Five AI risks every trustee should ask about
Five plain-English questions that help a board understand how staff use AI, and where it could go wrong.
Ifedayo · · 2 min read
Most small charities did not decide to adopt AI. It arrived quietly: someone used a chatbot to draft a funding bid, someone else summarised meeting notes, a volunteer tidied a spreadsheet. None of that is wrong. But trustees are responsible for oversight, and you cannot oversee what nobody has written down.
Here are five questions that open the conversation without needing any technical knowledge.
1. Which AI tools are people using, and for what?
Start with an inventory. Ask each team, in one short message, which tools they use and what for. You will probably be surprised, and the list is the foundation for everything else.
2. What information goes into them?
The biggest practical risk is data. Donor records, beneficiary details, staff information and financial data should not go into a free AI tool without a clear decision that it is safe to do so. Ask whether there is a written rule, and whether people know it.
3. Who checks what comes out?
AI tools produce confident text and tidy-looking numbers that can be wrong. If anything produced with AI reaches funders, the board or the public, someone should check it against the source first.
4. What happens if something goes wrong?
If someone pastes the wrong spreadsheet into a chatbot, would they know who to tell? Would they feel able to? A one-page incident plan and a no-blame culture matter more than any software.
5. Who owns this?
Someone needs to own AI use: keeping the list of tools, updating the rules and reporting to the board. It does not need to be a technical person. It does need to be a named person with time to do it.
Where to start
If you can answer all five confidently, you are ahead of most organisations. If not, our free AI & Controls Self-Check takes five minutes and shows you where to start.