Researcher, Recursive Self-Improvement Safety

Develop technical mitigations for AI safety risks from recursive self-improvement, including oversight systems, auditing, and behavioral monitoring.
- Design and implement pre-deployment risk assessments, control measures, and training interventions for loss-of-control scenarios.
- Build scalable oversight practices and automated auditing approaches effective across superhuman model capability regimes.
- Conduct rigorous testing, red-teaming, and experiments to understand model misalignment and safety-relevant capability gaps.
- Requires exceptional technical execution, strategic research taste, and ability to prioritize in domains with weak feedback loops.
- Bonus: prior work in ML research, AI alignment, verification, or related safety domains.

OpenAI is a frontier AI research and product company, with teams working on alignment, policy, and security. We post specific opportunities at OpenAI that we think may be high impact. We do not necessarily recommend working at other positions at OpenAI. You can read concerns about doing harm by working at a frontier AI company in our career review on the topic, including concerns about OpenAI in particular. Note that there have also been concerns around OpenAI's HR practices.