The Authorization Artifact Test: Applying the Impossibility Result to Ex-Ante Authorization Requirements

This regulatory application note operationalizes the impossibility result established in On the Impossibility of Observability-Based Authorization (Meyman, 2026; Technical Note, Version 1.4.0; concept DOI 10.5281/zenodo.19647542) for requirements that call for authorization before execution. It introduces the Authorization Artifact Test, a two-prong, regime-neutral instrument for determining whether a candidate governance architecture can, in principle, satisfy the artifact-related threshold of an applicable pre-execution authorization requirement. The test asks whether an action-bound verdict exists before execution and whether an independent third party can reconstruct that verdict from the authorization artifact and its bound materials, under a declared replay mode, without access to the governed system. Reconstruction may proceed under State-Replay or Protocol-Replay as defined by the Five Tests Standard; the artifact may carry the required materials or reference them through authenticated bindings recorded in a self-contained Evidence Package. Reconstruction that depends on vendor-controlled interpretation, vendor-controlled live services, mutable interpreters, or unpreserved live state does not satisfy the test. If either prong fails, the architecture cannot satisfy such a requirement, regardless of latency, automation, sophistication, or vendor description. Passing both prongs establishes only the Authorization Artifact Test threshold: it does not establish non-bypassability, fail-closed enforcement across the declared execution topology, ownership of authority, governed escalation, provenance, or the admissibility of the inputs grounding the verdict, and it does not establish ABIM Output Integrity, Input Integrity, full Replay Integrity, or 5TS conformance. The note applies the test, in conditional form, to contexts arising under the EU AI Act (Article 14), the General Data Protection Regulation (Article 22), the HIPAA Security Rule technical safeguards (45 C.F.R. § 164.312), DFARS 252.204-7012, and the NIST AI Risk Management Framework's GOVERN function. It does not interpret those instruments and does not assert that any of them expressly requires the artifact defined here. Whether pre-execution authorization is required is determined by the applicable authority or organization; the test determines whether the architecture can produce the corresponding authorization artifact. Where such a requirement has been established, the note derives the consequences of the impossibility result for architectures commonly described as guardrails, monitoring systems, observability platforms, and human review of system outputs. The dividing line is structural: a model-generated candidate held pending at a non-bypassable runtime authorization boundary may be the subject of a conforming pre-execution verdict emitted by that boundary; a verdict rendered only after release, or dependent on unreconstructable observation of the governed system, is not authorization. Human resolution may participate in the governed procedure by supplying authority-bound input; the runtime authorization boundary emits the operative action-bound verdict on which execution depends. The contribution is classificatory rather than prescriptive. The note is an architectural analysis, not legal advice, a legal opinion, or a compliance determination, and it does not introduce new doctrine or prescribe implementation. It supplies an operational instrument for applying the formal result in regulatory hearings, conformity assessments, audits, supervisory reviews, and standards-body work, distinguishing architectures that produce pre-execution, independently reconstructable verdicts from those that do not. Version 1.2 aligns Prong Two with the State-Replay and Protocol-Replay modes, recognizes that an authorization artifact may carry or reference its bound materials, places verdict emission exclusively at the runtime authorization boundary including after authorized human resolution, states the relationship between the Authorization Artifact Test threshold and conclusions under the Authorization Boundary Integrity Model and the Five Tests Standard, limits deployment-level positive results to the conditions actually evaluated, and narrows the test's positive claim to the artifact-related threshold of an applicable requirement. Intended audience: regulators, regulatory counsel, standards-body participants, and conformity-assessment bodies. This note is part of the FERZ research program on deterministic AI governance. The full corpus is available at https://zenodo.org/communities/ferz/.