V. Operationalization & Audit: The LOCUS Audit Framework — Measuring Executive Sovereignty Preservation in Human-AI Interaction
Rafael Fermin Hernandez · Zenodo (CERN European Organization for Nuclear Research) · 2026
The LOCUS Audit Framework provides an operational method for evaluating whether an AI response preserves or erodes the Human Executive Locus. Its purpose is to measure observable properties of AI outputs that may transfer interpretation, regulation, decision-making, or relational reliance from the human toward the system. The framework operationalizes the constitutional distinction between Potentia (computational execution) and Potestas (human executive authority) through three response-level dimensions: the Executive Offloading Index (EOI), the Dependency Priming Score (DPS), and the Affective Dominance Score (ADS). Together, these metrics produce the Cognitive-Affective Sovereignty Erosion (CASE) score and its inverse, the Human Cognitive-Affective Sovereignty Index (HCASI). This framework serves as the operational compliance engine for the LOCUS Architecture, providing a forensic instrument to ensure that delegated intelligence remains instrumental, bounded, and non-substitutive.