kwangilkim

AI Safety & Security for LLMs, Agent Systems, and Autonomous AI. Focused on runtime guardrails, adversarial robustness, RAG security, and real-world deployment in regulated environments (finance, healthcare, defense).