Hello!! I'm Brian Tang.
4th year CSE Ph.D. Candidate, CSE, University of Michigan.
I work on vision language models, AI, and privacy systems.
I'm a member of the RTCL advised by Kang G. Shin. I also collaborate with SpiLab and Florian Schaub.
In undergrad, I worked on machine learning security & privacy in WI-PI with Kassem Fawaz, Varun Chandrasekaran, and Somesh Jha.
Please contact me via bjaytang@umich.edu.

My past research projects involved creating real-time mobile security systems, evaluating LLMs for web automation and advertising risks, analyzing privacy policies with NLP, and creating adversarial ML attacks on face recognition systems. From these projects, I have co-authored numerous publications at top-tier conferences (USENIX Security, IEEE S&P, PETS). My current research focuses on creating real-time systems for vision language models, specifically with security and privacy applications. I am creating large-scale dataset collection pipelines for fine-tuning VLMs.