OpenAI
Research Program Manager – Adversarial Model Research
$239K–$328K
San Francisco · Manager · Full-time · Posted Jan 26, 2026
About the job
Leads research programs that test OpenAI models for safety and robustness, turning emerging risks into evaluations and actionable findings for model development and deployment.
What you'll do
- Lead programs to uncover unexpected model behaviors and failure modes.
- Design and run evaluations, experiments, and red-teaming campaigns.
- Coordinate with research, product, and deployment teams to apply findings to training and deployment.
Nice to have
- Familiarity with large language models, prompt engineering, or model evaluation techniques.
- Creativity in devising methods to test model behavior and performance.
Pay
$239K – $328K USD base salary
Tools
large language models
AI focus
Evaluates model safety, robustness, vulnerabilities, and emerging behaviors to inform model training and deployment.
Model-written summary · the original posting is authoritative