OpenAI
Model Policy Manager, National Security
$207K–$335K
San Francisco +2 · Hybrid · Manager · Full-time · Posted Oct 16, 2026
- Location
- San Francisco
Washington, DC
US - Remote
About the job
Shapes model behavior for U.S. government use, especially national security, by developing policies and translating them into training and evaluation criteria.
What you'll do
- Develop policies that guide safe and useful model behavior.
- Build evaluations, identify policy gaps and model failures, and use findings to improve policies and training.
- Collaborate with research, engineering, and domain experts to support safe, reliable deployment.
What they ask for
- Relevant experience in AI safety, policy, or risk assessment.
- Strong judgment and ability to turn complex safety questions into practical policies.
- Technical fluency to work hands-on with model data and evaluations.
- Eligibility and willingness to obtain and maintain a U.S. government security clearance.
Pay
$207K – $335K (USD, base salary)
AI focus
Develops model policies, training criteria, and evaluations for safe and useful behavior in national security applications.
Model-written summary · the original posting is authoritative