CIRIS has a new look. Visit the new site โ†’
InstallEpistemic WebSafetyHow it worksResearchAccordGitHub

โ† All papers

Framework paper

CIRISAgent Framework

An open-source ethical AI framework for accountable autonomy

Eric Moore ยท 2026-01-02 ยท DOI: 10.5281/zenodo.18137161

Abstract

The framework paper. A 22-service architecture organized around explicit action verbs and ethical reasoning, with transparency built into the structure rather than bolted on afterward.

Key findings

  • Service catalog: identity, events, federation, verification, safety, lens, persistence, conscience, judge, accord, scoring, and the rest.
  • Each service is independently inspectable; the architecture's defense is the structure, not a model-level guardrail.