Deterministic Governance Is Multi-Dimensional: Beyond Authorization in AI Systems

This paper introduces the Multi-Dimensional Deterministic Governance Model, a vendor-neutral framework for evaluating AI governance architectures in regulated environments. It argues that execution-time authorization, while necessary, addresses only one of five architectural dimensions required for deterministic governance. The five dimensions are: governance ontology (stabilizing the meaning of governed terms), constraint codification (transforming policy into machine-verifiable rules), deterministic decision logic (guaranteeing identical outcomes for identical governed state), proof-carrying decisions (enabling independent replay and verification of decision correctness), and execution authorization (enforcing outcomes at the action boundary with fail-closed behavior). The paper provides a diagnostic evaluation framework — five questions, one per dimension — that organizations can use to assess whether a governance system provides governance completeness or only execution enforcement. It demonstrates that missing dimensions create distinct, observable failure modes that no amount of strength in the remaining dimensions can compensate for. This work is intended for enterprise buyers, regulators, CISOs, general counsel, and governance architects evaluating AI governance infrastructure for high-consequence environments including healthcare, financial services, government, and legal operations. It is published as a preprint to support open dissemination and prior-art establishment. Related publications: A Taxonomy of AI Governance Approaches (DOI: 10.5281/zenodo.18275969), Versioned Meaning: How to Make Ontologies Audit-Stable (DOI: 10.5281/zenodo.18328587), Governance Laundering: A Taxonomy of Failure Modes in AI Compliance Architectures (DOI: 10.5281/zenodo.18746522), The Enterprise AI Governance Buyer's Guide (DOI: 10.5281/zenodo.18002693).