← All updates
MethodologyAugust 8, 2026

How we keep crisis data honest

Accuracy is the foundation of everything we do. A map of global crises is only as trustworthy as the data behind it, so we designed our process around one principle: if a claim can't be traced to a source, we don't publish it.

Where data comes from

Crisis discovery is assisted by an AI research pipeline that scans recent coverage from major news outlets, UN agencies, and established NGOs. We instruct the model to require recent, credible reporting — and to prefer under-covered regions and issues over headline-chasing.

Every published crisis stores its own citations: the specific articles and reports supporting its title, description, and figures.

How we score severity

Each crisis receives a severity score on a 1–10 scale based on two things: the scale of the affected population and the urgency of the situation. Severity drives ordering and visibility so the most acute situations surface first — it's a signal of urgency, not a measure of moral weight.

What we verify — and what we don't

We moderate every candidate against editorial criteria before publication: is it verifiable, geographically specific, and significant in scale? An AI critic flags weak sourcing and inconsistency.

We're also honest about our limits. This is a young project with a small team. Crisis descriptions are AI-assisted and moderated automatically; our human verification program is still scaling. We flag uncertainty rather than hide it.

How you can help

If you spot an error, tell us at contact@globalimpact.earth. We correct mistakes quickly and publicly.