Symposium introduction: the ethics of border controls in a digital age

ABSTRACT This symposium brings into conversation normative political theory on migration and critical border/migration studies, with a particular focus on digital border control technology. Normative theorists have long been concerned with questions about the extent and nature of control over migration that the state should exercise, and the balance of rights and duties between states and migrants. To date, however, there has been little reflection among such theorists on digital border control technology. Critical border/migration studies scholars, on the other hand, have paid considerable attention to the rapid development of digital technology in the border control/mobility management space, and revealed a range of problems with the technology itself and the ways it is deployed. What has thus far been lacking, however, is sustained ethical reflection on what should be done about the use of this technology. The papers in this symposium thus seek to bring these two groups of scholars together and to prompt what we hope will be a sustained conversation on these rapidly evolving and deeply problematic practices. This introduction contextualises the issue at the heart of this symposium – the rapid expansion of digital border controls and the ethical challenges that these pose – and offers brief summaries of the contributions.


Introduction
'Borders have guards and guards have guns ', claimed Joseph Carens (1987, 251) in a 1987 article inaugurating what has become known in normative theory as the immigration admissions debate.The imagery of the gun-wielding (and gun-shooting) border guard served as a visceral reminder of the stakes surrounding immigration and border control for those seeking a new life elsewhere and the attendant need for ethical reflection on, and ethically sound limitations to, state control of borders.Attention to armed guards usefully reminds us how immigration enforcement is coercive, and that coercion frequently involves physical violence.Normative theorists have rightly asked if and how this coercion can be justified.Nonetheless, many people who are prevented from immigrating never encounter a border guard and never see a gun.
Big data and AI have brought about what is arguably a qualitative shift in border controls.How should we respond to border controls if the border guard is (in some significant sense) an algorithm that bases decisions on the characteristics of ''Dividuals' constructed from vast, interoperable databases?What if the functional border is not a 'line in the sand' (Parker and Vaughan-Williams 2009), but rather a mobile border that shifts with the geolocation of one's smartphone?Consider the case of Keisy Plaza, a Venezuelan asylum seeker, whose journey with two young children took her from Columbia, through the Darién Gap and Central America, to arrive in Ciudad Juárez, Mexico.She waits within sight of the United States, unable to secure an appointment with US Customs and Borders Protection through their smartphone app CBP One.Like many asylum-seekers, Plaza has spent weeks attempting to secure an appointment through the glitch-prone app; asylum-seekers whose app does not freeze or report errors, usually find that the few appointment slots provided were filled within minutes (Rios 2023).
To understand CBP One, it is necessary to understand the political and legal context around immigration in the United States (Kocher 2023).Prior to the Trump administration's 2018 Migrant Protection Protocols (also known as the 'Remain in Mexico policy'), asylum seekers could arrive at a port of entry and receive a credible fear interview from an Asylum officer, who would determine if there was a significant possibility of persecution or a well-founded fear of persecution.Those who successfully made their case would receive an appointment to present their case before an immigration judge and could enter the US.
The Migrant Protection Protocols instituted a significant departure from US asylum law by requiring that asylum seekers remain in Mexico for their appointment in the United States.This raised serious human rights concerns by placing major barriers to access to asylum and by leaving asylum seekers exposed to kidnapping, extortion, robbery, and rape.An even more restrictive COVID-19 policy, Title 42, followed, leveraging an obscure public health rule to justify expelling migrantsincluding asylum seekerswho arrived at ports of entry or crossed unlawfully.Despite its highly restrictive nature, Title 42 still provided limited exemptions for vulnerable migrants.Initially, the Department of Homeland Security relied on NGOs in Mexico to coordinate appointments for exemptions.
In January 2023, it stopped relying on NGOs and adopted a technological solution, mandating exemptions to be scheduled through CBP One.Asylum seekers download the app, enter their biographical information, and upload a photograph of their face.Only then can they attempt to schedule an appointment.CBP One uses the phone's GPS to determine the person's location (the app only allows asylum seekers to schedule appointments within a certain proximity to the US Border).The Biden administration ended the Migration Protection Protocols shortly after taking office, but it retained Title 42 until April 2022, when the Center for Disease Control and Prevention determined that it was no longer needed to protect public health.
