Authority Resolution Framework: A Five-Domain Ontology for AI Governance
A new paper on arXiv (2608.15832) introduces the Authority Resolution Framework (ARF), designed to help guide decision-making in AI systems through a five-domain ontology. The idea is that as AI becomes more autonomous, it’s not just about being technically skilled; these systems also need to assess whether they have the right to act based on the context. ARF includes authority from different roles, informal influences, business concepts, formal processes, machine-readable permissions, and real-world situations. It features the Authority Relation (AR) as a key element that connects an actor, action, object, context, justification, and a measure called the DNA-Coefficient to analyze authority discrepancies. The framework also provides a machine-readable format for authority using JSON-LD.
Key facts
- Paper arXiv:2608.15832v1 introduces the Authority Resolution Framework (ARF).
- ARF is a five-domain ontology for representing and resolving authority.
- The five domains include organizational roles, business concepts, codified processes, machine-readable permissions, and external context.
- ARF defines the Authority Relation (AR) as a cross-domain primitive.
- The DNA-Coefficient measures divergence between documented and practiced authority.
- The framework uses JSON-LD for machine-interpretable representation.
- The paper argues that AI systems must determine authorization, not just capability.
- The paper is announced as a new submission on arXiv.
Entities
Institutions
- arXiv