In the sprawling settlements of Cox’s Bazar, the lives of more than a million Rohingya refugees are increasingly defined not by their histories, aspirations, or rights, but by data points and biometric codes. What began as a massive humanitarian registration effort to document a population uprooted by genocide in Myanmar has evolved into a system that can feel less like protection and more like surveillance. As Rohingya families queue for food, seek medical care, or enroll their children for basic services, they encounter a reality that is deeply quantified, highly conditional, and remarkably disempowering. In this new humanitarian paradigm, being counted has become more significant than being cared for.
Humanitarian actors often speak of registration, verification, and biometric identification as essential tools for ensuring that assistance reaches intended recipients and that operations run efficiently. In Bangladesh, the Government and UNHCR have carried out joint biometric registration exercises for Rohingya refugees, enrolling fingerprints, iris scans, and other identifying information into elaborate databases. These systems were initially presented as ways to ensure accurate identity documentation and improve service delivery. By mid-2018, joint verification exercises had begun in the camps of Kutupalong and Nayapara, with biometric data captured to support identity management and the issuance of identity cards to refugees who had fled persecution in Myanmar.
According to UNHCR, biometric identity management systems capture unique identifiers — such as fingerprints and iris scans — along with vital biodata and family links, which are then used to generate identity documents. For many Rohingya refugees, these were the first forms of official identification they had ever received. In 2019, for example, more than half a million Rohingya received such identity cards jointly issued by the Government of Bangladesh and UNHCR, marking a significant milestone in registering a population that had arrived en masse after waves of violence in Myanmar.
Biometrics were promoted — publicly and within operational planning — as a fraud-proof means of confirming identity, ensuring that assistance could be properly targeted, and aiding in family reunification and service provision. On paper, they represented progress in documenting one of the world’s largest stateless populations, a group that had fled systematic discrimination and denial of citizenship in Myanmar.
Yet beneath this narrative of technical efficiency and targeted aid lies a more complex, and often troubling, set of consequences. Gathering and managing data on marginalised, vulnerable groups is not a neutral exercise; it shapes power relations, boundaries of belonging, and mechanisms of control. As The New Humanitarian’s commentary from 2017 observed, Rohingya refugees had learned that data collected about them “can be used to further diminish their human rights,” even when it was framed as part of humanitarian planning.
These early warnings have since become all too real.
In 2021, Human Rights Watch documented how at least 830,000 names and biometric details of Rohingya refugees collected in Bangladesh were submitted by the Bangladeshi government to Myanmar as part of repatriation assessments. Although the UNHCR sought to clarify its role, the report noted that the data included information derived from UNHCR’s registration exercises, such as non-digital thumbprint images originally gathered for humanitarian purposes. The sharing of such information — without transparent consent mechanisms — raised immediate concerns about refugees’ safety and rights, given Myanmar’s history of persecution and forced displacement.
Human rights advocates argued that biometric databases hold particularly sensitive information that cannot simply be treated as humanitarian infrastructure. When a person’s fingerprints, iris scans, family ties, and birth details are stored in a central database, those identifiers become both a lifeline and a potential liability. The implications of mishandling such data, or of sharing it with authorities who have previously perpetrated ethnic violence, are profound.
For many Rohingya refugees, distrust of biometric data was not abstract. In 2025, reports surfaced that around 400 Rohingya families in the Kutupalong and Nayapara camps had been cut off from essential food rations, cooking fuel, soap, and access to medical services because they declined to participate in a renewed biometric update exercise. These families — long registered and reliant on humanitarian assistance — found themselves effectively excluded from the very systems designed to support them because they refused to provide updated biometric data.
According to operational notices, the registration update exercise required attendance by all eligible refugees for biometric enrollment and renewal of family information. Those who failed to attend risked being “inactivated” in the database and subsequently losing access to assistance. For many Rohingya, refusing biometric enrolment was not a whimsical choice. It was a protest against how their identities were being recorded, managed, and used in ways that extended beyond mere humanitarian delivery into realms of control and conditionality.
Some refugees voiced deep concern not just about the mechanics of biometric collection but about what the resulting identity documents signified. Community accounts suggested that updated biometric ID cards did not explicitly acknowledge “Rohingya” identity — a significant issue for a stateless group that has endured systematic denial of recognition and citizenship. For people whose very identity had been erased within Myanmar, the absence of explicit acknowledgement on identification documents in exile was experienced as a form of renewed erasure.
