Board governance · 7 minute read
The board questions that expose AI operating gaps
Questions that move governance from policy confidence to inspectable operating evidence.
- Board & Executive
- Trust
- Global
- Decide
Decision lens
Can leadership see who owns material AI use, what authority it has and what evidence supports operation?
Boards do not need model-level detail, but they need confidence that material uses, accountable owners, authority, evidence gaps and incident readiness are visible.
Ask about material use, not tool adoption
Focus reporting on important workflows, decisions and affected stakeholders. A model count or licence total says little about operating exposure.
- Which uses can create material impact?
- Who owns each outcome?
- What remains outside the inventory?
Ask what the system may do
Require explicit authority, prohibitions, approvals and stop conditions for agents that access data, use tools or influence decisions.
- Which actions require people?
- Can permissions prove the boundary?
- Who can stop operation?
Ask how evidence stays current
Evaluation, monitoring, incidents, supplier changes and realised value need recurring governance after release.
- What triggers reassessment?
- Which gaps remain open?
- How are benefits and risk reviewed together?
Practical checklist
Evidence to bring into the decision.
- 01Material-use inventory
- 02Named accountability
- 03Authority view
- 04Evidence and remediation status
- 05Incident and value reporting
Continue with evidence
Connect the perspective to an operating method.
These questions support governance discussion. They are not legal advice, director guidance, certification criteria or an audit programme.