The opportunity
About the Team The team’s mission is to accelerate the secure evolution of agentic AI systems at OpenAI. To achieve this, the team designs, implements, and continuously refines security policies, frameworks, and controls that defend OpenAI’s most critical assets—including the…
What you'll do
Architecting security controls for agentic AI: design, implement, and iterate on identity, network, and runtime-level defenses (e.g., sandboxing, policy enforcement) that integrate directly with the Agent Infrastructure stack.
Building production-grade security tooling: ship code that hardens safety monitoring pipelines across agent executions at scale.
Collaborating cross-functionally: work daily with Agent Infrastructure, product, research, safety, and security teams to balance security, performance, and usability.
Influencing strategy & standards: shape the long-term Agent Security roadmap, publish best practices internally and externally, and help define industry standards for securing autonomous AI.
Strong software-engineering skills in Python or at least one systems language: (Go, Rust, C/C++), plus a track record of shipping and operating secure, high-reliability services.
Deep expertise in modern isolation techniques: experience with container security, kernel-level hardening, and other isolation methods.
What they're looking for
- Hands-on network security experience: implementing identity-based controls, policy enforcement, and secure large-scale telemetry pipelines.
- Clear, concise communication that bridges engineering, research, and: leadership audiences; comfort influencing roadmaps and driving consensus.
- Bias for action & ownership: you thrive in ambiguity, move quickly without sacrificing rigor, and elevate the security bar company-wide from day one.
- Cloud security depth on at least one major provider (Azure, AWS, GCP),: including identity federation, workload IAM, and infrastructure-as-code best practices.
