The Walls Have Eyes: Surviving Migration in the Age of Artificial Intelligence

By Petra Molnar

With a foreword by E. Tendayi Achiume

A chilling exposé of the inhumane and lucrative sharpening of borders around the globe through experimental surveillance technology

In 2022, the U.S. Department of Homeland Security announced it was training “robot dogs” to help secure the U.S.-Mexico border against migrants. Four-legged machines equipped with cameras and sensors would join a network of drones and automated surveillance towers—nicknamed the “smart wall.” This is part of a worldwide trend: as more people are displaced by war, economic instability, and a warming planet, more countries are turning to AI-driven technology to “manage” the influx.

Based on years of researching borderlands across the world, lawyer and anthropologist Petra Molnar’s The Walls Have Eyes is a truly global story—a dystopian vision turned reality, where your body is your passport and matters of life and death are determined by algorithm. Examining how technology is being deployed by governments on the world’s most vulnerable with little regulation, Molnar also shows us how borders are now big business, with defense contractors and tech start-ups alike scrambling to capture this highly profitable market.

With a foreword by former UN Special Rapporteur E. Tendayi Achiume, The Walls Have Eyes reveals the profound human stakes of the sharpening of borders around the globe, foregrounding the stories of people on the move and the daring forms of resistance that have emerged against the hubris and cruelty of those seeking to use technology to turn human beings into problems to be solved.

• Language: English
• Publisher: The New Press
• Release date: May 21, 2024
• ISBN: 9781620978672

Petra Molnar

Petra Molnar is a lawyer and anthropologist specializing in migration and human rights. She co-runs the Refugee Law Lab at York University and is a faculty associate at Harvard’s Berkman Klein Center for Internet and Society. Petra has crossed many borders and worked all over the world including Jordan, Turkey, the Philippines, Kenya, Colombia, Canada, and various parts of Europe. Petra’s work has appeared in the New York Times, Al Jazeera, Wired, The Guardian, and many other outlets. The author of The Walls Have Eyes: Surviving Migration in the Age of Artificial Intelligence, she splits her time between Toronto, New York, and Athens.

Book preview

Cover: The Walls Have Eyes, Surviving Migration in the Age of Artificial Intelligence by Petra Molnar

THE WALLS HAVE EYES

Surviving Migration in the

Age of Artificial Intelligence

PETRA MOLNAR

Logo: The New Press

To all those who migrate

by choice or by force,

in hopes of a more joyful world

There are places in the world where rules are less important than kindness.

—Carlo Rovelli

There’s really no such thing as the voiceless. There are only the deliberately silenced, or the preferably unheard.

—Arundhati Roy

Contents

Foreword

Author’s Note

Introduction

1:The Wall Bleeds Rust: Robo-Dogs in the Sonoran Desert

2:Smart Borders Kill: Technological Violence at the Fringes of Europe

3:If We Go There, We Will Go Crazy: Refugee Camps as Digital Prisons

4:Recognizing Liars: AI Lie Detectors, Voice Printing, and Digital Incarceration

5:Data Is the New Oil: The Silicon Savanna and Data Colonialism in East Africa

6:All Roads Lead to Jerusalem: A Lucrative Border Industrial Complex

7:The Politics of Exclusion and Fear

8:Strategies of Resistance

Afterword

Acknowledgments

Reading List for Uncertain Times

Notes

Foreword

As I sit down to write this foreword in Silicon Valley, I am surrounded by countless billboards extolling the virtues of artificial intelligence, the apparent panacea for all our societal ills. Data Powers Business, reads one. Being in the epicenter of so much technological development, I cannot help but be struck by the messages that do not make it to the billboards—the messages advertising, for example, how power and cutting-edge technology go hand in hand, perpetuating vast divides between those who benefit and profit, and those against whom the technology becomes a weapon. The Walls Have Eyes is a powerful and creative intervention in this regard, showing the strength of storytelling as a form of resistance, while revealing the realities of technological border violence that will never make it to the billboards.

