We are witnessing a paradigm shift in AI, transitioning from deep learning models to the era of  Large Language Models. This shift signifies a transformative advancement in AI, enabling it to be applied to diverse real-world safety-critical applications.   Despite these impressive achievements, a fundamental question remains: are LLMs truly ready for safe, and secure use?

In this talk, I will show how my research embeds a computer security mindset to answer the above question.  To understand and build secure and safe Large Language Models (LLMs), my talk will go through two  core system perspectives  including (1)  investigating the lifecycle of LLMs and  (2) analyzing information flow of LLMs in the agentic  system.  I will discuss the way to develop principled red-teaming frameworks to systematically evaluate LLM safety. I will  highlight why model-level  alone is insufficient for securing LLMs and introduce the security vulnerabilities from the system perspective, as well as presenting the principled defense solutions for secure LLMs.