大模型智能体安全研究员-智能安全
Core
Systematically discover, model, and assess security risks for large model agent systems, including prompt injection, tool misuse, and data leakage, while designing defense architectures and conducting red teaming.
Role type
Senior IC AI security researcher (large model agents)
Builds
Standardized security benchmarks, automated attack simulation tools, and red teaming frameworks for agent systems
Domain
AI safety / Large Language Models / Agent Systems
Deliverable
production ML models
Required skills
Threat modeling, Prompt injection defense, Red teaming, Python programming, Risk assessment, System architecture design
Preferred skills
Academic publications in AI security, CTF/AI security competition experience, Security engineering, Traditional cybersecurity background
Technologies
Python, LLM frameworks, RAG, Function Calling, ReAct
Responsibilities
Design input/output filtering and sandboxing strategies for tool calls; Build and execute automated adversarial testing frameworks; Collaborate with product and engineering teams to integrate safety components into the agent product lifecycle.
Seniority
Senior, hands-on IC