jobs

Researcher, Agent Safety, Oversight and System Mitigations

About the TeamThe Agent Safety team works to ensure that increasingly capable AI agents act safely, exercise sound judgment, and remain aligned with user intent. Our mission is to reduce the probability of severe unintended outcomes from in

Company

OpenAI

Location

San Francisco

Status

Open

Posted

2026-09-03T16:34:21.597+00:00

OpenAI's Agent Safety team seeks a researcher or engineer to build oversight and system-level mitigations so capable agents can act autonomously without harm. Work includes designing and evaluating controls such as agent-based review, sandboxing, process isolation and permission boundaries; red-teaming end-to-end agentic systems for data exfiltration and unsafe tool use; and reducing missed harmful actions, unnecessary blocks and approval burden. The role partners with a Codex harness engineering team to productionize controls. Strong systems or security instincts matter; a background in AI control or security is welcome but not required.