AISYSTEMS knowledge layer · Public technical reference
A stable vocabulary for AI audit evidence.
AI Core is a public, descriptive ontology for consistent terminology across AI observation, audit artifacts and long-term technical evidence. It defines language and boundaries; it does not execute audits.
Clear boundary
A reference layer, with a deliberately narrow role.
AI Core keeps terminology stable while implementation, decision-making and enforcement remain outside the ontology.
What AI Core provides
- Stable definitions for audit and observation contexts.
- Immutable releases addressable by version and hash.
- Human-readable documents and canonical JSON-LD.
- A shared vocabulary usable by independent systems.
What AI Core does not do
- It does not execute audits or modify observed systems.
- It is not a compliance framework or certification scheme.
- It does not rank, score or enforce outcomes.
- It has no regulatory authority.
Conceptual position
From an observed system to traceable evidence.
Source Domain
The AI system, model, dataset or process being observed remains logically separate.
Audit Domain
A version-bound Audit Plan and Runner describe and perform observation without changing the source.
Evidence Layer
Evidence and manifests preserve the result together with version, time and ontology anchors.
Immutable archive
Releases remain available for reproducibility.
Changes are published under a new version. Earlier documents remain accessible at their original URLs.
Machine-readable access
Stable endpoints for people and systems.
Use the versioned JSON-LD for reproducible references. The root navigation files point agents to the current stable release.
One public ecosystem
A technical reference within AISYSTEMS.
AI Core shares one public layer with the AISYSTEMS company site, documentation and AI INDUSTRY knowledge publication.