Title: Building and Stress-Testing Robust AI Agent Systems for Real-World Deployment
Date: Tuesday, May 5th, 2026
Time: 13:00 - 14:30 PM EDT
Location: Online
Zoom: https://gatech.zoom.us/j/5256702657?pwd=bWyj8saWbSVdLlrhb7PEHJYlZ5bTO5.1

Yanzhe Zhang
Ph.D. Student in Computer Science
School of Interactive Computing
Georgia Institute of Technology
https://stevenyzzhang.github.io/website/

Committee members
Dr. Diyi Yang (advisor) - Computer Science Department, Stanford University
Dr. Zsolt Kira (co-advisor) - School of Interactive Computing, Georgia Institute of Technology
Dr. Kartik Goyal - School of Interactive Computing, Georgia Institute of Technology
Dr. Polo Chau - School of Computational Science & Engineering, Georgia Institute of Technology

Abstract
Large language models (LLMs) are rapidly evolving from passive text generators into autonomous, multimodal agents that can perceive, reason, and act in the real world. This transformation unlocks new capabilities, while also reshaping human–AI interaction and introducing novel safety risks. In this thesis proposal, I present my research on building and stress-testing robust AI agent systems. First, I introduce methods for constructing and evaluating agent systems that enable new capabilities, including automatically generating websites from visual designs and generative interfaces that structure user interaction with AI systems. Second, I examine emerging risks in these agents, showing that computer-use agents are vulnerable to pop-up attacks and that large-scale simulation can systematically uncover privacy risks. Finally, I propose a framework for studying the fundamental reliability challenges underlying these failures, with the goal of understanding and improving robustness in real-world deployments.