安全产品经理（Agent安全方向）-AI创新业务
Core
Design and implement the end-to-end runtime security system for AI Agents, covering task planning, tool invocation, execution environments, and security quantification to support safe deployment across business lines.
Role type
Product Manager (AI Agent Security)
Builds
AI Agent security capabilities and governance frameworks for internal business lines
Domain
Artificial Intelligence / Cybersecurity
Deliverable
production ML models
Required skills
Product planning, risk analysis, abstracting attack scenarios, data-driven strategy design, security metrics definition, cross-functional collaboration
Preferred skills
Technical background (code/CLI/logs), familiarity with Shell/file systems/network/processes, experience in security evaluation/red teaming/data platforms
Technologies
AI Agents, Prompt Injection, Tool Invocation, Supply Chain Security, Sandbox Security
Responsibilities
Lead product planning for Agent runtime security; design core security modules for prompt injection and data leakage; collaborate with red teams to abstract attacks into product solutions; build security evaluation and go-live standards; track global security trends and optimize internal frameworks.