For the 30 million refugees in the world today, a United Nations High Commissioner for Refugees (UNHCR) ID Card is a golden ticket to food, shelter, education, and cash, all provided by the helping hand of the international community. In addition, these ID cards initiate the process of applying for asylum or resettlement. What does it cost to earn one of these extraordinary documents? Only the rights to your own photograph, iris scans, fingerprints, personal and family history, health data, and all available prior legal documentation. According to a 2020 UN Report, over 37 million refugees had indeed given up this information and been registered with the UN’s Biometric Identity Management System (BIMS).
While there are many other intergovernmental organizations (IGOs), non-governmental organizations (NGOs), government agencies, and private organizations crucial to the provision of essential refugee services (many of whom have their own systems for collecting and monitoring biometric data), the UNHCR in particular is at the center of protecting refugees’ rights and safety. Biometric data helps the UNHCR in this regard by tracking movement and size of vulnerable populations, digitizing and saving personal documents, streamlining the resettlement process, and even assisting with healthcare delivery. This new intervention is part of a long history of technological advancements in government, each step of which is always accompanied by a complex ethical debate. When it comes to refugee assistance, how do aid organizations, which are often funded by governments that are indirectly responsible for their displacement, balance those refugees’ right to privacy against the very real bodily threats that they face? How much data can governments ethically collect? In order to understand these questions, it is important to understand the history of data collection as a means to control migration, the methods for using the new technology, and foreseeable future risks.
When it comes to refugee assistance, how do aid organizations, which are often funded by governments that are indirectly responsible for their displacement, balance those refugees’ right to privacy against the very real bodily threats that they face?
A Brief History of Data and Refugee Population Control
Using data and identity documents to control migration is both a very new and a very old phenomenon. For those living in a developed country in the 21st century, having an ID, passport, or other identity document seems completely natural, so much so that it would be difficult to imagine life without it. These kinds of personal documentation, however, were not widely used before the interwar period. As discussed by John Torpey’s The Invention of the Passport, passports were developed to regulate the massive flows of migration after the First World War and the tumultuous consolidation of nation states. However, displaced Russians and Armenians without a government to claim them were a problem for this new system. In response to the needs of these first contemporary refugees, the League of Nations issued them Nansen passports, which obligated signatory countries to confer citizenship rights to the owner while within their borders. This was officially the start of what many call the international refugee regime, whereby international organizations track and monitor all displaced persons.
The UN first began using biometric data to identify refugees in 2004 when they partnered with Microsoft to create user interface software called Project Profile and proGRESS. These technologies were implemented in 30 different countries to capture refugee photographs and basic biographical information. The stated goal was primarily to make identification easier, faster, and more uniform among various agencies and organizations. In addition, protecting valuable documents is difficult and sometimes impossible for those fleeing their homes. These technologies would help to alleviate concerns about damaged, stolen, or misplaced documents from a refugee’s home country, which has historically been a major barrier to receiving aid or resettlement. Aid workers could now perform a fingerprint, iris, or facial scan to match someone with their records and verify their identity.
According to a January 2020 UN report, eight out of every 10 refugees registered with the UNHCR, or about 37 million, had already been biometrically tagged and entered into the organization’s Global Distribution Tool (GDT).
The UNHCR began to expand Project Profile in 2013 with the development of Biometric Identity Management System (BIMS). Officially launched in 2015, this system uses facial recognition technology and iris scans in addition to fingerprints to seamlessly link individuals to their identification and personal documents. The registration process takes just seconds and allows the UNHCR to grant refugees a UN ID card. According to a January 2020 UN report, eight out of every 10 refugees registered with the UNHCR, or about 37 million, had already been biometrically tagged and entered into the organization’s Global Distribution Tool (GDT), which was developed along with BIMS to track the dispensation of aid amongst a variety of organizations. The UNHCR also rolled out a mobile app called UNHCR VERIFY-MY, which scans UN ID Cards and then displays the biographic data associated with the ID to provide secondary identification verification for aid-providers.
