news
Running Codex safely at OpenAI
May 8, 2026
OpenAI describes how it runs Codex securely using sandboxing, approval flows, network policies, and agent-native telemetry to support coding-agent adoption. The setup matters because it reduces the blast radius of agent actions while giving operators the controls and observability needed for safe, compliant use.
How OpenAI runs Codex securely with sandboxing, approvals, network policies, and agent-native telemetry to support safe and compliant coding agent adoption.
Source: openai.com