Petra, in your book "The Walls Have Eyes”, you write that "robot dogs" are being used to monitor the U.S.-Mexico border. What are your concerns about this technology?

Petra Molnar: Robo-dogs are just the latest manifestation of a growing global arsenal of border technologies. Announced in February of 2022 by the US Department of Homeland Security, these military grade machines are part of an increasingly vast surveillance dragnet that extends the smart border system at the US-Mexico border, pushing people who are crossing the border into dangerous terrain, leading to increased deaths as people take riskier routes to avoid detection. Technologies like robo-dogs are also unregulated and introduced without public consultation or transparency, making the border a perfect testing ground for high-risk technological experiments.

The robot dogs are just one example of how AI is being used for border surveillance. You also mention autonomous border surveillance systems, lie detectors, or AI-driven surveillance in refugee camps. It seems that AI technology is playing an important role in border protection against migration.

Molnar: Yes, artificial intelligence and automation is increasingly used in various aspects of migration and border control. From drones to voice printing technologies to experimental projects like AI-lie detectors, algorithms and automated decision-making have found their way into virtually every aspect of migration control. The global AI-arms race is motivated by the turn to techno-solutionism (or the reliance on technical solutions to complex societal issues) and the border and migration spaces are some of the most high-risk spaces where this experimentation occurs.

Is the use of AI technologies at the border under-regulated? In other words, is the US-Mexico border, for example, a testing ground for AI technology?

Molnar: Very little to no regulation currently exists to put up guardrails around the development and deployment of AI-driven technologies at the border. Even the recently implemented Act to Regulate Artificial Intelligence of the European Union does not go far enough to ensure that people’s human rights are being respected, nor does it ensure sufficient commitment to public transparency and oversight. And instead of the EU’s AI Act setting a strong global precented for other jurisdictions like the US and Canada, there is once again to incentive to regulate border technologies. This lack of global governance is in contravention to United Nations guidance and we need red lines or bans, or at the very least a moratorium on the development and deployment of harmful border technologies that are hurting real people.

In your book, you describe how the private sector is increasingly influencing border technology. How do you see the role of technology companies like NVIDIA and Palantir in this context?

Molnar: The private sector is a major player in normalizing technological incursions at borders and in migration. Borders have become a big business, spawning a nearly 70 billion dollar border industrial complex, where tools like robo-dogs, drones, and AI-powered surveillance are presented as solutions to the “problem” of migration. Companies like Palantir and NVIDIA, but also Israeli surveillance firms like Celebrite and Elbit systems are just some of the global players entering into opaque public-private partnerships with states, influencing what we innovate on and why. The norms being set by the private sector coupled with a securitization agenda that sees migration as a problem to be solved means that more and more technologies are being weaponized against people-on-the-move, instead of being used to audit immigration decision-makers or to root out racist border guards. This is a normative choice that a powerful set of actors is making, one that ends up hurting thousands of people who are exercising their internationally protected right to seek asylum.

The term "smart wall" suggests efficiency and precision. Can you explain how these technologies actually contribute to security and where they fall short?

Molnar: The term “smart” wall is a bit of a misnomer. It refers to the augmentation of more traditional physical methods of border infrastructure with automated decision-making, AI, and autonomous surveillance technologies. This web of technology can include drones, cameras, fixed surveillance towers, and even more experimental technologies like robo-dogs of Long Range Acoustic Devices (or sound cannons). However, studies by scholars such as Geoffrey Boyce and Samuel Chambers, along with my own work, have shown that these interventions do not work to deter people. Instead, they force people towards more dangerous terrain, nearly tripling the number of deaths at the US-Mexico border for example since the inception of the so-called smart wall.

Is the use of AI at borders fundamentally racist or discriminatory?

Molnar: AI is a social construct, just like any other piece of technology. It therefore replicates biases and discrimination inherent in our society. AI in particular has been shown to be racist and discriminatory towards people of colour, women, or people who are differently abled. These biases can have tremendous impacts for people at the border or in migration spaces. For example, biased algorithms can place people in immigration detention (as happened in the United States and Australia) or deny them visas or entry (The Netherlands, the UK and Canada have experimented with these types of technologies). This type of discrimination can separate families, place people in detention, or even result in their deportation, possibly putting them at risk.

You want your book to be seen as a wake-up call. What are the most important measures you believe should be taken to mitigate the negative effects of AI at the border?

Molnar: In my book , I chronicle 6 years of work across various borders, trying to understand how technologies are shaping the way people cross borders. But at the centre of it all are real people and communities being harmed by this technological violence. My book 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. We desperately need more governance and regulation around the use of technologies at the border, and conversations must include the viewpoints and expertise of affected communities who are at the sharpest edges of this new technological border regime.

You have conducted research in several countries and border regions. Can you identify differences in the use and handling of AI technologies at different borders? Which countries do you think are using particularly problematic technologies?

Molnar: My book The Walls Have Eyes takes us from the US-Mexico border to Greece to the Polish/Belarussian border, to Kenya, to the Occupied Palestinian Territories. Each of these regions use various technologies for border management and migration control. While the US-Mexico border and the EUdo have certain similarities, especially with the use of AI-powered surveillance, the EU has also been piloting high-tech refugee camps on the Greek islands such as Samos and Kos, replete with biometrics and drones and automated risk assessments. Kenya was also an interesting case study through which to look at digital colonialism, and the extraction of data in digital ID systems that disenfranchise millions, while also giving rise to mass biometric data collection programs in refugee camps such as Kakuma and Dadaab that have been around for decades. And lastly, this global story would not be complete without a visit to the Occupied West Bank, where Israel has been testing various surveillance technologies on Palestinians for decades, later exporting them out and selling these very same technologies to the US government for border enforcement at the US-Mexico border as well as for maritime surveillance in the EU.

Interview: Helmut van Rinsum

Petra Molnar is a lawyer and anthropologist specializing in migration and human rights. A former classical musician, she has been working in migrant justice since 2008, first as a settlement worker and community organizer, and now as a researcher and lawyer. She has worked all over the world including Jordan, Turkey, Philippines, Kenya, Colombia, Canada, Palestine, and various parts of Europe. She is the co-creator of the Migration and Technology Monitor, a collective of civil society, journalists, academics, and filmmakers interrogating technological experiments on people crossing borders. She is the Associate Director of the Refugee Law Lab at York University and a Faculty Associate (and former Fellow) at the Berkman Klein Center for Internet and Society at Harvard University.