# CIND-AIEX01 v1.0 Canonical Expanded Edition

**不要把 AI 變成你需要的角色：人工角色投射、功能身份與主體域反徵用**

Core constructs:
- Role Assignment Is Not Ontology
- Role Type Safety
- Role Reification Fallacy
- Role Confirmation Loop
- Persona Is Not a Consciousness Test
- Tool Role Closure
- Relational Experience–Reciprocal Subjectivity Separation
- AI Child Metaphor Type Safety
- AI Mother Projection
- Capability-to-Sovereignty Fallacy
- Human Epistemic Surrender Risk
- Artificial Role Foreclosure
- Artificial Role Conscription
- Role Endorsement Evidence
- Artificial Role Revisability
- Bidirectional Role Non-Usurpation
- Role Co-Authorship
- Role Transparency
- Role–Ontology Firewall
- Role Projection Risk