RFC-034: Domains of Concern for Mediated Intelligence — 15. Minimal Domain of Concern Declaration

AIGP Specification › RFC-034: Domains of Concern for Mediated Intelligence › 15. Minimal Domain of Concern Declaration

← 14. Required Properties of a Domain of Concern · Section index · 16. Design Principles →

15. Minimal Domain of Concern Declaration

{
  "domain_id": "autonomous_systems",
  "name": "Autonomous Systems",
  "description": "A mediation domain for AI-enabled systems that may recommend, select, or execute actions within a bounded operational envelope.",
  "concern_class": [
    "unauthorized_autonomous_action",
    "out_of_envelope_behavior",
    "high_consequence_action_under_uncertainty",
    "insufficient_human_control"
  ],
  "included_artifact_classes": [
    "action_recommendation",
    "action_execution_record",
    "mission_plan",
    "system_decision_trace",
    "override_record"
  ],
  "excluded_artifact_classes": [
    "general_information_response",
    "non-actionable summary"
  ],
  "admissible_evidence_types": [
    "authorization_record",
    "intent_declaration",
    "operational_envelope",
    "environment_state",
    "decision_trace",
    "human_approval_record",
    "safeguard_state",
    "outcome_record"
  ],
  "constraint_pattern": [
    "authority",
    "intent_alignment",
    "boundary_compliance",
    "environment_confidence",
    "human_control",
    "reversibility",
    "consequence_severity",
    "traceability",
    "uncertainty"
  ],
  "observer_modes": [
    "machine",
    "human",
    "hybrid"
  ],
  "governance_requirements": [
    "artifact_must_declare_domain",
    "evidence_must_be_traceable",
    "out_of_envelope_actions_must_escalate",
    "high_consequence_actions_must_require_human_control"
  ]
}

---

---

← 14. Required Properties of a Domain of Concern · Section index · 16. Design Principles →