This piece by Dr Philip Proudfoot (Institute of Development Studies), Helene Juillard (Key Aid Consulting), and Prof Bertrand Taithe (HCRI), for the Humanitarian Archive Emergency project.
The humanitarian sector is losing its memory. Funding cuts, closures, and the slow decay of digital infrastructure are all destroying archives, datasets, and records. A new global research initiative wants to find out how much has already gone and what can still be saved.
In early 2025, the main USAID website, along with its vast archive of publicly funded datasets, programme evaluations, and operational reports, went offline. Agency staff had their emails deactivated; internal systems were locked. This was not the work of hostile hackers, nor the consequence of some external catastrophe. It was the result of a political decision and executed with the banal efficiency of a server being unplugged.
Groups of researchers, archivists and former employees, working through back channels and contractual workarounds managed to save portions of USAID’s programme files. But many vital resources, like the Development Experience Clearinghouse, which held around 200,000 documents, representing the accumulated paper trail of billions of dollars in aid spending, remain inaccessible.
Given global power asymmetries, the USAID case was dramatic because it was visible. But across the humanitarian sector, a quieter, more diffuse version of the same crisis is unfolding every day, and it is almost certainly hitting grassroots and local organisations harder.
As the system retracts, organisations are shuttering offices, shedding staff, and cutting budgets. When they do, their records go with them. Not just files and folders, but also intangible institutional memory that lives in how an organisation understood its own work: the datasets tracking disease outbreaks, the evaluation reports documenting what worked and what didn’t, the email chains recording how life-and-death decisions were made. All of this is at risk of disappearing, if not already gone.
And often this happens not because someone pressed delete, but through the more mundane mechanics of institutional neglect: a webpage going offline, a subscription lapsing, or even a hard drive boxed up and forgotten in a cabinet nobody has the key to anymore. Erasure takes many shapes.
Saving Lives, Preserving Files
The current funding environment makes the data crisis especially acute. OCHA has warned that data about crisis-affected populations faces the highest risk to continued availability. Anecdotally, we know that many humanitarian organisations do not maintain adequate formal archiving practices to begin with. Digital records exist in shared drives, personal laptops, and printed documents in filing cabinets that might be cleared away when an office lease expires. Smaller organisations, national NGOs, local research institutions, community-based groups, are especially exposed to disappearing without a trace.
The fragility of digital records compounds the problem. Unlike paper, which degrades slowly and visibly, digital materials can become inaccessible in an instant — through format obsolescence, corrupted storage media or digital rot, loss of a password or simply the expiry of a platform subscription.
Yet digital preservation is also a specialist discipline, and very few humanitarian organisations have the capacity or resources to practise it. And in a system where every budget line must be justified against the ever-lingering imperative of ‘saving lives,’ the argument for a data preservation specialist is a difficult one. The irony, of course, is that the lives being saved tomorrow often depend on the files being preserved today. These records are theirs often and not ours to lose.
Lessons Lost
The words “lessons learned” appears across humanitarian documents. It is one of those phrases that has been repeated so often it has become meaningless, almost like a bureaucratic incantation. But lessons, by definition, require a record. They require someone, somewhere, to have written down what happened, what went wrong, what worked despite the odds, and for that record to still exist when the next crisis hits. Without archives, there are no lessons. There is only the endless repetition.
Understanding the scale
Many of you reading this will have a sense that loss is occurring. Yet we do not have any accurate measurement of its scale. And before we can talk about what to save, or how to save it, we need a map of what is disappearing, and where particular threats cluster.
We are part of a new research initiative, the Humanitarian Archive Emergency project, funded by the Wellcome and Leverhulme trusts led by Bertrand Taithe, which is trying to address that gap. The project is convened by the Humanitarian and Conflict Response Institute at the University of Manchester and supported by Elrha, IDS, Key Aid. It brings together a coalition of interested parties and experts in a steering committee of archivists, researchers and practitioners. The global survey is our first step. We are inviting anyone working in or around the humanitarian sector to identify archives, records, and datasets that are threatened, damaged, or already gone.
The survey [LINK] — available in English, French, Spanish, and Arabic — is not intended as a simple audit. It is closer to a census: a first attempt to understand the global contours of the data crisis that has, until now, been discussed largely in anecdote and alarm.
How many organisations have lost access to their own records in the past two years? How many datasets are sitting on digital infrastructure that could disappear next month? How many archives exist only because one person in one office has kept them safe?
These are not abstract research questions. For a sector that claims to run on evidence, they ought to be existential. Every programming decision, every funding allocation, every accountability mechanism, every lessons-learned exercise depends on access to records of what has been done before. When those records vanish, the sector does not simply squander its history, it loses the capacity to learn from it.
Humanitarian Archives Emergency
The HAE project is designed to respond on two fronts. The first — the survey ,now open — aims to produce a landscape assessment: a map of what exists, what is endangered, and what drivers of risk are most pressing. The findings will feed into a set of vulnerability case studies and a series of regional convenings bringing together practitioners, archivists, and researchers across the Americas, West Africa, the Middle East, and Asia-Pacific.
The second phase will develop, with support from the Archives and Digital Media Lab, a proposal for an ethical framework for prioritising preservation, a triage system, in effect, for deciding what to save when you cannot save everything. As we know, the history of humanitarian archiving has overwhelmingly favoured institutions headquartered in the Global North, while the records of organisations based in the global majority often are the ones closest to affected communities. Whose archives count? Whose records get rescued? Whose voice should be heard in future?
—
The survey is open now. If you work in the humanitarian sector — or you are adjacent to it — and know of archives, datasets, or records that are at risk, we want to hear from you