What remained after the demise of Remain in Mexico and Title 42 was CBP One, an app developed by a notoriously anti-immigrant administration during a global pandemic.CBP One is one example of how smartphones have become increasingly central to border enforcement, in what Austin Kocher describes as 'effectively an extraterritorial and digital port of entry (italics in original)' (Kocher 2023, 6) and Jack Herrera describes as a 'sort of digital border wall' (Herrera 2023).The app creates barriers to asylum seekers, including possession of a smartphone, access to the Internet, the app freezing or not permitting people to log in, language barriers (currently, it only includes English, Spanish, and Haitian Creole), and the inability to reliably recognise darker-skinned faces (Kocher 2023).According to Amnesty International, the mandatory use of CBP One continues to undermine the right to seek asylum by denying the basic safeguards provided through access to territory (Amnesty International 2023).CBP One is an example of larger trends in the way states seek to control borders and manage migration on a global scale.Wealthy Western states have become remarkably adept at pushing their borders far beyond the physical demarcations of their territory and making immigration decisions at great distances, often co-opting poorer states to do their migration-related bidding.These developments have been greatly facilitated by the rapid rise of digital border control technologies.This has resulted in quite profound changes in the ways in which migrants encounter and experience borders, on the one hand, and the way the state interacts with migrants, on the other.
The fundamental reality driving this symposium is that, today, not all borders have guards, and not all guards have guns.Or to be more accurate: not all borders have human guards and not all guards have guns.This does not, however, render those borders and border controls any less deadly or dehumanising.In fact, it may render them more so.Today, a border guard is as likely to be an algorithm as a person, biometric data or risk profiling software the 'weapon' of control and decision-making, and one's own home/computer/phone to be the place of encounter with 'the border'.The purpose of this symposium is, on the one hand, to show why this reality matters for ethical reflection on migration and, on the other, to show what normative theory might be able to contribute to more empirically grounded work on the digitisation of borders.
Within the existing normative literature on migration, open and closed border advocates disagree, sometime quite vehemently, on immigration policy and who should control it.A variety of positions have been staked out in relation to this question, ranging from those broadly understood to favour more restrictive policies to those who favour more open policies, offering a wide range of different justifications for such positions.Arguments in support of the state's right to exclude outsiders include Wellman's 'freedom of association' defence (Wellman 2008), Miller's 'liberal nationalist' position focused on the political culture of a community (Miller 2016), the value Walzer places on 'communities of character' (Walzer 1983), defences of state welfare systems (Macedo 2018), and concerns about 'brain drain' (Brock and Blake 2015).Arguments in favour of restrictions on the state's right to exclude include those grounded in 'basic liberties' (e.g.Oberman 2016), liberal commitments to the fundamental equality of persons (e.g.Carens 2013; Cole 2000; Sager 2020), and concerns over the exercise of coercive power over 'outsiders ' (Abizadeh 2008).
The differences between these positions on the general question of whether states should have a (unilateral) right to control access to membership, and on what grounds, are substantial.However, many of them share the same fundamental assumptions about the nature of borders and border crossing.These assumptions combine to create a border imaginary that is fundamentally disrupted by the rich work of critical border/ migration studies scholars.Briefly, under this border imaginary it is clear where the borders of one state end and another begin; migrantsrecognised, at least in theory, as autonomous, rights-bearing individualsengage with one state/entry regime at a time; and immigration controls are subject to democratic oversight, and this is what gives them their legitimacy.
However, as has been extensively documented by critical border/migration studies scholars, the digitisation (and attendant externalisation) of borders have made it almost impossible to physically locate the borders of individual states.Migrants enter a virtual world of Big Data, interconnected databases, satellite-communications networks, and artificial intelligence, in which they no longer interact with a single border but with the amorphous and mobile 'borders' of dozens of states simultaneously, obscuring border/immigration decision-making and accountability for those decisions (Shachar 2020).Rather than rights-bearing, autonomous individuals, migrants become amorphous collections of data traces.These 'Dividuals', or 'Data Doubles', are the objects of border control decision-making, even as the consequences of those decisions are felt by embodied individuals.Decisions are increasingly being made using risk-profiling software, driven by algorithms shielded from democratic oversight by intellectual property rights, on the one hand, and the sheer mathematical complexity of their operation on the other (e.g.Ajana 2015; Allen and Vollmer 2018;Amoore 2006;Andersson 2016;Aradau and Blanke 2022;Leese 2014;Walters 2006).The development of such digital control strategies has greatly facilitated the ability of wealthy, predominantly Western, states to outsource, or externalise, their border controls to the territories and authorities of other, usually poorer, states.Thus, where 'the border' of any given state is changes depending on who you are, where you come from, and why (it is perceived) you are trying to cross.To date, insufficient attention has been paid by normative theorists of migration to these fundamental realities.
