General Analysis
AI/Cyber Researcher
San Francisco, CA · Full-time · Posted Oct 29, 2026
About the job
Research AI cybersecurity capabilities and risks, develop evaluation challenges and assessment methods, and investigate ways to mitigate misuse and other threats.
What you'll do
- Create and expand security evaluation challenges, including CTF-style exploitation tasks and network attack simulations.
- Research capability elicitation, jailbreak methods, evaluation generalization and coverage, and risks such as AI misuse, model weight theft, and loss of control.
- Publish findings and deliver research results to customers.
What they ask for
- Strong background in vulnerability or security research, or deep research experience with frontier AI models.
- Ability to design and implement novel research approaches for emerging AI challenges.
- Publications in top-tier venues such as NeurIPS, ICML, ICLR, or USENIX Security, or comparable applied work.
AI focus
Evaluate frontier models' cyber capabilities, develop elicitation and jailbreak methods, assess evaluation coverage and generalization, and research mitigations for AI-related security risks.
Model-written summary · the original posting is authoritative