Google’s cyber threat intelligence division identified hackers on Monday who used AI to “fully automate vulnerability discovery and exploitation without a human in the loop for the first time.”