E2 AI Epistemic Legitimacy Anchors Tensors Skin

This paper formalizes epistemic legitimacy in artificial intelligence using the anchor–tensor–skin hierarchy as a standing-preserving classification scheme. Anchors fix non-negotiable constraints, tensors carry load-bearing structure, and skin represents expressive variation, with legitimacy arising only when these roles are respected. The analysis is necessity-based and eliminative: any AI system that attempts to generate legitimacy through performance, learning, consensus, or post-hoc validation is shown by exhaustion to violate standing and fail closed. The treatment is declarative and non-procedural, recording the structural conditions for AI epistemic legitimacy without proposing architectures, training methods, or operational pipelines.