Offensive Security Agent Engineer

Build and lead OpenAI’s offensive security agents that continuously discover, validate, and remediate vulnerabilities across infrastructure and applications. You’ll establish the agents’ architecture, evaluation strategy, and operating model, integrating internal context with live feedback to test cloud and Kubernetes environments, web applications, endpoints, and external attack surface. The work includes end-to-end vulnerability-management loops, human-in-the-loop safety controls, and production-grade infrastructure for safe, observable, continuous execution.