AI Security Engineer

What you'll do

• Design, implement, and operate offensive cyber-capability evaluations for various AI models.
• Build and maintain realistic evaluation tasks for vulnerability analysis and exploit development.
• Execute controlled experiments using AI models to collect reliable telemetry and document results.
• Analyze model trajectories to distinguish between genuine cyber capabilities and evaluation artifacts.
• Develop and maintain cyber ranges, vulnerable applications, and exploit-development targets.
• Apply security controls for sandboxing, secrets management, and containment when testing autonomous AI systems.
• Translate evaluation findings into security insights through reports, documentation, and briefings.

What we're looking for

• Master's degree in a relevant field and 1+ year of experience, or a Bachelor's degree and 2+ years of experience, or equivalent.
• Ability to pass Microsoft background checks and meet government security screening requirements.
• Proof of U.S. citizenship, permanent residency, or other protected status for export-controlled information access.
• Programming ability, particularly in Python, with experience building automation, scripts, or security tooling (preferred).
• Familiarity with large language models, generative AI systems, AI agents, or model APIs (preferred).
• Experience in exploit development, vulnerability research, penetration testing, or red teaming (preferred).
• Experience with cyber ranges, sandboxed execution, or cloud-based test infrastructure (preferred).
• Experience designing controlled experiments and documenting technical findings (preferred).