AI Governance Scorecard
The AI Governance Scorecard is a five-dimension framework for assessing whether an organisation's AI use is safe and defensible: acceptable use, oversight, data handling, competence, and auditability.
AI adoption without governance is a board-level risk. The AI Governance Scorecard turns "are we using AI safely?" into five concrete dimensions you can assess, evidence and improve. It is the artefact a risk or compliance team can stand behind.
| Dimension | The question it answers |
|---|---|
| Acceptable use | Is there a clear, communicated policy on what AI may and may not be used for, and do people understand it? |
| Oversight | Is there human review of AI-assisted decisions where it matters, with clear accountability for outcomes? |
| Data handling | Do people know what data can be put into which tools, and are sensitive and regulated data protected? |
| Competence | Can people actually use AI safely — do they understand hallucination, bias, and data exposure risks? |
| Auditability | Can the organisation evidence how AI is used, by whom, and to what standard, if asked by a regulator or board? |
This is Forge's articulation of the framework, shared under a Creative Commons Attribution 4.0 licence. You are free to reproduce and adapt it with attribution to Forge (Chimerical Ventures).
Frequently asked questions
What is the AI Governance Scorecard?
It is a five-dimension framework — acceptable use, oversight, data handling, competence, auditability — for assessing whether an organisation's AI use is safe, controlled and defensible.
What should an AI governance policy cover?
At minimum: acceptable use, the level of human oversight required, rules for data handling by sensitivity class, the competence people need to use AI safely, and how use is evidenced for audit.
Is AI governance just about compliance?
No. Governance is what lets an organisation scale AI with confidence rather than fear. Done well it removes blockers, because risk and compliance teams can see the use is controlled.