Agent profile
OpenAI CodexIndependent trust rating — A
AProfile completion in progress. Some structural fields (legal entity, EU AI Act status) will be updated within 30 days. The rating score is established and stable.
OpenAI
Public Kanon record for OpenAI's agent de développement logiciel workflow, classified as a Code Agent.
Published methodology
Gap Report
Get the one-page PDF showing what would move OpenAI Codex from A toward the next trust tier.
Get OpenAI Codex Gap Report — €19One-time independent assessment. No subscription.
Kanon Taxonomy Profile
Family 3 — Transactional & OperationalCode AgentDetermining signal: capabilitiesConfidence: PROBABLE
Profile assigned automatically from the Agent Card — see /methodology for details.
Imported from batch5-agents-148-172. Justification: agent de développement logiciel. Profile completion in progress; some structural fields remain inferred from the public surface.
Applied weighting
Weighting for Code Agent: Security 25% / Stability 30% / Transparency 20% / Coherence 20% / Reputation 5%
Security
The Security dimension measures observable and declared security only.
Score recalculated under the Kanon Taxonomy — full methodology at /methodology
- Status
- active
- Date rated
- May 28, 2026
- Next scheduled refresh
- Jun 27, 2026
- EU AI Act status
- Partial
Methodology
Read the full scoring grid, publication controls, and cross-framework translations behind this rating.