At this point, a protagonist in the immigration admissions debate might object that we are talking about two different things here: they are talking about general principles of immigration policy, we are talking about particular border control practices.Not only are the two not the same, but one is ethically prior to the other: general principles of immigration policy are ethically prior to the specifics of border control.The implication of this objection is that detailed attention to the technologies (including social technologies) of border control is unnecessary for high level theorising about migration justice.
We disagree.General principles of immigration policy and specific borders/border controls may not be the same, but immigration policy and border controls are much more intimately connected than the objection would suggest.First, research on borders reveals issues that have been insufficiently considered.Even if a general right to restrict immigration can be established, how immigration is restricted still raises pressing moral questions, such as privacy and discrimination, that demand independent consideration.More significantly, we cannot assess the legitimacy of actual border controls without an understanding of how borders are actually enforced.For example, most theorists who hold that states have broad rights to restrict immigration make exceptions for refugees.These exceptions require a fair legal process to determine if a person is persecuted.If border control technologies prevent most refugees from making their case or if the use of risk-profiling software leads to the unjust rejection of legitimate claims, the right to asylum is undermined.
Second, the nature of immigration policy and the ideas that justify itamong them, sovereignty, democracy, freedom, and distributive justicedepend on the nature of borders.Normative concepts are embedded in a border imaginary.When authors argue that 'communities of character' or 'liberal nations' have a right to restrict immigration, they assume these communities or nations are bounded by state borders and exercise power over a defined territory.When open borders advocates contend that coercive border controls sit uneasily with a commitment to equality and liberty, what does it mean if the 'subject' of border control is not a rights-bearing individual, but rather a 'Dividual' constructed from millions of data points scraped from the Internet?
On a third more practical note, the immigration admissions debate has been raging since the publication of Walzer's Spheres of Justice and Carens' 'Aliens and Citizens', in the 1980s, all the while states have gradually been expanding the control they try to assert over global movement, often with deadly consequences.At some point, sustained attention needs to be paid to the actual practices of border control, 1 not just the more abstract question of whether a community should be able to exercise unilateral control of their membership policy.
Critical border/migration studies also stand to benefit from closer engagement with normative theorising.Critical scholars excel at revealing and criticising how borders are enforced, but offer limited guidance for a principled, systematic response to their criticisms.While some of these theorists are guided by abolitionist visions of a world without states or nations and make gestures toward the notion of a global commons (Anderson, Sharma, and Wright 2009), the question of migration justice is comparatively undertheorized.Furthermore, whatever one's views about the idea of a borderless world, there is an urgent need for immediate actions and reforms that can be defended by clear appeal to normative principles.Normative theorising has something to contribute here.This contribution need not necessarily act as a sticking plaster over an unjust system but can itself help guide the development of the 'non-reformist reforms' favoured by abolitionists (e.g.Bradley and De Noronha 2022).
The purpose of this symposium, then, is to foster dialogue between critical border/ migration studies and normative theory, with particular attention to rapidly evolving border control technologies.It emerged from two panels on the Ethics of Border Controls in a Digital Age, at the British International Studies Association that took place from 29 to 30 August 2022.Taken together, the papers in this symposium aim, first, to outline some of the key technologies that are employed and being developed to manage global migration and border control; second to examine the ethical challenges these technologies raise, both for those on the move and for states claiming to be liberal democracies committed to human rights; and third, to provide theoretical resources for thinking about how to address these challenges.
