Researcher, Recursive Self-Improvement Safety

Models are becoming increasingly capable—moving from tools that assist humans to agents that can plan, execute, and adapt in the real world. Mitigating the frontier risks resulting from these capabilities is paramount to OpenAI’s ability to continue deploying models safely. The Preparedness team is dedicated to addressing these critical risks. Our work includes: Measurement. Monitoring and predicting the evolving capabilities of frontier AI systems. Mitigation. Keeping misalignment safeguards, alignment tools, and on track to adequately address extreme threats that might arise in the future. Coordination. Setting mitigation targets by maintaining OpenAI’s preparedness framework , and partnering with other staff to achieve these targets. This is urgent, fast-paced work that has far-reaching implications for the company and for society.

Preparedness is hiring strong technical executors to support preparations for accelerated AI development, which may culminate in recursive self-improvement. This work relies on anticipating misalignment risks that might exist in the future, but might not exist now; so it’s especially important that people in this role are tasteful and strategic. The role is wide-ranging, covering any mitigation for loss of control risk, spanning the design and implementation of better pre-deployment risk-assessment , control measures , RSI-relevant training interventions, and turning one’s technical work into established institutional practices and external-facing communications. Below is a subset of our focus areas: Scalable oversight: Establishing practices for model misbehavior monitoring and oversight which remain effective in superhuman model capability regimes, with a focus on bridging from today’s monitoring approaches to future-proof ones. Automated auditing: As model capabilities increase, we’ll increasingly rely on automated approaches for finding the most severe forms of model misalignments. We’ll both need to sift through large swaths of production traffic to find the most egregious misalignments, and reliably elicit tail risks before deployment. Rigorous monitorability: Rigorous testing and red-teaming of our measurements of model misbehavior related to loss-of-control (e.g. reward hacking, sandbagging, scheming). This includes better understanding monitorability , and e.g. preparing for potential losses of Chain-of-Thought monitorability. Model behavior science : Design experiments and evaluations to understand the extent to which models are problematically misaligned, or their safety-relevant capabilities lag behind dangerous capabilities. This may include training model organisms of misbehavior for behaviors not currently present in production, or training interventions to increase safety-relevant capabilities. Coordination and verification : Prototype technical mechanisms for verifying compliance with potential future AI safety agreements. AI R&D risk measurement: Track progress toward automation of technical staff to inform OpenAI’s near-term investments in alignment and security. Maintaining and strengthening RSI safety cases: We’re especially interested in identifying and addressing blindspots of mitigation areas which we may have missed. Generally, our team alternates between performing rigorous hypothesis-driven research and turning our insights into interventions or control systems which impact production models, with occasional support of engineering teams. In this role, you will: Carefully consider the problems OpenAI might face in the future and how to prepare for them. Turn an open-ended objective like “prepare for future misalignment threats” into a much more concrete direction (e.g. “stress-test monitors for scheming”) – prioritizing the work that is most useful to start right now. Execute quickly, building scrappy prototypes, and then improving them iteratively until they become established components of our safety