In my work both as a law professor focusing on the international law of migration and displacement and as the former United Nations special rapporteur on contemporary forms of racism, racial discrimination, xenophobia and related intolerance, as well as in my own experiences as a migrant, I have seen firsthand the impacts of an increasingly dystopian world order replete with terrifying border technologies. Ironically, some technologies are marketed to signal our transition to a better, gleaming future. Others more clearly signal their exclusionary function. Drones equipped with tasers, iris scanners in refugee camps, and massive surveillance infrastructures at nearly every frontier are just a few examples woven into the tapestry of The Walls Have Eyes.

When I was first beginning to grasp the rapid rise of what some have come to call digital borders, everyone I spoke to recommended I connect with Petra, who then and now remains (among other things) the leading legal expert on the human rights implications of digital border violence. I eventually had the good fortune to meet her. Her expertise was unparalleled, but even more than that, I discovered that she was a force to be reckoned with in the best kind of way. Petra was clearly driven by a passion for radical border justice and a vision that centered the agency and self-determination of those subject to the most oppressive regimes of border injustice. The Walls Have Eyes is only one manifestation of this passion and vision.

Since meeting her, I have eagerly followed Petra’s work from border to border, witnessing the careful stitching together of a truly global story of power, violence, innovation, and joyful resistance from one of the leading thinkers on these issues. I have had the privilege of witnessing the birth of this book from its early inception nearly four years ago, when Petra and I first started collaborating and dreaming up our own project, the Migration and Technology Monitor—an archive and a community striving to foreground and uplift the expertise of people on the move to tell their own stories of border violence, digital borders, and the contestation or even co-optation of new technologies.

Technology has the power to liberate but it also has the power to decide who lives and who dies. Thousands have already succumbed to manufactured graveyards produced by governmental policies from the Mediterranean Sea to the Sonoran Desert, punished for the very human impulse to seek safety and provide for their families. This impulse must be front of mind for us all as our planet faces an environmental reckoning.

Technology and its development cannot be divorced from historical structures of oppression, labor exploitation, and various forms of imperialism, including settler colonialism. There can be no real way out of the most pressing global crises without meaningfully addressing these legacies precisely because technology is neither neutral nor objective. It is fundamentally shaped by the racial, ethnic, gender, and other historically rooted inequalities prevalent in society—often exacerbating these inequalities. Significant responsibility for ensuring a just future lies especially with those nations that, in the past and the present alike, benefit from the power structures that now determine the terms of technological innovation, often with little regard for the vast majority subject to those innovations.

The Walls Have Eyes could not be more timely. We desperately need human stories to illuminate the very real harms that border technologies create and perpetuate: stories like that of Little Nasr, a teenager with severe scoliosis at the Greece-Turkey border trapped not only by racist border policies but also an ever-growing surveillance dragnet that ensnares all who come near. Or that of Zaid, who poignantly reflects on his journey, surviving pushbacks, surveillance, and dehumanization at the hands of the European asylum system. The Walls Have Eyes is unique and courageous, weaving together powerful personal stories with Petra’s razor-sharp analysis.

Ultimately, Petra shows that any discussion of border technology must center its systemic harms, while insisting on upholding fundamental rights. We must pay attention to who gets to speak on these issues, and we must listen for the silences. Affected communities and civil society must drive the conversation around the development of technology in migration and border spaces. The tech community—policymakers, developers, and critics alike—must push the conversation beyond reform and toward the abolition of technologies that cause harm, destroy communities, separate families, and exacerbate the deep structural violence continually felt by Black, Indigenous, racialized, LGBTQI+, disabled, and migrant communities. Note that technologies deployed and perfected at the border do not stop there. The AI lie detectors, surveillance towers, and robo-dogs that The Walls Have Eyes so deftly describes are bleeding over into other facets of public life that impact us all.

The Walls Have Eyes is a clarion call to action; an invitation not to give in to despair, but rather to follow the lead of those on the frontlines of border violence, including those who show up time and again to expose and resist human rights abuses.

