Rethinking the Use of Biometric Systems for Refugee Management

In 2002, the United Nations High Commissioner for Refugees (UNHCR) first used biometric technology to manage the repatriation, or returning, of Afghan refugees to Afghanistan. After the Taliban regime was toppled in Afghanistan, UNHCR provided assistance packages to refugees who wanted to return home. Iris scanning was used to make sure that each refugee received a package only once, and to prevent ‘false refugees’ from trying to claim a package. UNHCR claims that using this biometric measurement system for refugee management in Afghanistan has been a success because it achieves the goal of confirming the legitimacy of claimants. Since then, UNHCR has been expanding its use of iris scans to gather personal data from refugees. In a recent article, Katja Lindskov Jacobsen questions the benefits of using iris scans to gather data about refugees. The author argues that UNHCR fails to recognize some significant possible risks associated with the use of iris scans for refugee management.

Although the successful use of iris scans to verify the identity of refugees in Afghanistan seems promising, the database typically only holds around 200,000 biometric templates. In a refugee crisis, such as the one unfolding in and around Syria, the database may need to hold millions of templates. As the size of the database grows larger, more false matches will be made.
With an iris scan expected error rate of three percent, the author calculates an expected 60,000 false matches in a refugee population of two million.

When false matches are made, it is often difficult for refugees to argue against the results because they may lack trusted identification documents to prove that there is actually a mistake in the system. This can jeopardize the living conditions of the refugees because gaining access to UNHCR assistance is contingent on passing the biometric check, as well as on the biometric system running smoothly. For example, a problem with the biometric registration system in 2013 led to the denial of assistance to 6,500 refugees in Mauritania.

A humanitarian mission by an international organization like UNHCR is usually funded by donor states and must be approved by host nations to operate. Donor countries and host nations may use funding and their rights-to-operate as leverage to access iris data collected by UNHCR. A humanitarian organization is then placed in a difficult position. If it refuses to provide the data, it will not be able to operate and provide assistance to any of the refugees. However, giving out personal data violates privacy, as there are no regulations that control how the data is used. Data collected for humanitarian purposes could be used for malicious purposes by third parties that have access to the data.

UNHCR has implemented protocols to protect the privacy of the Afghan refugees, including removing the links that connect the codes describing the iris to the refugees’ personal information. However, a new problem arises from this change: false matches can no longer be corrected. If a refugee claims that there is a false match, officials cannot check whether the matched iris actually belongs to another person or, alternatively, confirm that the refugee is lying to attempt to claim assistance twice. This link removal also does not equate to full privacy. It still may be possible to identify the refugees by matching the UNHCR database to another database where the irises continue to be linked to personal information.

UNHCR has been using iris scan technology since 2012 to allow Syrian refugees in Jordan to access their cash assistance. As of January 2016, 630,000 Syrian refugees in Jordan had registered for the program. The organization sees this program as a great success, as it has provided refugees with quicker access to financial assistance, and there have been no major problems with the biometric system thus far. UNHCR is planning to expand the use of the technology to provide aid to Syrian refugees in other countries. It is important, however, for UNHCR to step back and reconsider the risks associated with the biometric system. False matches can deny much needed assistance to refugees with legitimate claims. Comprehensive analysis of the technology should be carried out to ensure the privacy of refugees, as well as their access to assistance.

Article Source: Jacobsen, Katja Lindskov. “Experimentation in Humanitarian Locations: UNHCR and Biometric Registration of Afghan Refugees.” Security Dialogue 46(2), 2015.

Featured Photo: cc/(Vladimir Arndt, photo ID: 67944545, from iStock by Getty Images)