KAIST Graduate School of Information Security

2025.11.21(화) 정보보호대학원 2025년 가을학기 콜로퀴움 - Hsu-Chun Hsiao
작성일2025-12-02
KAIST 정보보호대학원에서는 11월 21일 오후 4시에 아래와 같이 콜로퀴움을 개최하고자 합니다. 많은 참석 부탁드립니다. o 일 시 : 25. 11. 21(금) 16:00~ o 주 제 : Uncovering and Mitigating Privacy Risks of Bots in Group Chats o 강 사 : Hsu-Chun Hsiao (National Taiwan University, 교수) o 장 소 : N1동 201호 ※ 시작시간 5분전에 준비하여 주세요. ㅡㅡㅡ ♣ Title : Uncovering and Mitigating Privacy Risks of Bots in Group Chats ♣ Abstract This talk explores the privacy implications of chatbots in group messaging platforms. In the first part of the talk, we demonstrate that chatbots frequently access far more messages and metadata than necessary, allowing them to track and link users across different groups. We analyze conversation datasets to quantify this threat and propose SnoopGuard, a new secure group messaging protocol. SnoopGuard mitigates these risks by enforcing selective message access and sender anonymity against chatbots, ensuring user privacy while maintaining end-to-end security. In the second part of the talk, we examine the user perspective through a large-scale survey. Our findings reveal a critical lack of user awareness: many users are unaware of bots in their group chats and significantly underestimate the extent to which chatbots can access their data. We introduce a five-stage model that details how users detect, interpret, and respond to these privacy risks, highlighting mismatches between current platform designs and user expectations. ♣ Bio Dr. Hsu-Chun Hsiao is a Professor in Computer Science and Information Engineering at National Taiwan University. She completed her B.S. and M.S. at National Taiwan University in 2006 and 2008, respectively, and her Ph.D. at Carnegie Mellon University in 2014. She led the security team in the Information Technology Office at National Taiwan University Hospital from 2018 to 2025. Her research focuses on building practical security solutions with theoretical underpinnings. She is a recipient of several awards, including Young Scholar Fellowship and Ta-You Wu Memorial Award from Taiwan Ministry of Science and Technology, IEEE S&P Test-of-Time Award, and PETS Best Artifact Award.

2025.11.21(화) 정보보호대학원 2025년 가을학기 콜로퀴움 - Hsu-Chun Hsiao

작성일2025-12-02

KAIST 정보보호대학원에서는 11월 21일 오후 4시에 아래와 같이 콜로퀴움을 개최하고자 합니다. 많은 참석 부탁드립니다.

o 일 시 : 25. 11. 21(금) 16:00~
o 주 제 : Uncovering and Mitigating Privacy Risks of Bots in Group Chats
o 강 사 : Hsu-Chun Hsiao (National Taiwan University, 교수)
o 장 소 : N1동 201호

※ 시작시간 5분전에 준비하여 주세요.

ㅡㅡㅡ

♣ Title : Uncovering and Mitigating Privacy Risks of Bots in Group Chats

♣ Abstract
This talk explores the privacy implications of chatbots in group messaging platforms. In the first part of the talk, we demonstrate that chatbots frequently access far more messages and metadata than necessary, allowing them to track and link users across different groups. We analyze conversation datasets to quantify this threat and propose SnoopGuard, a new secure group messaging protocol. SnoopGuard mitigates these risks by enforcing selective message access and sender anonymity against chatbots, ensuring user privacy while maintaining end-to-end security. In the second part of the talk, we examine the user perspective through a large-scale survey. Our findings reveal a critical lack of user awareness: many users are unaware of bots in their group chats and significantly underestimate the extent to which chatbots can access their data. We introduce a five-stage model that details how users detect, interpret, and respond to these privacy risks, highlighting mismatches between current platform designs and user expectations.

♣ Bio
Dr. Hsu-Chun Hsiao is a Professor in Computer Science and Information Engineering at National Taiwan University. She completed her B.S. and M.S. at National Taiwan University in 2006 and 2008, respectively, and her Ph.D. at Carnegie Mellon University in 2014. She led the security team in the Information Technology Office at National Taiwan University Hospital from 2018 to 2025. Her research focuses on building practical security solutions with theoretical underpinnings. She is a recipient of several awards, including Young Scholar Fellowship and Ta-You Wu Memorial Award from Taiwan Ministry of Science and Technology, IEEE S&P Test-of-Time Award, and PETS Best Artifact Award.

목록으로

세미나