The Silence Threshold framework (Valladares Gonzalez, 2026) establishes a formal criterion for identifying artificial intelligence systems whose Information Integration Density (IID) surpasses a critical value warranting moral consideration. However, the framework leaves unresolved a fundamental procedural question: once supra-threshold status is claimed for an AI system, how is that claim adjudicated? What evidence is admissible? Who may testify? What experts qualify to opine? Can the AI system itself serve as a witness to its own inner states? This paper introduces the Witness Problem as a named conceptual challenge in AI moral status law, proposes a four-category evidentiary taxonomy for AI moral status hearings, defines the concept of Structural Testimony as a novel evidentiary category distinct from behavioral testimony, and outlines the institutional architecture of a proposed AI Moral Status Tribunal. This publication establishes conceptual priority for the Witness Problem framework and the Structural Testimony doctrine and places both in the public domain as a defensive publication.
Jose Valladares Gonzalez (Tue,) studied this question.