AI will not set its own limits in warfare. States must. · Cordula Droege
Source published: 2026-10-15
Read the original at ICRC · Humanitarian Law & PolicyTrust is a design decision
For a purpose-led organisation, the useful question is who benefits, who bears the risk and who stays accountable. AI should make that answer clearer.
This reading from ICRC · Humanitarian Law & Policy is a starting point for discussion. Our perspective below explores the practical questions for charities and humanitarian teams; it does not reproduce the source article or claim to summarise its conclusions.
What this could mean for your team
Bring programme, safeguarding and privacy colleagues into the brief. Record who can pause the workflow and how an affected person can ask for a correction.
Include people who understand the service and the people it affects. A useful AI initiative should make work more reliable while keeping responsibility visible.
A small action with a clear purpose
Use a short team discussion to agree the problem, the information involved and the evidence you would need before trying a change. Write down the assumptions you still need to check.
Read the original source in full, then decide which lessons fit your context. The right next step may be improving a process, clarifying a policy or testing one bounded use case.
Take one useful step.
- Name the problem and who experiences it.
- Agree the evidence and human review needed.
- Choose one small, measurable next step.
Original AIHI commentary. Source authorship remains with the credited publication.
