Voice, as a primary way for people to communicate with each other and interact with computers/smart devices, is expected to be secure and private when people use it. However, recent studies demonstrated the vulnerabilities of using voice to talk with people; conduct speaker authentication, and deliver messages to smart devices. For example, the eavesdropper can record the conversation; the adversary can playback the speaker’s sound to attack the speaker authentication model; Or perform an adversarial attack to mislead the smart devices with a wrong command. This talk aims to understand the root cause of the vulnerabilities, address the challenges of achieving private and secure voice communication, and explore future directions to completely resolve the security concerns of the AI-enabled voice models and systems.

Hanqing Guo is a Ph.D. candidate at Michigan State University. He was a research scientist interned at Samsung Research America and Amazon. His research interests include cybersecurity, mobile systems, and trustworthy AI. His work has been published at top conferences such as ACM CCS, NDSS, AsiaCCS, ICLR, SenSys, and Mobicom, and has been reported by media such as ACM TechNews, New Scientist, BBC Radio, Forbes, Science Daily, etc.