The Chinese government already has a history of attacking US infrastructure, raising concerns of current security standards for cyber physical systems. With the advent of AI, these attacks can be developed and deployed at a faster rate. Prof. Madnick of MIT demonstrated with simulations that these attacks can cause physical damage, such as explosions, and would require much more time to get systems back up and running versus a ransomware attack. While escalating to physical attack has a natural deterrent for nations of possible causing international conflict, non state actors can implement these attacks more easily with AI and possibly escalate terrorism. However, Prof. Tehlia of Teshiva university points out that AI can also help with defenses and detection.
