Cryptography for Emerging Technologies

ABSTRACT

Ensuring the security and privacy of emerging technologies is crucial in the modern world. This talk presents my research on enhancing security and utility for three key technologies: outsourced databases, generative AI, and private information retrieval. We will outline leakage-abuse attacks against end-to-end encrypted database systems to show vulnerabilities in current proposals. Next, we show how cryptography is used to build watermarks for generative AI outputs — improving the robustness and utility of the watermarks over prior work. Finally, we boost the state-of-the-art private information retrieval schemes to match known lower bounds and provide more efficient database updating.

BIOGRAPHY

Alex Hoover recently received his PhD from the University of Chicago, where he has continued as a postdoctoral scholar. He also works as a part-time postdoc with collaborators at Columbia University. His primary research interest is in developing cryptography for new technologies; however, he is broadly interested in many different fields within and outside of computer science. Recently, most of his research has focused on understanding the security and efficiency of end-to-end encrypted systems, developing safety tools for generative AI (e.g., watermarks), and improving the state of the art for private information retrieval. He has previously worked on election complexity and quantum computing and has ongoing work in cryptographic complexity and secure data pooling. In his free time, Alex enjoys learning about other disciplines, such as the natural sciences and philosophy, as well as reading fantasy/fiction.