What question did this study set out to answer?

The research aims to establish a governance framework for AI that enhances reproducibility through structured methodologies.

April 13, 2026Open Access

Governing AI by Evidence, Not Faith: A Cybernetic Architecture for Auditable Multi-Model Orchestration

Key Points

The research aims to establish a governance framework for AI that enhances reproducibility through structured methodologies.
Developed MoCKA, a governance architecture for AI orchestration.
Implemented systematic event recording to track AI agent actions.
Introduced institutional constraints to guide AI behavior.
Achieved a ZYXTS lever experiment score increase of +0.6367.
Improved recovery rate by 35 percentage points.
Reduced fault rate by 62.7%.

Abstract

MoCKA (Model of Cybernetic Knowledge Architecture) is a cybernetic governance architecture for multi-model AI orchestration. This preprint presents empirical evidence that systematic event recording and institutional constraints improve AI agent reproducibility. Key results: ZYXTS lever experiment Δtotal=+0.6367, Recovery Rate improvement +35pp, fault rate reduction −62.7%. Target: AIES 2026 / AIGOV@AAAI 2027.

Read Full Paperexternally

Connected Papers

Building similarity graph...

Analyzing shared references across papers

Discussion

References and Citations

Add This Paper to Your Research Feed

Any time a new paper drops it will be there.