AI for Peace: Empowering Humanitarian and Peacebuilding Organizations through Responsible Technology

Artificial intelligence is moving out of pilots and into the daily work of humanitarian and peacebuilding organisations. Most of the attention goes to field-facing applications — early warning, mediation support, information integrity. Much less goes to the work that actually consumes small organisations: donor reporting, financial control, project tracking and audit preparation. For an organisation of a handful of people, that administrative load often determines whether the programme work happens at all. 

This session addresses both, and gives most of its time to the second. ICT4Peace will moderate a discussion on where responsible technology genuinely strengthens the capacity of under-resourced organisations, and where it does not. Globethics will address ethics by design and the institutional conditions for responsible deployment; Collective Intelligence for Good will speak to tool development; PoliSync to how organisations put such tools into practice in their daily activities. 

The panel includes a live demonstration of a practical tool of the kind a small organisation can build and operate for itself. The illustrative case is one most participants will recognise: a project funded by two donors in two currencies, with overlapping but differently structured reporting requirements, where every cost has to be recorded once, allocated correctly between funders, evidenced once and reported twice. The demonstration runs on a synthetic dataset; no live organisational data is shown. It is deliberately specific about where AI does the work — reading information out of documents, categorising it, drafting narrative sections — and where ordinary automation and good process do it instead. 

The session closes with an interactive discussion on practical pathways for responsible adoption in operational contexts, and carries its conclusions forward to the Globethics workshop "Before the Tool: What Must Change Inside Organizations for Responsible AI in Crisis Contexts", with which it is designed to run as a sequence.