IS 400 Colloquium: Yaman Yu

Generative AI is rapidly transforming how young users engage with technology, extending beyond education and entertainment to emotional support, virtual relationships, and even risky behaviors. Yet, parents often underestimate the extent and complexity of these interactions. As AI chatbots become more human-like with advancements in Large Language Models (LLMs), traditional moderation struggles to address emerging safety risks, including boundary violations and harmful content.

In this talk, I will present our work on identifying and mitigating youth safety risks in AI-driven interactions. Through an in-depth analysis of real-world usage, we uncovered unique challenges young users face. To address these, we developed YAIR, a risk taxonomy capturing key safety concerns, and designed an LLM-powered parental control system to help families navigate AI use effectively.

Bio

Yaman Yu is a Ph.D. candidate in the School of Information Sciences at the University of Illinois Urbana-Champaign, advised by Dr. Yang Wang. Her research bridges security and human-computer interaction (HCI), focusing on safeguarding AI systems and mitigating harms from AI misuse, particularly for vulnerable populations. Her work has been published in top security, privacy, and HCI conferences, including the IEEE Symposium on Security and Privacy, USENIX Security, and the ACM Conference on Human Factors in Computing Systems (CHI). She has received a Best Paper Honorable Mention at CHI, won the Berkeley LLM Agent Hackathon, and her research has been featured in media outlets such as The New York Times, Business Insider, and the Illinois News Bureau.

About

The iSchool Colloquium (IS 400) is a venue for presentation and discussion of research and professional activities by faculty, students, staff, and guest speakers. The colloquium is hosted in person and virtually. All are welcome to attend.

Questions? Contact Yingjun Guan.