jobs
Researcher, Cyber Safety Training
About the Team Safety Research at OpenAI is responsible for ensuring our most capable models can be safely deployed in the real world. The Safety Training team develops the research and engineering…
En inglés
- Company
- OpenAI
- Location
- San Francisco
- Status
- Open
- Posted
- 2026-11-21 00:40 UTC
OpenAI's Safety Training team seeks a researcher to shape frontier model behavior on cyber safety. The role involves designing realistic cyber safety evaluations, studying over-refusal and harmful assistance, researching post-training methods, and moving them into large training runs at scale. It asks for hands-on model training or fine-tuning experience, practical security judgment or reinforcement-learning practice, and ability to build and debug training experiments. Collaboration with cybersecurity researchers, trust and safety, legal, policy and other research teams is central; the distinctive focus is balancing help for legitimate security work against misuse.
Original job posting