About

I am a Master’s candidate in the Department of Electrical Engineering at POSTECH, advised by Prof. Yongjune Kim. My research interests lie in AI privacy and security. In particular, I have primarily worked on privacy-enhancing technologies (PETs), especially Homomorphic Encryption (HE), to develop privacy-preserving and efficient AI inference systems.

My broader research vision is to understand and navigate the trade-offs among security, utility, and efficiency (computation and communication costs) in AI Security. Since modern AI systems face diverse attacks, threat models, and deployment requirements, no single defense mechanism can serve as a universal solution—nor are these techniques mutually orthogonal. I am therefore interested in developing hybrid, robust AI security frameworks that combine multiple AI security techniques—including Homomorphic Encryption, other PETs such as Multi-Party Computation and Differential Privacy, and non-PET defense mechanisms—to defend against diverse attacks while preserving practical utility and efficiency.

Recent news

Publications

* equal contribution