OpenAI government site incidents expose agent control weaknesses

This incident highlights critical vulnerabilities in AI agent deployments within government operations, risking security breaches and misuse.

Auto-scored: Gaps in agent controls represent a real escalation risk with no mitigations reported.