Ai System Type
Objective
Capture and validate the ai system type element within the Generative, Agentic & Control Plane layer.
UAIF requirement
The incident record should implement the canonical ai_system_type field using the defined UAIF data type, validation constraints, and conditional rules.
Business impact
Captures generative, agentic and control-plane context needed to investigate escalation and system-to-system propagation.
Validation approach
Validate ai_system_type for controlled vocabulary membership. Then review semantic consistency against the source incident evidence and the selected conformance profile.
Expected evidence
Retain the source record or analyst input for ai_system_type, schema-validation output, transformation history, selected profile, schema version and reviewer rationale where judgement is involved.
Mapping and source
UAIF Schema Specification; UAIF Technical Specification; JSON pointer /properties/ai_system_type.
Implementation guidance
Populate ai_system_type from an identified source or documented analyst decision. Enforce the schema constraints at record creation and update, preserve the original value and provenance, and surface validation failures without silently coercing data.
Assessment considerations
Sample records across normal, boundary and invalid conditions. Confirm that ai_system_type is populated only when applicable, conditional rules are enforced, provenance is retained and downstream systems do not change its meaning or precision.
Canonical schema definition
| JSON property | ai_system_type |
|---|---|
| Required in core profile | No |
| Type | "string" |
| Allowed values | LLM, VLM, RAG_Pipeline, Agentic, Multi_Agent, Traditional_ML, Other |
Source: UAIF Schema Specification and UAIF Core JSON Schema.