The use of APIs in the humanitarian sector is on the rise. Across UN80 and Humanitarian Reset discussions, leveraging the use of technology to facilitating quick, interoperable and structured data sharing is front and centre.

As an early adopter, DTM launched its first API in June 2024, with subsequent upgrades to improve the interface and user experience. Since its release, DTM’s API has been accessed over 75,000 times, providing open, well-structured, and timely displacement data
It is clear that APIs are likely to continue to shape the humanitarian data ecosystem as a core data generation and sharing modality. Leveraging APIs to make data more readily available and accessible to enable local capacities for re/use is vital. The fact that 83%  of data from the Humanitarian Data Exchange (HDX) is currently sourced from APIs is a testament of this shift towards open access.

To mainstream this practice across the humanitarian data ecosystem, tremendous work goes into working with operational field teams. This requires development of well-structured data schemas that work cross-nationally, enhancing data pipelines and data governance, amongst many other efforts by professional data providers/stakeholders.

While there is much to be done, DTM is continuously improving to expand the data availability from its API, but also increasing data accessibility for a variety of different users. 

DTM API 3.0 Upgrade: Unlock Drivers, Origins and Sex
The latest version of DTM API 3.0, released on 22 August 2025, now includes key data points on drivers of displacement, sex of those displaced and the broad origins  of where people came from. Data on drivers, origins and sex will complement existing IDP numbers at admin level 0,1 & 2. Naturally, all administrative boundaries are P-Coded which have been pulled from OCHA’s COD database to ensure interoperability. In cases where data was not collected, or data quality is an issue, the information will be marked as such. 

Compiling and commonly structuring 15 years of data from DTM teams currently operating in 91 countries (previously 101 at its peak in 2024) has not been an easy task and came with multiple obstacles.

Elevating the User Experience

The success of our API is based on our users. An intuitive and seamless user interface is foundational to widespread adoption and data use. Hence, in DTM API 3.0, a quick and simple registration page was added, along with the use of user authentication tokens for API calls.

For DTM, this authentication integration is valuable for user-centric design and future UX enhancements for the following reasons:

• Better understanding who DTM’s API users and community are.
• Gaining insights into what data fields are most useful (and least useful) for users.
• Preventing abuse of the API through potential malicious or excessive usage.
• Adherence to Data Responsibility best practices, which is a joint effort, and would not be possible without the pooled expertise, resources and practices of Humanitarian Data Standards Working Group and Inter-Agency Data Responsibility Working Group.
• And most importantly, to strengthen user engagement by proactively notifying them about upcoming API changes and gathering their feedback on future developments and priorities.

Where next: Improving accessibility

As we continue to make more of DTM's data available we are exploring ways in which we can make the data accessible and useful for a wider range of stakeholders, often with different levels of skills regarding analytical or visualization tools, but also with different information and analysis needs.

We are testing approaches likeModel Context Protocol which will allow users to query the API data in plain language using Large Language Models (LLMs) such as Copilot or ChatGPT. Users will be able to  query the API data with plain language requests such as “how many people were displaced in Haiti in 2024?”, “which Homrada (admin 3) in Ukraine reports the highest level of IDP women?”, or “which district in Iraq hosts the highest proportion of climate displaced IDPs?”

We hope this approach can help reduce barriers to the use of DTM data and encourage the use of DTM displacement data in decision-making and analysis among a broader audience.

To get started on using the API,register here, where all documentation, and code libraries for Python and R can also available.

This work was supported byTheComplex Risk Analytics Fund (CRAF’d).

 

¹ The Rise of APIs in the Humanitarian Sector – The Centre for Humanitarian Data

²Origins are largely per the majority group in each location/area as recorded by DTM and should only be used as a generalisation.