What question did this study set out to answer?

The paper aims to clarify misconceptions about the relationship between deterministic computation and authorized decision-making in AI systems.

March 28, 2026Open Access

Deterministic Engines and Deterministic Governance: Why Correctness Is Not Permission

View Full Paper

Ask AI

Bookmark

Authors

This analysis reveals flaws in equating deterministic systems with authorized decision-making, suggesting new frameworks for governance.

The paper aims to clarify misconceptions about the relationship between deterministic computation and authorized decision-making in AI systems.
Conducted a systems-architecture analysis of deterministic engines and governance
Distinguished three layers of governance: engines, safety configurations, and authorization
Introduced proof-carrying authorization artifacts and disqualifying tests for governance claims
Highlighted the insufficiency of deterministic outputs for ensuring compliance and liability
Established the importance of independent verification for governance claims
Clarified the distinction between logs and proof artifacts in governance

Edward Meyman (2026) studied this question.

View Full Paper

Ask AI

Bookmark