
This role within OpenAI's Safety Systems team focuses on defining model policies for frontier cyber risk, ensuring AI systems align with human values in high-stakes cybersecurity contexts. Key responsibilities include designing policy frameworks for dual-use risks, translating complex threat models into behavioral specifications, and collaborating with engineering teams to implement measurable safeguards. The position sits at the intersection of offensive security, AI safety, and policy implementation. It offers the appeal of working at the forefront of AGI safety with a mission to protect humanity from misuse while enabling legitimate defense. The role is based in San Francisco with a hybrid schedule requiring three days in the office, supported by a collaborative culture and comprehensive on-site amenities.




















