Loading...
Mid LevelFull-timeRustAWSGitAIML
Salary Range
$97,750 - $132,250 /year
This salary is estimated based on similar roles. The actual salary may vary.
About the Role
We are looking for a Security Researcher to join OpenAI’s Codex Security team. This role focuses on mapping, characterizing, and prioritizing cross-layer vulnerabilities in advanced AI systems.
Key Responsibilities
- →Conduct deep security research on real-world software systems to discover complex vulnerabilities across large codebases and distributed architectures.
- →Investigate and validate vulnerabilities discovered by AI-driven security agents, including building proofs-of-concept and exploit demonstrations.
- →Partner with engineering teams to improve automated vulnerability discovery, validation, and remediation workflows as part of product development.
- →Build high quality security datasets and evals that will help advance model’s cybersecurity capabilities
- →Train and improve AI models used for vulnerability discovery, validation, and automated remediation by developing datasets, evaluations, and feedback loops grounded in real-world security research.
- →Publish technical writeups, research insights, and vulnerability analyses that advance the state of application security.
Requirements
- ✓Have strong experience in vulnerability research, exploit development, or offensive security.
- ✓Have deep experience with cutting edge offensive-security techniques
- ✓Are fluent across AI/ML infrastructure (data, training, inference, schedulers, accelerators) and can threat-model end-to-end.
- ✓Operate independently, align diverse teams, and deliver on tight timelines.
- ✓Communicate clearly and concisely with experts and decision-makers.
- ✓Care deeply about improving the security of widely used software and open-source infrastructure.
- ✓Are a strong developer who can work in a small energetic team
- ✓Key technical challenges
- ✓System-level vulnerability discovery — identifying complex vulnerabilities that span multiple services, trust boundaries, or components.
- ✓High-confidence validation — distinguishing real exploitable vulnerabilities from speculative or theoretical issues.
- ✓Scaling security research with AI agents — guiding automated systems to analyze millions of commits while maintaining research-level rigor.
- ✓Automated exploit and proof-of-concept generation — building reproducible demonstrations of vulnerabilities within sandboxed environments.
- ✓Building large systems that work within OpenAI's enterprise architecture
Free to sign up. Takes 30 seconds.
