Standing Requirement for Artificial Epistemic Agents

This paper specifies the necessary conditions for admissibility that preserve standing in artificial epistemic agents. It shows that admissibility is not achieved by performance, learning, optimization, or internal coherence, but only by satisfying standing-preserving constraints on reference, scope, and commitment. By exhaustion, all alternative criteria—probabilistic validation, success metrics, governance certification, or architectural sophistication—are eliminated as insufficient. The result fixes admissibility for artificial agents as a binary structural condition rather than a graded capability. No algorithms, architectures, or implementation procedures are provided; the analysis is necessity-based and eliminative.