As part of our research on the Global Data Justice project, we are investigating how actors in the humanitarian sector are embracing new technologies and how digital innovation is changing humanitarian values. It is an exciting time for the sector as it incorporates new technologies into its operations, including drones, advanced data analytics, and biometrics, among other innovations. But the use of technology is neither straightforward nor apolitical, and it is important that we understand how the datafication of aid might disrupt the humanitarian cause in unexpected ways.

This morning the BBC reported that the World Food Programme (WFP) is partially suspending food aid in areas of Yemen under the control of the Sana’a-based authorities (the Houthis) because of their refusal to accept WFP’s introduction of a biometric registration system for aid distribution, as first reported by The New Humanitarian. For WFP, biometrics are an important means to control food aid and to prevent fraud in aid distribution. WFP argues that in this case “the integrity of our operation is under threat and our accountability to those we help has been undermined” by the Houthis’ refusal to allow the use of biometrics to facilitate the delivery of food aid.

As experts have already begun to highlight, WFP’s decision to suspend aid in this case sets a critical precedent. It appears to be the first time that a humanitarian organization has withdrawn assistance in response to local resistance to the use of a technology such as biometrics. It is worth pointing out that biometrics are not mandatory for registration with other humanitarian agencies, such as UNHCR. WFP’s decision to suspend aid also ought to be understood in a wider context: WFP has previously stated that it has a “moral imperative to leverage technology to achieve efficiencies” and that the organization is on a “very aggressive digital transformation journey”. This appears to be the next step in that journey.

The Houthis have stated that they “refuse the enrolling of beneficiaries in a biometrics programme because it is counter to national security” and that the collection of biometric data by WFP is part of an “intelligence operation”. It is not easy for outsiders to the conflict to evaluate whether the Houthis’ concerns are founded, or what process occurred within WFP and the UN to reach this decision. The WFP’s statement notes that the decision has “the support of the entire United Nations system”, indicating that it was discussed across multiple agencies and presumably reflects a policy decision on the part of the UN’s leadership, as well as an operational one on the part of WFP.

The larger problem this conflict over biometric registration surfaces, however, is that the Houthis are not raising concerns over data protection or privacy. They are pushing back on issues of politics and sovereignty, and cannot be answered by promises to handle data responsibly. What is being questioned here is who holds power over whom, and how they should be able to wield that power. It is not the first time this question has been asked in relation to the humanitarian system: the status of neutrality conveyed by the Geneva Convention and the UN Convention on various international actors has never been stable or uncontested. Neutrality is a political status that is conferred partly due to moral and functional necessity, but as importantly represents a political settlement between powerful actors.

Those with the power to identify and verify also hold the power to determine access to resources. This is true of humanitarians, as well as governments and military commanders, even if they draw their power from different sources. Biometrics and ID systems are currently one main flashpoint for the often-fragile balance of power that allows humanitarians to fulfil their role in the world’s most volatile places: they now link the WFP, for example, to Palantir through contracts and highly visible news reports, and through the firm to its investors and clients (the CIA, the Trump administration, and links to US national security infrastructure). They also link to delivery systems which parties to conflicts may want to influence, and provide ways of redefining people between categories (refugee, member of an ethnic group, combatant, aid recipient) in ways which also feed into constituencies and power. There are many reasons to think about ID databases as sources of power, and many reasons to demand they are kept neutral — or not used. Not having biometric ID systems in use does not reduce the power of humanitarian actors, but erecting them does create new forms of power. For instance, sovereignty is a question of legitimacy, but it is also performative: the presence of technologies that link UN agencies to other actors allows the Houthis to perform sovereignty (and possibly gain it in doing so) by resisting them.

This is not to argue that the UN should abdicate its mission and bend to the will of parties to the conflicts where it operates. It is, however, to say that databases are inevitably tools of power, and people — particularly in conflict situations — can be expected to have an opinion on this. Digital databases are a ‘many hands’ problem: they have multiple uses (distribution of goods but also anti-fraud; inclusion and recognition but also categorisation and exclusion). They are rarely run and used by just one actor, and digital data does not eventually become invalid: its uses only multiply over time. The security of the database is the security of the people in it, and power over the database is power over the people in it. It is not just the Houthis who will question this power as digital and biometric ID systems become increasingly sophisticated, and humanitarian agencies need to succeed in their efforts to demonstrate they are aware of the new power balance biometric ID creates.