Has the era of fully automated cyberattacks begun?

Google has revealed the first documented use of an AI system capable of fully automated vulnerability discovery and exploitation, targeting French websites in media, entertainment, and education. The attackers are believed to be Russian-speaking.

Google has disclosed the first documented use of an AI system capable of fully automated vulnerability discovery and exploitation, without human intervention in the execution stages. Google's Threat Intelligence Group, in a report published with cybersecurity firm Mandiant, stated that attackers believed to be Russian-speaking used a system comprising multiple AI agents to conduct continuous operations for identifying software vulnerabilities and developing tools to exploit them. The activities targeted French websites in the media, entertainment, and education sectors. Researchers found stolen data from targeted entities on a server used by the attackers to develop and test their tools. Politico reported that Google researchers said this is the first time they have observed advanced language models being used to fully automate both vulnerability discovery and exploitation. This marks a shift from using AI as a hacking aid to operating systems capable of independently executing successive stages of attacks.

Has the era of fully automated cyberattacks begun?