For many organisations, the first AI decisions have already been made informally: staff experimenting with chat tools, and suppliers switching on new features in software you already use. That makes it a governance question now, not a future one.
These five questions give a board a practical starting point.
- Where are we already using AI? Start with an honest inventory. Include the tools staff use day to day and the AI features built into existing systems, not just formal projects.
- What problem are we trying to solve? The strongest uses start with a real customer, member or operational problem. Be wary of adopting technology first and looking for a purpose second.
- What could go wrong, and who would be affected? Consider privacy, accuracy, fairness and reputation. Pay particular attention to personal information about customers, members, clients or volunteers, and check your obligations under privacy law.
- Who is accountable? Name an owner, agree a simple policy on acceptable use, and decide how and how often the board will hear about it. Decisions that matter should still rest with people.
- Do we have the capability to use it well? Tools are the easy part. Good outcomes depend on data quality, staff skills and the confidence to change how work gets done.
A practical first step
A focused board or leadership session can answer most of these questions in a few hours: map current use, agree the organisation's position and principles, and identify one or two opportunities worth testing. From there, AI becomes a managed part of the strategy rather than something happening around it.