083/10, 7 June 2016 and ECtHR, K.H. and Others v. Slovakia, Application no. 32881/04, 28 April 2009. See ECtHR, Leander v. Sweden, Application no. 9248/81, 26 March 1987; ECtHR, Z. v. Finland, Application no. 9/1996/627/811, 25 February 1997; and ECtHR, Segerstedt-Wiberg and Others v. Sweden, Application no. 62332/00, 6 June 2009.

51 See ECtHR, Leander v. Sweden, Application no. 9248/81, 26 March 1987; ECtHR, Z. v. Finland, Application no. 9/1996/627/811, 25 February 1997; and ECtHR, Segerstedt-Wiberg and Others v. Sweden, Application no. 62332/00, 6 June 2009.

52 According to Article 7 of the Directive, to be regarded as legitimate, data processing must: be carried out after freely given, specific, and informed consent is obtained; be necessary for the performance of a contract to which the data subject is a party; be necessary for compliance with a legal obligation or for the performance of a task carried out in the public interest or in the exercise of official authority vested in the controller or in a third party to whom the data are disclosed; be necessary in order to protect the vital interests of the data subject; and be necessary for the purposes of a legitimate interest.

53 See chapter 1.

54 In Europe, legislation on the use of drones mainly derives from the Commission Delegated Regulation (EU) 2019/945 of 12 March 2019 on unmanned aircraft systems and on third-country operators of unmanned aircraft systems and the Commission Implementing Regulation (EU) 2019/947 of 24 May 2019 on the rules and procedures for the operation of unmanned aircraft. Issues related to the use of drones are also addressed by the Frontex Regulation, which at Article 7 lays the basis for the Agency to buy the necessary technical assets and to collect and use the information acquired through drones and satellite.

55 ECtHR, S. and Marper v. United Kingdom, Application no. 30562/04 and 30566/04, 4 December 2008.

56 ECtHR, S. and Marper v. United Kingdom, Application no. 30562/04 and 30566/04, 4 December 2008, § 47.

57 On the criminalisation of migrant populations through borders and databases, see Nina Amelung (2021).

58 See CJEU, Joined Cases Tele2 Sverige AB and Watson, C-203/15 and C-698/15, 21 December 2016.

59 A similar reasoning was advanced by the Court in earlier cases: ECtHR, Leander v. Sweden, Application no. 9248/81, 26 March 1987, § 59; and ECtHR, Dalea v. France, Application no. 54/1996/773/974, 19 February 1998.

60 Regulation no. 2019/817 on establishing a framework for interoperability between EU information systems in the field of borders and visa, 20 May 2019, and Regulation no. 2019/818 on establishing a framework for interoperability between EU information systems in the field of police and judicial cooperation, asylum and migration, 20 May 2019.

61 Regulation 2019/816, adopted in April 2019.

62 It is often argued that the Regulations on interoperability do not provide any specific protection for the processing of biometric templates – i.e. digital representations of biometric samples – stored in the sBMS. On the issue, see among others Cristina Blasi Casagran (2021, 450).

63 This development aligns with the introduction of two types of biometric identifiers – four fingerprints and a facial image – under the Entry/Exit Regulation. The Regulation has provided for the use of facial recognition technology for verification purposes at the border. On the issue, read Niovi Vavoula (2021, 475).

64 In particular, Annex I of the Proposal distinguishes between: (a) Machine learning approaches, including supervised, unsupervised and reinforcement learning, using a wide variety of methods including deep learning; (b) Logic- and knowledge-based approaches, including knowledge representation, inductive (logic) programming, knowledge bases, inference and deductive engines, (symbolic) reasoning and expert systems; (c) Statistical approaches, Bayesian estimation, search and optimization methods.