E. Tendayi Achiume

Alicia Miñana Professor of Law at the

University of California, Los Angeles

Former United Nations Special Rapporteur on Contemporary

Forms of Racism, Racial Discrimination,

Xenophobia and Related Intolerance

Author’s Note

The Power of Storytelling as an Act of Resistance

Borders are violent. Yet they are also spaces of tremendous resistance and solidarity, often in very unexpected ways. Storytelling and story sharing is one form of resistance—and a profound and crucial element in any attempt to illustrate the opaque world of border technologies. Over the years, people have been incredibly generous in sharing their stories with me, often in the most difficult of times. Being entrusted with other people’s stories is a great privilege, one which I do not take lightly. But collecting stories is not without its pitfalls. How, for instance, can I offset its inherently extractive nature? How can I avoid potentially damaging tropes? Yes, it’s true that we are slowly moving beyond victimization narratives and the disempowering notion that those who have been marginalized need to be given voice. Indeed, as writer Arundhati Roy reminds us: There’s really no such thing as the ‘voiceless.’ There are only the deliberately silenced, or the preferably unheard. But these difficult questions of who gets to speak to which issues nonetheless remain.

This discomfort sits at the center of this book—but it is a discomfort I do not want to turn away from. While I do have my own migration experience and hyphenated identity that always motivates my work in this space, I do not and cannot attempt to speak for all. Nor do I want to fetishize human suffering or add to the spectacle of violence at the world’s borders. Instead, what helps me is the idea of story stewardship. An author like Potawatomi professor and science writer Robin Wall Kimmerer serves as a useful example. Her storytelling is a profound meditation on the concept, drawn from various Indigenous traditions, that recognizes the storyteller’s obligation to acknowledge their own power and privilege over other people’s narratives, and to hold space for stories to tell themselves.

With each conversation and growing relationship, I found these issues of witnessing, extracting, and capturing becoming more complex as I grappled with how best to convey the atrocities that continue to occur near shores and borders around the world without becoming part of the exploitative surveillance machine myself. It is my hope that by infusing this book with as much of the lived experiences of the people I worked with as I can, it will illuminate the testing grounds of these dangerous technologies and bring them from the realm of the abstract to the realm of the real. Documenting this system itself is risky business—trying to pin down opaque decisions, tracking down secretive private-sector players, and personally witnessing horrific conditions.¹ Often it’s hard to do it well in these difficult conditions. Equally fraught, however, is the question: How does one carefully and critically document very real human rights abuses that are by their very nature difficult to see?

Unfortunately, I often see researchers and journalists taking photographs of people on the move when they are at their most vulnerable, or gathering stories without thinking through the tremendous responsibility this entails. I have learned much about the ethics of visual imagery from documentarians and filmmakers who I have been lucky to work with. I was on the Greek island of Lesbos after a horrific fire destroyed Moria, the largest refugee camp in Europe, in the fall of 2020. Thirteen thousand refugees were confined to a sunbaked stretch of road and teargassed as they were evacuated, and then forcibly transferred into a hastily constructed new camp. While on an official tour of the new camp, I witnessed photojournalists hurriedly pushing their cameras into a tent filled with laughing children, and another taking moody portraits of a Black man bathing in the ocean.

All photographs in this book were carefully selected so as not to replicate problematic depictions of racialized bodies that are so common in representations of migrants. As we should consider: What does it mean to press the shutter or write the story, to freeze, capture, and archive a moment? Who gets to decide which images become symbols of crises? How can we make the truth known while also preserving individual integrity? Whose perspectives matter and which priorities take precedence? What does critical representation and meaningful participation look like? To avoid the usual tropes of poverty porn, showing people in distress in what may be the worst moment in their lives, we made sure that, unless a person is a direct participant in our projects, their face is not shown. This choice is not meant to dehumanize. Rather, it is meant to return at least some agency to people on the move. I am indebted to these learnings as well as the Forced Migration Review, which has a similar photo policy that I have adopted in my work.

In order to tell this global story of power, violence, and technology, I rely on a sometimes uneasy mix of law and anthropology. It is a slow and trauma-informed ethnographic methodology, one which requires years of being present in order to begin unraveling the strands of power and privilege, story and memory, through which people’s lives unfold. Sometimes an interview isn’t really even an interview—rather, it is sharing a meal together on the floor of a container inside a refugee camp, or a silent walk along the seashore. This practice does not yield immediate insights—and indeed, researchers (myself included) have often faced scrutiny about the efficacy and accuracy of our methods or even the validity of what we are doing for such long stretches of time in the field.² But it is through this slow unpicking that the real impacts of borders come to light.

