AI governance
What controls, roles and approvals govern the system?
Category distinction
Decision assurance vs AI governance
AI governance defines how an organisation intends to control AI systems. Decision assurance asks what formed a specific consequential organisational act when the organisation acted.
Maintained by VeriscopicLast reviewed 2026-09-24
Definition
The precise question
Decision assurance is decision-linked; AI governance is system- and organisation-level. The two can work together without answering the same question.
Reference model
From category to independent review
01
Decision assurance
02
Decision evidence
03
VES decision envelope
04
Evidence objects
05
Independent review
The envelope can link model output, runtime records, policy versions, authority records, claims documents, human rationale, external data and approvals while preserving each source's declared limits.
Insurance example
What this looks like in practice
A model approval record may show that a system was governed. A decision envelope can additionally show which version contributed to an underwriting exception and how a person exercised authority.
Evidence map
What a reviewer needs to see
- Model purpose, approval and operating conditions
- Specific input, output and system contribution at execution
- Human judgement, authority, commitment and declared limitations
Different questions
Decision assurance
What evidence, authority, conditions, contribution, judgement and commitment formed this act?
Where they connect
A model card, approval record, monitoring result or gateway receipt may contribute to a decision package. The package still needs the decision context and organisational commitment to be reviewable.
A practical insurance test
Could a reviewer explain why this claim or underwriting decision was made, rather than only proving that the model was approved? If not, governance evidence is necessary but not sufficient.
The VES connection
A published evidence structure
VES 1.2 can represent model governance records as Evidence Objects while preserving the wider decision-state around them.
Read VES 1.2 for the published Decision Envelope, Evidence Object and verification model.
Boundary
Decision assurance does not replace model governance, risk management or runtime controls. It prevents those records from being mistaken for the complete basis of a specific decision.
Frequently asked
What does decision assurance vs ai governance address?
Decision assurance is decision-linked; AI governance is system- and organisation-level. The two can work together without answering the same question.
Does this establish that a decision was correct or compliant?
Decision assurance does not replace model governance, risk management or runtime controls. It prevents those records from being mistaken for the complete basis of a specific decision.