AIPS DPP — Trust and Ontology Taxonomies
AIPS DPP reuses the canonical AIPS ontology vocabulary rather than inventing a passport-specific Product taxonomy.
APT
PrimaryDerived
Derived products additionally declare SingleUpstream or Composite derivation topology.
Intelligence Capability (IC)
Perception, Classification, Detection, Prediction, Generation, RecommendationRanking, Reasoning, Planning, Optimization, Control.
Agency / Autonomy (AR)
Inform, Recommend, Assist, ActWithApproval, AutonomousAction.
Embodiment (EM)
NonEmbodied, PhysicalWorldPerceptive, EmbodiedActuating, ClosedLoopAutonomous.
Intelligence Effect (IE)
Score, ClassificationEffect, PredictionEffect, DetectionAlert, RecommendationEffect, GeneratedContent, ExplanationEvidence, Decision, Plan, DigitalAction, PhysicalAction.
Topology Scope
Standalone, DomainCoupled, CrossDomainFederated, EnterpriseOrchestrated.
Qualification currentness
CURRENT, STALE, REVALIDATION_REQUIRED, INVALID.
Trust currentness
Dependency/governance/assurance currentness is recorded separately so a Product can be a valid AI Product while a particular evidence dimension requires refresh.
Risk and governance
R0-R4 is retained as the effective risk tier. Risk drivers and governance obligations use the canonical Stage 5 AIPS vocabularies; the passport merely projects them for trust and discovery.
Assurance metrics
Metric names remain extensible because appropriate evidence differs across learned, symbolic, agentic and Physical AI. The passport standardizes the assurance envelope rather than imposing ML-only metrics.