This methodology is also of course incomplete, and for every story, space, and context that is included here, there are missing pieces, silences. The chapters are snapshots in time and space, giving readers the opportunity to make their own connections about the stories told and the stories missed. Every person named in this book has confirmed they wish to appear, but some names and identifying features have been changed to protect people’s safety. To honor someone’s wish not to be named, sometimes I use singular pronouns when describing a particular border trip, even when there was more than one person present. Ethnographies differ from journalism in that they do not strive to be objective—indeed, no piece of writing is. The narratives in this book are all filtered through my lens, colored by my background, experiences, and blind spots, recognizing that memories are malleable and socially constructed. Moreover, in light of the ultra-rapid development of new technologies, I have tried to be as current as possible, but book publishing is a slow and deliberate process, and so whatever appears on these pages may have been already eclipsed by newer technologies. Any mistakes are, of course, my own.

Given the importance of voice and story, the choice of terminology throughout this book is deliberate. While rigid categories of refugee, asylum seeker, and migrant are used in law and policy to create particular narratives, in reality, these categories often elide much of the story. Therefore, wherever possible, in my work and in this book I use the term people on the move or people crossing borders to expand the terminology that people commonly use when they discuss human migration and how migration management technologies are experienced. This deliberate use of more inclusive terminology does not seek to undermine the fact that refugees face particular vulnerabilities and often experience higher risks and harms as a result of migration management technologies. It simply highlights that at the end of the day there are people at the center of it all. Moreover, all of us may in one way or another be affected by migration management technologies as we cross borders and move across the world. While its greatest impact is on traditionally marginalized communities such as refugees and asylum seekers, the ecosystem of migration management technologies affects us all.

A high-tech refugee camp opens on the Greek island of Kos, December 2021. (Petra Molnar)

Introduction

The Growing Panopticon of Border Technologies

Borders are both real and artificial.¹ They are what historian Sheila McManus calls an accumulation of terrible ideas, created through colonialism, imperial fantasies, apartheid, and the daily practice of exclusion. Today, there are millions of people on the move because of conflict, or instability, or climate change, as well as for economic reasons. People crossing borders whether by force or by choice are often talked about in apocalyptic terms, as a flood or a wave, terms that are underscored by xenophobia and racism. Migration turns the act of crossing a border into a criminal act, and the person on the move into the embodiment of illegality.

In recent years a new frontier has emerged at the cruel intersection of racism, technology, and borders. Some of the technology is old. Passports and physical border walls have always been used to separate and exclude people, but new technologies are making their way into immigration and refugee processing, now at a faster rate than ever before. Decisions such as whether to grant a visa or detain someone, which would otherwise be made by administrative tribunals, immigration officers, border agents, legal analysts, and other officials, are now made by machines through algorithms. Predictive analytics, which use large data sets to make predictions about human behavior, are used both in humanitarian emergencies to deliver aid and to project where people may be crossing borders. There has also been a concurrent rise in the use of biometrics, or the automated recognition of individuals based on their biological and behavioral characteristics. Biometrics can include fingerprint data, retinal scans, and facial recognition, as well as less well-known methods using technology that can recognize a person’s vein and blood vessel patterns, ear shape, and gait. Even more experimental are lie detectors relying on AI to determine who is telling the truth at the border, while voice-printing technologies analyze accents and patterns of speech. Meanwhile, the surveillance dragnet is only continuing to expand, with a growing arsenal of cameras, blimps, loud sound cannons, and even robo-dogs deployed to control borders.

Consequently, the color of your skin, the accent in your voice, and even your body become a passport read by an increasingly automated border regime that excludes some while welcoming others. Already violent global border policies are sharpened through the use of digital technologies developed for the purposes of border control and migration management.² These technologies separate families, push people into life-threatening terrain, and exacerbate the historical and systemic discrimination that is a daily reality for people on the move, especially those from marginalized communities—this is all too familiar if you’re Black or Brown or otherwise deemed suspicious and have had many uncomfortable interactions at airports or with immigration officers.