Amidst the development of all of these technologies that collect, store, and share refugee biographic and biometric data, the UNHCR also implemented its first major program that dispenses aid based on biometric data alone. Using technology from IrisGuard, the UN integrated iris scanning technologies into ATMs so that Syrian refugees could go directly to an ATM, scan their irises, and receive a direct cash allowance from the UNHCR. This solved the complex problems of transporting and delivering material aid and preventing fraud. The program was successful enough that the World Food Program (WFP) picked it up, and now Syrian refugees in Jordan can quite literally pay for their groceries with the blink of an eye. This technology, according to the WFP, “helps ensure that you deliver the intended benefit to the right person, at the right time, in the best way.” UNHCR reports of these programs have been overwhelmingly positive, and at the demands of donor and resettlement countries the use of biometrics for refugee processing has become almost ubiquitous worldwide.
Ethics of Data Collection
Scholars, activists, and advocates alike have raised concerns about the reliance on biometric data for providing humanitarian aid and immigration services. The first order of concern is whether or not the refugees are truly able to consent to the collection and use of their biometric data. While they technically have the right to refuse biometric scanning, there is a clear power dynamic between the refugees and the aid workers upon whom they may rely for food, housing, healthcare, and most importantly resettlement. This power dynamic has a large effect on coercing refugees into giving up their biometric data, even though they may distrust the aid workers. If they do refuse to complete biometric scans, they are effectively cut off from the registration process and unable to receive aid. This effect is multiplied in places where biometrics are used to receive cash or food from the grocery store, leaving refugees with little to no agency in this process.
Even if it were the case that refugees freely submit to biometric scanning, studies have shown that they are not adequately informed by aid organizations how, to whom, and for what purpose their data would be shared with the international community. Refugees interviewed by Caribou Digital expressed that language barriers and hasty practices kept them from understanding what paperwork was being filled out on their behalf or what legal status they had. This uncertainty only served to multiply their feeling of vulnerability. One interviewee remarked that not only did he not understand what it meant to be a refugee or asylum seeker, at the time of his registration he had never before seen a fingerprint scanner. No one explained to him what it was used for: “I thought they would burn me.” Similar interviews conducted in Ethiopia by The Engine Room researchers discovered refugees thought that aid workers were checking for health concerns when conducting iris scans.
One interviewee remarked that not only did he not understand what it meant to be a refugee or asylum seeker, at the time of his registration he had never before seen a fingerprint scanner. No one explained to him what it was used for: “I thought they would burn me.”
Even if refugees understand these “above the line” interactions with aid workers, they likely may not understand how their data is used and shared “below the line,” as demonstrated by investigative research from Caribou Digital. Within the UNHCR ecosystem alone, there are at least 50 registry and data management tools that share data. The World Food Program has its own program called SCOPE, which is used for distributing food aid as described above. Outside of this ecosystem, the UN also shares varying degrees of data with IGOs like the International Organization for Migration (IOM), UN Office for the Coordination of Humanitarian Affairs (OCHA), European Union Court of Justice, European Court of Human Rights (ECHR), EU European Border and Coast Guard Agency (FRONTEX), and the EU EURODAC program, among others. The extensive list creates a complex system of data sharing, making it difficult to trace where personal data is being spread, shared, and used.
The technologies that allow for biometric scanning are in many ways themselves discriminatory. Fingerprint scanners have a harder time reading those with darker skin or those who perform manual labor, making them more likely to have issues with false negatives or double entries and hence cutting them off from assistance. For Muslim women in Bangladesh, the process of taking iris scans or a headshot is invasive and frightening. Sometimes a refugee may have to walk long distances in order to reach a registration site, which can prevent pregnant women, the elderly, or disabled persons from registering. Because receiving food or monetary assistance is entirely contingent on each person registering with the UNHCR, these inequities of access could lead to significant disparities in wellbeing for the most vulnerable of an already vulnerable population.
Because receiving food or monetary assistance is entirely contingent on each person registering with the UNHCR, these inequities of access could lead to significant disparities in wellbeing for the most vulnerable of an already vulnerable population.
However, these challenges that refugees face must not cast them into too reductive of a victim narrative. In fact, in 2018 Rohingya refugees in Bangladesh staged protests against the UNHCR. They demanded that the UNHCR stop collecting biometric information until Rohingya refugee leadership could meet with the UNHCR to discuss the use of the data and the proper terminology of “Rohingya” was included on their UNHCR ID cards. They also demanded a guarantee that their data would never be given to Myanmar officials, who could potentially use it to harm repatriated Rohingya in the future. This fear stems from prior use of nationality verification cards (NVCs) by the Myanmar government to repress the Rohingya just on the other side of the border. These protests were important for helping the Rohingya gain agency in their own affairs. However, their efforts could not overcome the complex global bureaucracy of data collection and sharing, and UNHCR procedures remain unchanged.
