Byoungyoung Lee 이병영

Professor, Electrical & Computer Engineering
Seoul National University
Director of the CompSec Lab
byoungyoung@snu.ac.kr

CV Google Scholar CompSec Lab

I work on systems security — attack mitigation, vulnerability finding, and confidential computing — spanning kernels, browsers, and trusted hardware. My recent focus is AI agent security: studying realistic attacks on LLM-based agents, such as prompt and data injection, and building principled defenses against them. More on our research →

AI Agent Security

Realistic attacks on LLM-based agents — prompt and data injection — and principled defenses like DualView and Prompt Flow Integrity.

Systems & Kernel Security

Fuzzing, concurrency bugs, and exploit mitigation across OS kernels and browsers — Razzer, OZZ, PeTAL, DMGuard, TikTag.

Confidential Computing

Trusted execution and data protection for cloud and ML pipelines — DLBox, Obliviate, SGX-Shield, TEEMATE.

To prospective students: I'm looking for motivated students in systems or security research. If you're interested in working with me, please reach out at byoungyoung@snu.ac.kr.

Research Updates

Jul 2026
Agent Data Injection: an arbitrary click attack against web agents — Woohyuk Choi
Jul 2026
DualView: defending personal AI agents like OpenClaw against indirect prompt injection — Juhee Kim
Jun 2026
Bad Epoll: the bug missed by Mythos — Jaeyoung Chung

Recent Publications

All publications →
2026
USENIX Security DMGuard: Safeguarding Kernels from Page-Use-After-Free Vulnerabilities
Juhee Kim⭑, Jaeyoung Chung⭑, Dae R. Jeong, and Byoungyoung Lee
USENIX Security Symposium · to appear
Systems Security
⭑ co-first authors
IEEE S&P GHost in the Shell: A GPU-to-Host Memory Attack and Its Mitigation
Sihyun Roh, Woohyuk Choi, Jaeyoung Chung, Yoochan Lee, Suhwan Song, and Byoungyoung Lee
IEEE Symposium on Security and Privacy
Hacking Systems Security
Ack: NRF-RS-2023-00209093, IITP-2020-0-01840
IEEE S&P Site Isolation is Dead: How Site Isolation is Broken in Agentic Browsers and Extensions
Suyoung Lee, Seongho Keum, Changoo Lee, Dongwon Shin, Sanghyun Hong, Byoungyoung Lee, and Sooel Son
IEEE Symposium on Security and Privacy
AI/Agent Hacking Web/Privacy
Ack: RS-2026-25470691
IEEE S&P Heap Localization: Cache Side-Channel based Linux Kernel Heap Exploit Techniques
Yoochan Lee, Sihyun Roh, Hyuk Kwon, Byoungyoung Lee, and Thorsten Holz
IEEE Symposium on Security and Privacy
Hacking Systems Security
Ack: ERC-101045669, DFG-EXC-2092-390781972, IITP-2020-0-01840
ISSTA AgentBreaker: Evaluating Context-Aware Indirect Prompt Injection Risks in Modern Web Agents
Yongbi Son, Changoo Lee, Dongwon Shin, Byoungyoung Lee, Sanghyun Hong, and Sooel Son
ACM SIGSOFT International Symposium on Software Testing and Analysis · to appear
AI/Agent Web/Privacy
arXiv Agent Data Injection Attacks are Realistic Threats to AI Agents
Woohyuk Choi⭑, Juhee Kim⭑, Taehyun Kang, Jihyeon Jeong, Luyi Xing, and Byoungyoung Lee
arXiv preprint
AI/Agent Hacking
⭑ co-first authors
arXiv DualView: Preventing Indirect Prompt Injection in Personal AI Agents
Juhee Kim⭑, Woohyuk Choi⭑, Taehyun Kang, Youngmin Kim, and Byoungyoung Lee
arXiv preprint
AI/Agent
⭑ co-first authors
2025
NDSS DLBox: New Model Training Framework for Protecting Training Data
Jaewon Hur, Juheon Yi, Cheolwoo Myung, Sangyun Kim, Youngki Lee, and Byoungyoung Lee
Network and Distributed System Security Symposium
Confidential Computing AI/Agent
Ack: NRF-RS-2023-00209093, IITP-RS-2024-00438729, NRF-RS-2023-00335515
NDSS Secure Data Analytics in Apache Spark with Fine-grained Policy Enforcement and Isolated Execution
Byeongwook Kim⭑, Jaewon Hur⭑, Adil Ahmad, and Byoungyoung Lee
Network and Distributed System Security Symposium
Confidential Computing Systems Security
⭑ co-first authors
Ack: NRF-RS-2023-00209093, COMPA-2710002717, AFOSR-FA9550-24-1-0204
IEEE S&P TikTag: Breaking ARM's Memory Tagging Extension with Speculative Execution
Juhee Kim, Jinbum Park, Sihyeon Roh, Jaeyoung Chung, Youngjoo Lee, Taesoo Kim, and Byoungyoung Lee
IEEE Symposium on Security and Privacy
Hacking Systems Security
arXiv Prompt Flow Integrity to Prevent Privilege Escalation in LLM Agents
Juhee Kim, Woohyuk Choi, and Byoungyoung Lee
arXiv preprint
AI/Agent
IEEE Access Finding Device Driver Bugs With Fuzzing PCIe Configuration Input
Kyungwook Boo, and Byoungyoung Lee
Systems Security
2024
ACM CCS PeTAL: Ensuring Access Control Integrity against Data-only Attacks on Linux
Juhee Kim, Jinbum Park, Yoochan Lee, Chengyu Song, Taesoo Kim, and Byoungyoung Lee
ACM Conference on Computer and Communications Security
Systems Security
Ack: IITP-2020-0-01840, NRF-RS-2023-00209093, DARPA-FA8750-24-2-0002
SOSP OZZ: Identifying Kernel Out-of-Order Concurrency Bugs with In-Vivo Memory Access Reordering
Dae R. Jeong, Yewon Choi, Byoungyoung Lee, Insik Shin, and Youngjin Kwon
ACM Symposium on Operating Systems Principles
Systems Security Best Paper Award
Ack: NRF-RS-2024-00347516, NRF-RS-2024-00359979, IITP-RS-2023-00232728, TIPA-TIPS-00262147, K-Startup-20144069, KISTI-K24L4M2C5
BlackHat USA Bypassing ARM's Memory Tagging Extension with a Side-Channel Attack
Juhee Kim, Jinbum Park, Sihyeon Roh, Jaeyoung Chung, Youngjoo Lee, Taesoo Kim, and Byoungyoung Lee
Hacking
ATC A Secure, Fast, and Resource-Efficient Serverless Platform with Function REWIND
Jaehyun Song, Bumsuk Kim, Minwoo Kwak, Byoungyoung Lee, Euiseong Seo, and Jinkyu Jeong
USENIX Annual Technical Conference
Systems Security
Ack: RS-2023-00321688
ASIACCS SyzRisk: A Change-Pattern-Based Continuous Kernel Regression Fuzzer
Gwangmu Lee, Duo Xu, Solmaz Salimi, Byoungyoung Lee, and Mathias Payer
ACM Symposium on Information, Computer and Communications Security
Systems Security
Ack: ERC-850868, SNSF-PCEGP2-186974, DARPA-HR001119S0089-AMP-FP-034
arXiv TEEMATE: Fast and Efficient Confidential Container using Shared Enclave
Chulmin Lee, Jaewon Hur, Sangho Lee, and Byoungyoung Lee
arXiv preprint
Confidential Computing Systems Security

All publications →