OpenAI is looking to mitigate risks in advanced AI systems by designing evaluations, surfacing vulnerabilities, and collaborating with researchers to strengthen model reliability and public trust. The Technical Program Manager will lead initiatives to test the safety and robustness of OpenAI's models through creative experimentation and structured evaluation, transforming ambiguous risks into concrete research programs and influencing future model development and deployment.
Requirements
- familiar with large language models, prompt engineering, or model evaluation techniques.
Responsibilities
- Lead programs that explore unexpected model behaviors and identify failure modes.
- Translate vague or emergent risk signals into clear priorities and actionable research plans.
- Design and run creative evaluations, experiments, and red-teaming campaigns.
- Collaborate with research, product, and deployment teams to integrate findings into model training and deployment cycles.
- Develop repeatable systems for tracking model performance and understanding emerging behavior patterns.
Other
- Have strong experience in technical program management, with excellent organizational and communication skills.
- Are comfortable managing fast-paced, high-uncertainty projects and shaping them from the ground up.
- Are creative and resourceful in devising new methods for testing model behavior and performance.
- Can effectively coordinate across technical and non-technical stakeholders to drive alignment and execution.
- This role is based in San Francisco, CA. We use a hybrid work model of 3 days in the office per week and offer relocation assistance to new employees.
