AI governance paths explained
AI Governance Check uses the same four internal governance paths as the Practical AI Governance Toolkit: Low, Moderate, High Impact, and Stop / Escalate. Governance readiness is shown separately so missing safeguards do not silently downgrade the inherent routing decision.
The rules engine evaluates information sensitivity, AI authority, consequences, affected people, higher-impact domains, deployment context, vendor involvement, autonomy, and unresolved uncertainty. The result is an internal governance aid, not a statutory or regulatory classification.
Low
Routine, low-consequence use that affirmatively meets the baseline conditions: appropriate information, limited authority and consequences, no unresolved material uncertainty, and use within approved organizational boundaries.
Moderate
Additional exposure or uncertainty requires structured review. Typical triggers include non-public information, public-facing use, third-party AI, recommendations or limited automation, transparency concerns, or meaningful operational risk.
High impact
The use may materially affect people, rights, safety, regulated activity, consequential decisions, sensitive information, biometrics, or high-consequence operations. Enhanced assessment, specialist review where applicable, documented go-live approval, and residual-risk acceptance are required.
Stop / Escalate
A potentially prohibited or clearly unacceptable purpose, severe-harm condition without established safeguards, material approval conflict, or other override condition exists. Do not deploy or expand through the normal path until an authorized review determines the next step.
Ready to classify your AI use?
Start assessment