Agent Risk Registry

Prototype

A reference registry that classifies AI agents by risk with defensible, evidence-backed entries — every score is reasoned against quoted evidence and ratified by an accountable human, not self-asserted.

Entries are scored against the ARC 12-dimension classification model (four groups, each dimension tiered 1–3) — the Responsible AI Institute's published ARC framework (arXiv working paper; full citation in the repository README). This is an independent working prototype by a single author, not an official RAI registry, product, or endorsement.

The registry

These are the entries every other view draws on — filter by risk tier or name, and select one to read its full record: scored dimensions, the quoted evidence behind each score, provenance, and any challenges. Each entry carries two independent marks: a risk tier (how risky the agent is) and a trust level (how reviewed the entry is).

Risk tier