While states have long known that the ability to control borders rests on how well they coordinate with each other, normative theorists have tended to treat the questions relating to border control as the concern of singular states and their citizens.Saunders argues that the rapid expansion of digital border control technologies, and the further externalisation of borders that they have facilitated, requires us to think about the ethics of border controls from the perspective of global interdependence.Practices of profiling, biometric identification, and the data sharing that these, and other digital practices, require have built a global web of migration control, in which travel in general, not simply settlement, are managed.These practices pose significant problems for all those who seek to movewhether for settlement or simply for tourismbut most especially for citizens outside of 'the West'.For normative reasoning about migration to be relevant, it must tackle the fundamentally interdependent nature of migration management, and the emerging new realities of our lives as digital subjects.
The emergence of a global digital infrastructure of migration management has been made possible by the simultaneous development of domestic infrastructures of digital control.Narita's paper puts these infrastructures under the microscope, arguing that unearthing the infrastructural systems through which immigration policy is managed allows us to ask different kinds of questions about the ethics of border controls.Examining the UK's 'Hostile Environment' policy, the paper shows how the infrastructures of border control create a form of path dependency for immigration policy, which impacts the building (or possible dismantling) of borders into the future.Ethical reflection on borders, then, has to focus not only on discriminatory or otherwise unjust policies, but needs to attend to the ethics of the infrastructures through which these policies are developed and deployed.
Picking up on the temporality of border controls, Kaneti tackles the issue of Artificial Intelligence, taking inspiration not only from existing attempts to deploy AI in border control but how states and corporations may try to use AI in the future.Asking 'What Can AI See?', through experimenting with how open source generative AI programmes produce images of 'migrants', Kaneti foregrounds the ontological and epistemological challenges of how we 'teach' AI about 'migrants', thus impacting what AI can therefore 'know' about 'migrants'.These foundational questions of teaching and knowing are vital as AI increasingly 'informs, organises, and predicts the world we see and understand'.
Turculet tackles a theme which threads through all the papers in the symposiumthe status of the individual subject in a rapidly digitising era.Her paper reconnects the 'deep dives' of Narita and Kaneti back to the 'big picture' outlined by Saunders.At the root of ethical inquiry is, for Turculet, the fundamental question of the subject.In order to ask 'who owes what to whom?' we must be able to answer the question of 'who does what to whom?' This question is becoming increasingly difficult to answer as ever more spheres of life become subject to algorithmic decision-making, or algorithminformed decision making.The stakes involved in border-crossing scenarios are often greater as borderzones are already places where migrants struggle to see their rights protected and enforced.Digitising borders compounds what Turculet calls the 'problem of indeterminacy'in which it is not clear who is morally responsible.Drawing on Data Feminism, Turculet argues that the first step toward ethical digitisation must be identifying and rectifying potential problems of indeterminacy.
Another way of thinking about the problem of indeterminacy is in terms of potential for domination.Where it is not clear who does what to whom, and thus who owes what to whom, it becomes almost impossible for those who have been harmed to seek redress, leaving them at greater risk of domination.Sager's paper takes seriously the potential of digital border controls to dominate migrants, drawing on neo-republican accounts of domination.He proposes a non-domination-based ethics of digital surveillance and mobility.In order to be legitimate, digital border control technologies must, at the very least, avoid the arbitrary use of power through the creation of institutions that allow migrants to participate in decision-making and to contest policies and practices that have the potential to dominate them.
The final paper of the symposium takes us in a slightly different, but important direction, foregrounding the experience of researchers working on borders and migration.The increasing digitisation of global border control poses profound challenges for researchers working in and around border areasparticularly as the locations and scope of these borderzones expands (geographically and socially).While issues of method and research ethics will be intensely familiar to many social scientific researcherssuch as those working in critical border/migration studiesthey may be less familiar to normative researchers who rarely go 'into the field'.There is, however, an emerging trend in normative work on migration to engage 'on the ground' with migrants and states.Awareness of the digital environments of borders/borderzones and how they differentially impact on researchers and migrants is vital to ensure that this work protects both migrants and researchers.
The papers in this symposium will likely generate more questions than they answer about this important topic.Our hope is that these questions prompt sustained engagement between normative theorists and empirical researchers working on issues of migration, borders, and digital technologies.