Critics have questioned the ethical foundation of digital identity systems in humanitarian contexts. The Bangladesh-UNHCR biometric programme did not always ensure that refugees were fully informed about how their data would be stored, shared, or applied. Fieldwork and commentary have highlighted that refugees were sometimes told to register primarily to receive aid, without a clear explanation of potential risks or uses of their biometric data beyond documentation. In similar contexts, internal audits have found that UN staff implementing biometric programmes lacked a complete understanding of data protection policies, which in turn undermined meaningful protection for refugees’ privacy and agency.
The role of biometric data in managing refugee populations also intersects with national policy goals. Governments use data to identify, track, and manage displaced people as part of broader administrative agendas. In Bangladesh’s case, biometric registration was tied to broader discussions about repatriation, camp management, and the organisation of refugee services. In some cases, governments have explicitly stated that data might be used to support returns, a prospect that provokes anxiety among refugees who fear forced repatriation to dangerous conditions in Myanmar.
These dynamics illustrate a paradox: a system designed to protect can instead make refugees more vulnerable. When biometric data is used to determine eligibility for essential services such as food and healthcare, those who opt out or question the process can be left without the basic means of survival. When identity information is shared with third parties, including governments with records of violence against the group in question, that same information can become a tool of repression or coercion. And when identification documents fail to recognise the very identity that refugees claim, the identity that was stripped from them in their country of origin — the result can feel like a second erasure.
Viewed through this lens, the humanitarian embrace of data-driven identity management raises fundamental questions about power, consent, and the meaning of protection. Identity, once a deeply personal and meaningful aspect of human life, becomes a dataset to be processed, updated, and indexed. Those who administer assistance are also the custodians of this data, and the relationships of authority and dependency this creates are rarely symmetrical. Refugees find themselves needing to comply with bureaucratic requirements not simply to access aid, but to maintain their very presence in the database that determines their eligibility for survival.
The consequences of these systems extend beyond service delivery. Refugees who lack secure, functioning identity documentation often find themselves cut off from basic rights and mobility. The fear of exclusion from databases can drive compliance even when refugees harbor deep distrust of how their data will be used. This creates a situation where the collection of data becomes a condition of life itself, rather than a tool for empowerment.
In this environment, the line between protection and control blurs. The language of humanitarianism — identity cards, registration exercises, verification processes — masks a deeper reality: the data collected shapes who is seen, who is counted, and who is cared for; it also shapes who is overlooked, excluded, or rendered invisible.
For Rohingya refugees, the challenge of identity has always been central. Denied citizenship and documentation in Myanmar, they arrived in Bangladesh already stripped of formal political belonging. What was meant to be a mechanism to restore some measure of administrative order has in many cases become a system that replaces personhood with a profile line in a database. Rather than empowering refugees, the current biometric regime risks reducing them to entries in a ledger — users of services rather than subjects of rights.
As humanitarian actors and governments continue to refine biometric and data systems, these developments highlight an urgent need to interrogate not just the technical efficacy of such systems, but their ethical and political implications. Protecting refugees requires more than counting them. It requires respecting their agency, safeguarding their privacy, and ensuring that data meant for protection does not become a vehicle for exclusion.
If the humanitarian project is to be true to its principles, it must reckon with how technologies of identification are transforming the meaning of refuge. In a world where the Rohingya are already stateless, identity should not become another frontier of erasure.
References
- Joint Bangladesh/UNHCR biometric verification exercises for Rohingya refugees. (UNHCR)
- UNHCR Biometric Identity Management System (BIMS) capturing fingerprints, iris scans and biodata. (UNHCR)
- Over half a million Rohingya received biometric ID cards as part of joint efforts. (UNHCR)
- The Rohingya are the world’s largest stateless population in Cox’s Bazar. (UNHCR)
- Risks of data use raising concerns about human rights. (The New Humanitarian)
- Bangladesh submitted biometric names to Myanmar for repatriation, raising safety fears. (Human Rights Watch)
- UNHCR accused of improper biometric data sharing without consent. (ODI: Think change)
- Families losing aid after refusing biometric updates. (rohingyarefugee.news)
- Registration update rules conditioning assistance on attendance. (help.unhcr.org)
- Identity concerns on updated biometric cards lacking “Rohingya” label. (rohingyarefugee.news)
- Internal audits finding weak implementation of data protection norms. (WIRED)
- Potential use of data to support repatriation raises refugee anxiety. (fmreview.org)
