Why do you care about AI Existential Safety?

AI has been, and will continue to be, integrated into real-world systems across various domains, including critical ones. Just like conventional software systems, it is crucial to validate and verify the safety and security of such systems to prevent malicious manipulation and catastrophic consequences. We need software and systems perspectives to guide the design and deployment of AI in ways that ensure safety, reliability, and security.

Please give at least one example of your research interests related to AI existential safety:

My research focuses on the security and safety of AI-enabled systems, aiming to empower system providers and individual users to counteract attacks. I am broadly interested in a range of topics in security and privacy relating to machine learning, including adversarial generative AI, security/privacy of LLM agents, and machine learning for security.