C15 INCYBER Adversarial AI Challenge Accepted!

Digital apocalypse or springboard to innovation? Generative AIs can learn from our vulnerability databases, logs and CTFs, and compromise a system in seconds. But they can also detect the behavior of their malicious brethren. As long as AIs need to communicate via networks, they will use IP addresses and leave traces in our application logs. This session is an invitation to turn adversity into a collective response. Faced with such an adversary, unity is strength. Let’s draw up together a concrete, actionable plan to reduce our surface exposure to AI and exploit their weaknesses. Yes, offensive AIs have an Achilles’ heel.

Allez à la source