The AI-augmented humanitarian: Responsible automation across the global development and humanitarian aid programming cycle
Abstract
Humanitarian agencies, donors, and technologists are increasingly embedding artificial intelligence (AI) and automation within humanitarian and development programming. Driven by donor fatigue amid never-ending aid flows, operational demands for agility and speed, and promises of improved efficacy, some see AI as the answer to the sector’s crisis, epitomised by the Trump administration’s 2025 US foreign aid pause,1 often referred to as part of the wider Global Aid Reset.2 While the sector maintains institutional caution rooted in its ‘do no harm’ principles, AI-enabled practices are quietly entering the operational infrastructure of global development, from large language model-assisted e-mail and report writing to international non-governmental organisations headquartersprototyping programming cycle automations, predictive analytics, anticipatory action,early warning systems, and generative mediation tools. This paper proposes a frameworkfor the AI-augmented humanitarian, emphasising AI as augmentation of human decisionmaking rather than replacement, grounded in participatory governance, adaptive learning,and ethical accountability. Drawing on philosophy of technology, sociotechnical systemstheory, and applied case evidence from the global development sector, the paper arguesthat AI’s greatest potential lies in enhancing humanitarian experts’ judgment, amplifying marginalised local voices, and enabling more timely and responsive action, while rigorously attending to risks of bias, automation dependency, and power asymmetry. The framework offers practical pathways for responsible AI integration that honour both ethical commitments and operational imperatives. This article is also included in The Business & Management Collection which can be accessed at https://hstalks.com/business/.
The full article is available to subscribers to the journal.
Author's Biography
Kristian Dalsgaard Svendsen is a researcher and practitioner specialising in geopolitics, monitoring, evaluation, research and learning (MERL), conflict transformation, and the responsible integration of artificial intelligence (AI) in humanitarian and development programming. He serves as Senior Expert with Conflict Management Consulting, undertaking MERL assignments for global iNGOs and advising on PeaceTech in fragile and conflict-affected contexts. As Council Member of the Danish Council for International Conflict Resolution, he contributes policy briefs, analysis, and commentary on non-military approaches to international security, conflict resolution, and the implications of emerging technologies for peace. He is currently prototyping AI-augmented evaluation tools for complex development systems, including predictive analytics for early warning and natural language processing-driven qualitative analysis as part of the AI-native startup MERLx. His work spans digital peacebuilding, participatory design, and adaptive management across the Horn of Africa, Middle East, South Asia, and globally, collaborating with United Nations agencies, international non-governmental organisations, and local civil society organisations. His current research focuses on AI-augmented evaluation methodologies, including explainable AI for causal inference and the ethical governance of automation in humanitarian programming. Kristian holds a Master’s degree from Roskilde University and is currently based in Barcelona, Spain.