Long Term Privacy and Security Implications for Refugees (Beyond Data Leaks)
Biometric technologies give aid organizations, host countries, and resettlement countries the ability to identify refugees, but it also allows them to single out those that they do not want to consider refugees. A key step of registering with the UNHCR and receiving aid is a process called refugee status determination (RSD), whereby the UNHCR and asylum countries determine whether or not an applicant qualifies for refugee status and asylum. During this process, they are forced to relive and even prove their own persecution in order to receive assistance from the international community, to the detriment of their own mental health. It also raises an important question — what happens to those who have not experienced enough or the “right kind” of trauma to meet the standards of the resettlement country?
There is a complex hierarchy of categories of migrants, including refugees, people in refugee-like situations, internally displaced persons, disaster-induced migrants, asylum-seekers, ‘economic’ migrants, or even illegals. Each of these categories are perceived as having varying levels of moral valence, and assistance is granted based on these designations. Therefore, while an internally displaced person and a refugee may have very similar experiences, they will be treated very differently by the international community. As an aid-worker using BIMS at the refugee camp in Kakuma, Kenya told researcher Fianluca Iazzolino, “There are many Kenyans to claim to be refugees [sic]. I understand that they too need help, but it should be the state to support them, not us. We cannot do the job of the state.” Biometrics and data collection help aid organizations and host countries to divide and categorize people in this way, picking out those who do not have the right identity to receive aid.
Biometrics and data collection help aid organizations and host countries to divide and categorize people in this way, picking out those who do not have the right identity to receive aid.
Beyond aid organizations, biometric data can be used by host or resettlement countries in alarming ways. While the official UNHCR policy is to share only minimal biographical information with host countries, that is not the case. As of January 2019, the UNHCR has started sharing all refugee biometric data with the US Department of Homeland Security (DHS), even the data of those whose application has been denied and who will never step foot in the US. The data is uploaded to the Homeland Advanced Recognition Technology System (HART), which also has the capability to share data with the Department of Defense (DOD), intelligence community, law enforcement, and even other countries. As data sharing continues to increase, a global database of political non-conformists is being created. Tracing back to the first form of refugee identity technology, the Nansen passport, refugees are definitionally a political outcast — someone who no longer fits into the global system of nation states. As Btihaj Ajana argues, refugee ID cards serve as a form of Agambenien inclusive exclusion, giving refugees access to food, shelter, and other forms of assistance while formalizing their status as outsiders.
Viewing refugees as outsiders or in some conceptions a threat to the nation state has contributed to the criminalization of refugees and asylum seekers. The DHS claims that sharing data with the DOD helps protect against fraud and stop potential terrorists from crossing United States borders (the UNHCR has rejected and even the CATO Institute has debunked this reasoning). The information is used to speed along processing of individuals they deem “low-risk” and put roadblocks to processing those they deem “high-risk,” including those who refuse biometric screening.
Consider refugees from South America and Mexico seeking asylum in the United States as an example. Even before President Donald Trump enacted policies to slow their immigration, these asylum-seekers faced significant barriers to achieving protected status despite the extreme violence and hardships that they fled, which were comparable in “severity” to that which refugees from other countries experienced. Political antipathy for Mexican and South American migrants and refusal to recognize them as refugees in public discourse is still one of those barriers. Even now that many of Trump’s policies have been reversed, asylum applicants are being processed one-by-one to sort out those who deserve asylum status, permanent resident status, or criminal status. This process of assigning a risk assessment to refugees is largely rooted in politics rather than real risk. That is not to suggest that an open border policy is a solution either. However, it is important to question how a system that already criminalizes those it aims to help could use intimate biometric information in the future.
Biometric data collection is becoming more and more normalized in migration policies and humanitarian aid practices, especially as COVID-19 has required the depersonalization of their day-to-day operations. In many instances, these technologies have helped to mitigate the risk of damaged or lost documents, provide humanitarian assistance, and speed up the processing of refugees and asylum seekers. However, running an efficient and cost-effective program is not the end goal for aid organizations — aiding refugees is. As biometric programs continue to grow it will be important to remember that a refugee does not become more worthy of assistance as more data about them is collected.
