PREPRINT · 17 SEPTEMBER 2026
Standing to Stop: Authority provenance before human oversight becomes real
A stopping point is not enough. The party at it needs presently valid, traceable authority, and the stop must actually change what can happen.
OWAI RESEARCH
Research behind
the architecture.
Our research examines the conditions under which a system’s conclusions may become actions. The programme spans stopping points, authority provenance, execution boundaries and the behaviour of governed systems over time.
Each publication has a summary on this site and a full, citable version with a DOI on Zenodo. They are preprints and working papers that provide arguments and formal models. Product assurance is assessed separately against implemented behaviour and deployment evidence.
SELECTED PREPRINTS & WORKING PAPERS
PREPRINT · 12 MARCH 2026
Architectural Constraints in AI Governance: Execution Legitimacy, System Trajectories, and Structural Boundaries in Probabilistic Decision Systems
Governance of AI-enabled operational systems has to live in the architecture that decides which state transitions are admissible — including across sequences of individually acceptable steps.
TECHNICAL NOTE · 13 JANUARY 2026
An Architectural Framework for Enforceable Decision Boundaries in AI Systems
Inference may propose actions but never authorise them. Authority is resolved outside the model, and execution is impossible without it.
PUBLICATION · 11 JANUARY 2026
Prompt Engineering as a Control Illusion
Prompts influence what a model says. They cannot decide what a system is permitted to do.
ESSAY · JANUARY 2026
The Missing Place for “Stop”: Why systems that work correctly still fail
Systems that work well for years can still drift into decisions nobody can defend — because no one is structurally placed to stop them before execution.