65 For a critical perspective on the ‘black box’ phenomenon, see Frank Pasquale (2015).

66 The Committee on the Elimination of Racial Discrimination (2008, para. 10), interpreting the Convention, provided a useful definition of forms of indirect discrimination, finding that «indirect – or de facto – discrimination occurs where an apparently neutral provision, criterion or practice would put persons of a particular racial, ethnic or national origin at a disadvantage compared with other persons, unless that provision, criterion or practice is objectively justified by a legitimate aim and the means of achieving that aim are appropriate and necessary».

67 See also chapter 1. Among the issues addressed, the Report proposes a detailed examination of how borders enforcement technologies may facilitate or exacerbate racial discrimination. It highlights the role of online platforms, particularly social media, which can be a breeding ground for discrimination and xenophobic hatred; it scrutinises racial profiling systems linked to the deployment of new technological tools and AI in security, border control, and social services as aggravating forms of racism, racial discrimination, and xenophobia; it discusses the implications of mandatory biometric data collection and identification systems, which are believed to result in direct and indirect discrimination based on race, ethnicity, national origin, descent, and religion, notably as migrant people and asylum seekers often lack control over the personal data they provide; it examines language recognition systems, such as those used by the German Federal Office for Migration and Refugees, which analyse an applicant’s spoken language sample to assess the plausibility of their claimed nationality; and finally addresses mobile data extraction and social media intelligence practices specifically targeting migrant and refugee populations. These practices are widespread in several countries, including Austria, Belgium, Denmark, Germany, Norway, and the United Kingdom.

68 Regulation no. 2016/399 on a Union Code on the rules governing the movement of persons across borders (Schengen Borders Code), 9 March 2016.

69 ECtHR, Gaygusuz v. Austria, Application no. 17371/90, 16 September 1996; and ECtHR, Mamatkulov and Abdurasulovic v. Turkey, Applications nos. 46827/99 and 46951/99, 6 February 2003.

70 As recalled by the EU Fundamental Rights Agency (2020, 9), States have in fact a positive obligation to put in place effective monitoring and enforcement mechanisms in order to secure people’s rights and freedoms, also when it comes to regulate digitalisation processes and the deployment of Artificial Intelligence systems.

71 The ECtHR jurisprudence has evolved through fundamental cases that are worth mentioning. Ruling on Loizidou v. Turkey (Application no. 15318/1989, 23 February 1995) and on Ilaşcu and Others v. Moldova and Russia (Application no. 48787/99, 8 July 2004), the Court recognized the responsibility for violations carried out outside the territory of the State, according to the criterion of ‘effective control of the territory’ where the violation takes place. In 2004, with the judgment on Issa v. Turkey (Application no. 3821/1991, 16 June 2004), this test has been extended through a functional approach to the ‘effective control over the person’ who suffers the violation by means of operating agents of the State: the Court has indeed stated at §71 that Article 1 of the Convention cannot be interpreted so as to allow a State party to perpetrate violations of the Convention on the territory of another State, which it could not perpetrate on its own territory. Moreover, with the definition of the so-called ‘Hirsi test’ (ECtHR, Hirsi Jamaa and Others v. Italy, Application no. 27765/09, 23 February 2012), the Court referred to the continuous or exclusive control de jure and de facto as the main benchmark or criterion. According to the Court, ruling on Hirsi, even though the pullback operation took place in high sea, Italy had exercised continuous and exclusive de jure and de facto control over the applicants as they were embarked on a military boat. Similarly, in Medvedyev (ECtHR, Medvedyev and Others v. France, Application no. 3394/03, 29 March 2010), the Court concluded that France was exercising at least de facto full and effective control over a boat during an interception operation in high seas, adopting a functional approach to extraterritorial operations. Furthermore, in Al-Skeini and in Jaloud (Al-Skeini and Others v. United Kingdom, Application no. 55721/07, 7 July 2011; Jaloud v. Netherlands, Application no. 47708/08, 20 November 2014), the ECtHR adopted a more functional conceptualisation of jurisdiction, establishing that the latter emerges where the State exercises forms of public power normally to be exercised by a sovereign government – for instance, the maintenance of security – and consequently brings people under its authority and control.