Before the Tool: What Must Change Inside Organizations for Responsible AI in Crisis Contexts

Artificial intelligence is increasingly present in peace processes, humanitarian response, and conflict-affected environments, but many of the real challenges are not technical. They are institutional. Current conversations often focus on how AI can improve humanitarian response. 

This session flips that question:   How can organizations better prepare internally to deploy AI responsibly in the first place? In fast-paced, ever-evolving crises and fragile settings, weak internal AI governance, unclear accountability, and inconsistent data practices can turn even well-intentioned tools, or even a lack of tools, into sources of risk. Without the right internal systems, AI can undermine trust, compromise data safety, and create unintended harm. 

This session brings together practitioners, policymakers, and change management experts to examine the organizational conditions that enable, or prevent, responsible AI deployment. Rather than focusing on tools in the field, it centers on institutional readiness: how decisions are made, risks are managed, and responsibility is assigned.