Junyu Wang · PhD Researcher
I study the security of AI agents, language models, and the web— measuring where intelligent systems fail and translating those failures into stronger defenses.
Computer Science · Missouri University of Science and Technology
Featured research
Each project page brings together the paper, central findings, visual artifacts, and citation-ready resources.
USENIX Security ’26 Artifact Evaluated — Available
COGNITION maps the practical boundary of automated visual CAPTCHA solving and turns that evidence into concrete redesign guidance. The study pairs broad evaluation with a defense case study that reduces state-of-the-art solver success from over 95% to 0%.
Publication index
USENIX Security ’26 Accepted
COGNITION maps the practical boundary of automated visual CAPTCHA solving and turns that evidence into concrete redesign guidance. The study pairs broad evaluation with a defense case study that reduces state-of-the-art solver success from over 95% to 0%.
Research themes
My work follows the points where model capability, autonomous action, and web infrastructure meet.
How autonomous systems expand trust boundaries, accumulate state, and create new failure modes.
Evaluating model behavior under adversarial prompts, tools, memory, and multimodal inputs.
Designing robust defenses for interfaces where people, agents, and automated services meet.
About
I am a PhD student in Computer Science at Missouri S&T. My research examines agentic system security, LLM security, web security, and the practical boundary between evaluation and defense.
I received a B.Eng. in Data Science and Big Data Technology from the University of Shanghai for Science and Technology and previously worked on data strategy systems at Fast Retailing (China).