Staff AI Security Specialist
Core
Build security controls for AI agent systems to prevent adversaries from subverting reasoning, tool-calling, and action capabilities.
Role type
Staff AI Security Specialist (applied security IC)
Builds
Security controls, threat models, red-team exercises, and evaluation benchmarks for agentic systems.
Domain
AI Security / Agentic Systems / Software Engineering
Deliverable
production ML models
Required skills
AI/ML or LLM system security, agent harness security (prompt/context/tool loops), security domain depth (app/offense/identity/detection), AI threat landscape knowledge (prompt hijacking, memory poisoning), Python development
Preferred skills
Runtime security decisioning, static/dynamic analysis tooling, adversarial ML research
Technologies
Python, MCP, OWASP Top 10 for LLM Applications, MITRE ATLAS
Responsibilities
Analyze security of AI agent systems and find subversion points; Design and build security controls (static analysis, runtime policy, behavioral detection); Red-team own agents and defenses; Build repeatable evaluations and benchmarks; Advise product teams on secure design; Track offensive/defensive AI security research.
Seniority
Staff, hands-on IC with mentorship
