← All sections
§19

The three guardrail layers

Adopted from Elementum. Guardrails sit outside the AI system, not inside the model. The model's built-in safety is not a substitute. Three layers, each catching a different class of failure:

LayerWhere it livesWhat it doesExample
Policy guardrailsDocuments + the registryDefines who can do what, with which data, under which constraints, before a system runs"AI cannot approve refunds > $5,000 without human review."
Workflow guardrailsThe orchestration layerTurns policy into deterministic approvals, RBAC, business rules, routingThe $5,000 threshold routes the request to a named reviewer with context
Runtime guardrailsExecution-timeInput filtering, output validation, tool allowlists, schema checks, semantic filters, kill switchesBlock PII in prompts; block tool calls outside the allowlist; block outputs that don't conform to the expected schema

All three are required. Policy alone is theater. Workflow without runtime checks misses prompt injection. Runtime without policy means each agent invents its own rules.