Model Policy Manager, National Security

Written by the T&S Network from OpenAI's posting.

Model policy work for a frontier AI model used by U.S. government customers, focused on national security applications. The role defines how models should behave in high-stakes scenarios and turns those rules into training and evaluation criteria, sitting in the Model Policy team within Safety Systems.

What you'd do

• Write behavior rules for models used in national security work, balancing safety with usefulness.
• Build evaluations to find policy gaps and model failures.
• Use evaluation findings to revise policies and training criteria.
• Turn hard safety questions into plain, workable policy wording.
• Work hands-on with model data and evaluation results.
• Partner with research, engineering, and domain experts on safe deployment.

What it asks for

• Background in AI safety, policy, or risk assessment.
• Strong judgment on complex safety questions.
• Hands-on technical skill with model data and evaluations.
• Active TS/SCI clearance or equivalent (preferred but not required).

Worth knowing

• Work is tied to U.S. government use of the models in safety-critical settings.

Terms in this role

Red teaming.

Testing a system, often an AI model, by deliberately trying to make it fail or misbehave, so problems are found and fixed first.

Our best summary of the posting, and it can be wrong. The employer's posting is what counts: read it in full on OpenAInorth_east before you apply. If we got something wrong, report this listing.