AI安全蓝军 - 安全与风控
Core
Develop autonomous security agents to automate offensive testing, vulnerability discovery, and threat modeling for the company's massive business assets and AI applications.
Role type
Senior AI Security Researcher (Offensive/Red Team)
Builds
Automated offensive security agents, vulnerability scanning workflows, and AI safety guardrails
Domain
Cybersecurity + Artificial Intelligence (LLM Security)
Deliverable
production ML models
Required skills
Offensive security (web/mobile/binary), penetration testing, ATT&CK methodology, Python/Golang/Rust/JavaScript, LLM architecture, Agent frameworks, MCP/Skill injection analysis
Preferred skills
Automated vulnerability discovery, AI safety research, security competition experience
Technologies
Python, Golang, Rust, JavaScript, LLMs, Agent Harness, MCP
Responsibilities
Design and execute automated penetration tests using AI agents; Develop core capabilities for autonomous threat modeling and attack planning; Build safety guardrails and audit trails for agent execution in production; Conduct red team exercises and vulnerability research for AI applications (prompt injection, skill poisoning); Create and maintain a knowledge base of offensive methodologies and agent skills; Establish data-driven evaluation metrics for automated security capabilities.
