AIGR · Institutional AI governance

Assessment

EvidenceReviewDecision

Governance assessment

Turn one AI system into a governed evidence record.

AIGR converts governance requirements into a controlled review: scope is fixed, evidence is assigned to owners, material blockers are separated from ordinary gaps, and the outcome can be explained to executives, procurement and risk teams.

ScopeOne defined system
EvidenceOwned + dated
Review states4 controlled states
OutcomeDecision-ready

Define exactly what is being assessed.

A governance opinion is only useful when the boundary is explicit. Before evidence review begins, the assessment record fixes the entity, named AI system, intended use, lifecycle stage, jurisdiction and evidence period.

ENTITY

The accountable legal entity and business unit.

SYSTEM

The named model, application or AI-enabled workflow.

LIFECYCLE

Development, validation, pre-production or production.

JURISDICTION

The regulatory context applied to the review.

PERIOD

The evidence window and assessment date.

Policy statements do not pass a control. Evidence does.

01

Relevant

The artifact answers the requirement being tested rather than an adjacent policy objective.

02

Current

The artifact reflects the system and period under review and has not been superseded.

03

Traceable

A named owner, source and date can be tied back to the organization’s operating environment.

04

Sufficient

The artifact demonstrates that the control operates—not merely that it was intended.

Six governed stages from intake to decision.

Each stage adds information without erasing the audit trail created by the stage before it.

01IntakeSystem, owner, use case and assessment boundary.
02ClassifyRisk context, sector overlay and evidence expectations.
03CollectArtifacts, owners, dates and source references.
04ReviewValidated, in review, not applicable or blocker.
05DecideScore, critical gates, conditions and governance outcome.
06MonitorExpiry, incidents and material-change triggers.
Assessment output

A board-readable view backed by reviewer evidence.

Governance recordControls + artifacts

Every material conclusion is traceable to the evidence reviewed.

Critical gatesNo masking

Severe failures are not averaged away by stronger domains.

RemediationOwner + next action

Open findings are sequenced into an accountable action plan.

Executive viewDecision signal

A concise outcome can be read without losing the underlying audit trail.

Enterprise assessment

Establish governance readiness for a defined AI system.

Request an Assessment