This workshop starts with the people already doing this work. In a survey of 211 NGO and international organisation staff for PeaceTech in Practice, 70% said tech experts do not understand the real needs of people working in conflict areas. This is a design problem, not just a policy issue. The people who build humanitarian and peace technology are rarely the ones who have to use it during a crisis, and the tools that result often do not match conditions in the field: low connectivity, old hardware, no maintenance budget. The same pattern shows up in who controls the AI infrastructure these tools rely on. More on that in the session. Open source has real potential here. It removes licensing costs that shut organisations out and can be built for low tech, low connectivity environments. But that only works if people using these tools are listened to when they are designed. Our data shows they currently are not.

"Do No Harm" is a well established principle in peacebuilding, but it is rarely applied to open source design decisions. This workshop draws on the work of the PeaceTech Alliance, a nonprofit Austrian network of peace organisations, technologists and universities. In groups, using sticky notes, participants map where Do No Harm applies to open source tools, how it could be measured, and where the current gaps are. The findings are brought together and shared after the conference.