Analyst Briefing
Where Agentic Governance™ sits among the categories.
A map for analysts who already track adjacent categories. Agentic Governance™ is not a substitute for AI-governance, model-risk, or prompt firewalls — it sits at a different altitude and consumes their signals as inputs to an authority decision.
The landscape
- AI-governance (model fairness, safety, policy) — answers “is the model well-behaved?”. Feeds the authority decision.
- Model-risk (SR 11-7 lineage) — validates models statistically. Orthogonal to runtime authority.
- Prompt firewalls (injection / jailbreak screening) — govern the prompt as input. A verdict feeds, but does not replace, the prompt-as-authority-act decision.
- Agentic Governance™ — governs the authority behind the act and resolves it to finality.
Why the altitude matters
A perfectly fair model can still take an unauthorised action; a validated model can still exceed its scope; a clean prompt can still spend authority it was never delegated. Each adjacent category closes a different gap. Agentic Governance™ closes the authority gap — and makes the closure provable.
In closing
The dividing line across the whole map: every act has authority; every authority has finality.