CS Colloquium: Securing Modern System is More Challenging Than Ever (and requires new and dedicated guardrails).

Speaker: Ben Nassi, Technion, Black Ha

Location: 60 Fifth Avenue, Room 150

Date: Wednesday, March 12, 2025

Over the past decade, an increasing number of systems and devices have gained Internet connectivity and been enhanced with sensing capabilities and AI. While these advancements have created a world of smarter, more automated, and highly connected devices, they have also introduced significant security and privacy challenges that cannot be effectively addressed with traditional countermeasures. The first part of this talk will focus on side-channel attacks. We discuss the challenges posed by the coexistence of functional devices with limited computational power (that do not adhere to Moore’s law) alongside sensors with ever-increasing sampling rates. We will explore how threats such as cryptanalysis and speech eavesdropping—previously accessible only to well-resourced adversaries—can now be executed by ordinary attackers using readily available hardware like photodiodes and video cameras. These attacks leverage optical traces or video footage from a device’s power LED to extract sensitive information. The second part of the talk will focus on AI security. We explore the unique challenges AI-powered systems face in the physical realm and a countermeasure to secure Teslas against time-domain adversarial attacks. Finally, we discuss the emerging need to secure LLM-powered applications against a new threat we call Promptware that exploits an application’s GenAI component to perform malicious activity. We explore Morris-II, a variant of Promptware that behaves as an AI worm.