Safety of GenAI Through the Lens of Security and Cryptography

In this 32-minute talk from the Simons Institute's "Theoretical Aspects of Trustworthy AI" series, Professor Somesh Jha from the University of Wisconsin-Madison examines the safety challenges of Generative AI through security and cryptography perspectives. Explore how principles from these established fields can be applied to address emerging concerns in AI safety, offering theoretical frameworks for building more trustworthy AI systems.