Description

Abstract

Algorithmic systems now make or influence millions of consequential decisions daily — allocating benefits, screening job applicants, setting bail, approving credit. Yet the institutions that authorize these systems frequently cannot verify whether they operate as intended, producing what this book terms the Sovereignty Gap: the measurable disconnect between institutional authority and algorithmic execution. When the gap is open, civil-rights harms accumulate faster than oversight can respond.

This book proposes and operationalizes the Luevano Standard — a tripartite framework of Consistency, Transparency, and Accountability — as a prerequisite for delegating institutional authority to automated systems. The Standard is named in honor of Angel Luevano and the landmark civil rights consent decree Luevano v. Campbell (1981), which established that facially neutral procedures with discriminatory effects violate Title VII. That precedent is directly applicable to algorithmic decision systems today.

The Luevano Standard is implemented through four engineering mechanisms: Logic-to-Monitor Synthesis (LMS), which compiles statutory requirements into runtime-verifiable predicates; Auditable Reasoning Traces (ARTs), which create tamper-evident records of every decision sufficient for legal review; Deterministic Execution, which ensures reproducible, auditable outputs; and Data Provenance, which records the lineage of every datum used in a decision. Oversight is provided by the Independent Luevano Registry (ILR), a proposed tripartite meta-governance body representing state, industry, and civil society.

The framework is grounded in runtime verification theory (Leucker & Schallhart, 2009), legal formalization scholarship (Sergot et al., 1986), and civil rights doctrine (Griggs v. Duke Power Co., 1971; Mathews v. Eldridge, 1976). It is positioned as a runtime enforcement layer that operationalizes existing governance frameworks including the NIST AI Risk Management Framework (2023), ISO/IEC 42001:2023, and the EU AI Act (Regulation (EU) 2024/1689).