Virginia Tech® home

Seminar: Balancing User Experience and Privacy in Wearable AI Systems

Swetha Rajaram

PhD Candidate
School of Information,
University of Michigan

Tuesday, February 17, 2026
9:30 - 10:30 a.m.
1100 Torgersen Hall

 

Abstract

As wearable AI systems, such as smartglasses and augmented reality displays, are increasingly used across everyday settings, a core challenge is meeting users’ unique interaction needs in these contexts while safeguarding their privacy. In this talk, I illustrate how designing wearable AI interfaces to be highly reconfigurable can both reduce interaction friction and mitigate privacy risks for users and bystanders. First, I show how gesture and haptic interaction techniques, working in parallel with traditional speech and audio communication, can enable more flexible conversations with AI assistants on display-free glasses. Next, through co-design studies with AR and security & privacy researchers, I establish a design space of privacy-friendly alternatives to traditional AR interactions that reduce data exposure while preserving functionality. Finally, I present Privacy Equilibrium, an optimization framework for facilitating system-driven “negotiations” of sensing capabilities, aiming to automatically balance user experience and privacy for wearable AI users and bystanders. Together, this research argues that realizing the benefits of wearable AI requires these systems to dynamically adapt to our complex, and at times competing, goals. I conclude by discussing open challenges and future research towards this vision.

Biography

Shwetha Rajaram is a PhD candidate at the University of Michigan School of Information. She studies topics at the intersection of human-computer interaction and usable security and privacy. Her research develops interactions with emerging technologies (e.g., augmented reality, generative AI), aiming to enhance human workflows while safeguarding privacy. Her work has been published in premier HCI venues, such as the ACM CHI Conference on Human Factors in Computing Systems (CHI) and the ACM Symposium on User Interface Software and Technology (UIST). More information can be found at https://shwetharajaram.github.io/