A Taxonomy of AI Governance Approaches: Distinguishing Visibility, Alignment, and Authorization

The term “AI governance” has become semantically overloaded, applied indiscriminately to logging, guardrails, model alignment, dashboards, and policy workflows. This paper introduces a formal taxonomy that distinguishes three fundamentally different governance problems—visibility (what happened), alignment (is the system generally safe), and authorization (was this specific action permitted under policy)—and maps common vendor “governance” claims to the problems they actually solve. The taxonomy defines deterministic AI governance as a pre-execution authorization layer in which identical governed state yields identical governance verdicts and the system emits verifiable decision artifacts sufficient for independent third-party offline replay. It specifies the normative structure of governed state, the Minimum Evidence Package (Normative), an enforcement triad (ALLOW / DENY / ABSTAIN), and disqualifying evaluation methods, including anti-laundering checks designed to prevent trust-based or vendor-dependent imitation of governance. This work is intentionally testable and disqualifying rather than aspirational. It is intended to give enterprise buyers, regulators, auditors, and researchers a shared vocabulary and concrete conformance criteria for evaluating governance claims in high-stakes, regulated deployments. Update (Jan 2026): Released a three-document companion kit while leaving the canonical taxonomy paper (v1.5) unchanged: an Executive Summary (2 pages) for board/C-suite orientation (including a capability comparison matrix, the Five-Point Diligence Test, and the complete AI governance stack), an updated Vendor Evaluation Supplement with a decision tree to determine when authorization-layer governance is required, and Evidence Package Examples providing three domain-specific, independently replayable worked artifacts demonstrating ALLOW, DENY, and ABSTAIN verdicts and the Minimum Evidence Package (Normative) requirements.