The increasing allure of using technological interventions at and around the border has very real impacts on people’s lives, which are exacerbated by a deliberate lack of meaningful governance and oversight mechanisms for these technological experiments. Border spaces serve as testing grounds for these new technologies, places where regulation is deliberately limited and where an anything goes frontier attitude informs the development and deployment of surveillance at the expense of people’s lives.

Unregulated technological experimentation is often justified under the guise of efficiency and security. These projects also play into a growing techno-utopian drive, weaponizing innovation against the unwanted for the benefit of those in power, employing tropes of security and making determinations of who deserves to be safe and who is a threat. The locus of power and technological development is concentrated in North America and Europe, primarily to be deployed on the so-called Global South.³ Labor exploitation is also alive and well, fueling data centers in places like East Africa and Latin America, key physical infrastructures that house the computing powers of the West, running on cheap labor that are never permitted to leave their countries. The development of technology also reinforces power asymmetries between countries and influences our thinking around which countries can push for innovation, while other spaces, such as conflict zones and refugee camps, become sites of experimentation and exploitation. The commitment to technological supremacy is also an approach that silences other possible solutions. The hubris of Big Tech and the allure of quick fixes do not address the systemic reasons why communities are marginalized and why people are forced to migrate in the first place.

My desire to understand how the interconnected systems of power, history, labor, politics, and economics underpin these technologies and their impacts throughout a person’s migration journey is what ultimately brought me to the world’s borders to get a firsthand look. But conversations about technology can be difficult to enter into—and indeed, people from different backgrounds and disciplines are often excluded because the field seems so specialized, its terms so inaccessible. Artificial intelligence, machine learning, automated decision-making systems, and predictive analytics are a group of overlapping terms and refer to a class of technologies that assist or replace the judgment of human decision-makers. Different disciplines and regulatory mechanisms also use different terms.⁴ Automated decision-making systems draw from fields like statistics, linguistics, and computer science, and use techniques such as regression, rules-based systems, predictive analytics, machine learning, deep learning, and neural networks, often in confusing combination with one another. For instance, five years ago, I didn’t even know what algorithms were, let alone how they are being deployed to manage people on the move. Nor did I particularly care, because when the person you are representing in court is getting deported tomorrow, you feel like you do not really have time to care about something as abstract and far away as surveillance and AI. But now, attention to such technology is central to my work; it’s a useful lens through which to try and understand how power operates in society.

While various technological innovations swirl around us today, automation and robotics have been around for a while. Ismail al-Jazari, the Muslim inventor from Anatolia (present-day Turkey) and whom some call the father of Robotics, first started documenting his various inventions in the late twelfth century. I wonder how he would feel about his work giving rise to robo-dogs and drones used to hunt people escaping persecution from the Middle East. People have long been fascinated by robots. But the word robot is actually very new to the English language. Its origins can be traced back to a Czech play, R.U.R. (full title: Rossum’s Universal Robots), written by the famous playwright Karel Čapek, in 1921; it is a word etymologically related to robota, meaning work. I grew up reading Čapek in elementary school but had no idea of his influence in this area.

But how does automation actually work? It works through algorithms, the basis for automated decision-making. Algorithms are what researcher Tarleton Gillespie calls a sets of instructions, similar to recipes, that organize data and act upon it to produce a desired outcome.⁵ Certain algorithms are trained using a large corpus of existing data, which allows them to classify and generalize beyond the examples in the training set.⁶ The training data can be a body of case law, a collection of photographs, or a database of statistics, some or all of which have been categorized according to the designer’s criteria.

Scholars, activists, and technologists have rightly criticized algorithms for being inscrutable.⁷ It is difficult to pry them apart and figure out how exactly they come to their decisions. This is because an algorithm’s source code, its training data, and other inputs may be proprietary and thus shielded from public scrutiny on the basis of intellectual property law or because they are confidential business assets. When algorithms are used in immigration and refugee matters and rub up against issues of national security, both input data and source code may be classified. And because people are not able to scrutinize the data to understand how the algorithms make decisions, the logic of the algorithms becomes less and less intelligible to humans.⁸ One of the main concerns about this situation is the introduction of bias. Algorithms are vulnerable to the