Disguising Attacks with Explanation-Aware Backdoors

Explore a 14-minute IEEE conference talk that delves into the concept of explanation-aware backdoors as a means of disguising attacks. Learn from researchers Maximilian Noppel, Lukas Peter, and Christian Wressnegger from KASTEL Security Research Labs at Karlsruhe Institute of Technology in Germany as they present their findings on this sophisticated cybersecurity threat. Gain insights into how attackers can potentially exploit explanation methods to conceal malicious activities and understand the implications for current security measures.