Lead security reviews for AI-powered applications, LLM integrations, AI agents, MCP-based architectures, and retrieval-augmented generation (RAG) systems.Develop secure design patterns for prompt handling, model access, tool permissions, context isolation, and AI data protection.Assess risks including prompt injection, indirect prompt injection, insecure tool use, model abuse, excessive agency, data leakage, and supply chain risks.Partner with AI engineering teams to implement guardrails, model governance, runtime monitoring, and secure AI deployment patterns.Establish security guidance for AI coding assistants and developer AI workflows.