Why it matters
Extends the governed threat surface beyond model and input integrity to the runtime governance evaluation layer. Existing cyber, identity, data, tool, memory, retrieval, and output controls remain necessary.
Control relationship
Authorized meaning must remain connected to the meaning operationalized at runtime.
SLIA is a threat class targeting the semantic representations or infrastructure on which runtime governance relies; it is not a component of the Semantic Control Plane.