Most AI adoption isn't a strategic decision. It's staff quietly using tools to cope with impossible workloads. If leadership doesn't know what's in use, they can't manage the risks.
AI shapes who gets prioritised, how needs are assessed, and how resources are allocated, including through backend systems like allocation algorithms and forecasting models you didn't build. Do you know where AI is shaping outcomes?
AI tools are built on data that reflects certain contexts: often Global North, English-language, well-resourced settings. They may perform differently across languages, cultures, and geographies.
Every AI system produces errors. In humanitarian work, that could mean someone is excluded from assistance, misidentified, or exposed. Tools for detecting bias exist, but none are being systematically applied in the sector yet.
When an AI tool contributes to a harmful outcome, who is responsible? If the answer is unclear, accountability doesn't exist.
Consent isn't enough. The people affected by AI-driven decisions should have genuine input into whether and how these tools are used, not be informed after the fact. If your organisation can't point to where that happens, it's a governance gap.
One team using one tool is manageable. But AI adoption scales fast. The time to build governance is now, before AI use becomes too widespread to manage.
Your organisation can confidently answer
questions.
Want to talk through your results? I offer a free introductory conversation to help you figure out where to start. No pitch, no commitment.
Message me on LinkedIn