AGENT INFRASTRUCTURE

OpenAI documents the security boundaries behind its persistent-agent deployment

By IDAKIM Digital · Published November 13, 2026

New engineering guidance separates agent execution from trusted control systems and combines sandboxing, identity and review.

OpenAI’s November 12 engineering post describes controls used for its internal OpenClaw agents. The architecture separates the trusted gateway from the agent harness so dynamically generated code runs inside an untrusted workload boundary. The guidance covers network allowlists, workspace restrictions, process isolation and credential brokering, alongside model-based action review. Separate agent identities support auditing and revocation, while restricted logs capture configuration changes and agent activity. OpenAI reports more than 500 persistent agents in its deployment. The post also discusses OpenClaw Enterprise, an existing open-source control-plane project still in pilot; this article is newly published deployment guidance rather than a claim that the project launched that day. Its example settings reflect OpenAI’s environment and require adaptation to another organization’s workloads.

Why it matters

Map the trusted control plane, credential boundary and permitted destinations before enabling unattended work. Test a denied action and audit access without exposing secrets in ordinary logs.

Source: OpenAI · November 12, 2026

Product claims are attributed to the linked publisher. IDAKIM Digital has not independently tested these claims.