IATI Director Yemesrach Workie on Making Open Data Sharing a Priority in the AI Era

In this Humanitarian AI Today episode, Yemesrach Assefa Workie (Yemi), Executive Director of the International Aid Transparency Initiative (IATI), speaks with producer and guest host Brent Phillips about open data as a strategic infrastructure for artificial intelligence. As an open data sharing framework widely used across the humanitarian community, IATI tracks over $3.9 trillion in funding across more than one million development and relief activities reported by nearly 2,000 organizations. The conversation explores how rapid advances in machine learning are reigniting interest in large structured data sets, allowing AI applications to seamlessly traverse IATI’s vast corpus and convert decades of decentralized global aid data into real-time operational insights. The interview dives into practical use cases where AI tools are already unlocking the value of IATI's machine-readable XML standard alongside the 500,000 unstructured evaluation documents, audits, and reports linked within its database. Yemi highlights real-world implementations, such as the UK Government’s Development Explorer for portfolio analysis, Germany’s AI-driven data quality checks, and IATI’s deployment of AI help desk agents. However, she cautions that AI capabilities are only as good as their underlying inputs, urging humanitarians to manage critical risks including poor data quality, loss of local crisis context, representation bias from varying publishing volumes, and over-reliance on automation over human judgment. A key takeaway from the discussion focuses on how publishing open data acts as a vital visibility multiplier for aid organizations in the AI age. As autonomous agents, LLMs, and decision-support systems become the primary interfaces for analyzing crisis responses, unstandardized or unshared operational data effectively renders an organization invisible to automated discovery. By channeling standardized data through open frameworks like IATI, actors of all sizes, particularly small civil society organizations and local NGOs, ensure their ground-level contributions, delivery partnerships, and operational footprints are accurately represented and indexed by global AI systems. Looking ahead to major sector gatherings like the AI for Development Summit hosted by DevelopMetrics, the episode underscores why structured open data standards are experiencing a major revival. By providing a standardized, verified baseline, large datasets like IATI prevent model hallucinations, enable cross-domain interoperability with geospatial or climate data, and lay the groundwork for futuristic decision-support tools. Concluding with a clear call to action, Yemi encourages organizations of all sizes to publish their data, refine entry quality, and leverage free IATI tools to position themselves at the forefront of AI-driven humanitarian action. Humanitarian AI Today is a community-produced podcast and a proud member of the Devex Podcast Network, an initiative dedicated to expanding the visibility of coverage across global health, development, and relief sectors. The series brings together community partners like the International Aid Transparency Initiative (IATI) and enterprise technology leaders like Neo4j, providing an open channel for humanitarian actors and technical teams to share updates on AI projects, discuss advances in machine learning, and collaborate on data-driven solutions for the sector.