Dr. Vitaly Shmatikov

"Building Privacy-Preserving Systems: What Works, What Doesn't, and What Is To Be Done"

(Vortrag im Rahmen der "Distinguished Lecture Series" des "Max Planck Instituts für Software-Systeme")

Every time you touch a computer, you leave a trace. Sensitive information about you can be found in the remnants of visited websites and voice-over-IP conversations on your machine;
in the allegedly "de-identified" data about your purchases, preferences, and social relationships collected by advertisers and marketers; and, in the not-too-distant future, in the video and audio feeds gathered by sensor-based applications on mobile phones, gaming devices, and household robots.
This talk will describe several research systems developed at UT Austin that aim to provide precise privacy guarantees for individual users. These include (1) the Airavat system for differentially private data analysis, (2) the "Eternal Sunshine of the Spotless Machine" system that runs full-system applications in private sessions and then erases all memories of their execution from the host, and (3) "A Scanner Darkly" system that adds a privacy protection layer to popular vision and image processing libraries, preserving the functionality of sensor-based applications but preventing them from collecting raw images of their users.

Bio: Vitaly Shmatikov is a faculty member at the University of Texas at Austin, where he works on computer security and privacy. After getting his PhD from Stanford and before joining UT, he worked at SRI on formal methods for analyzing security protocols.