AI Framework Autonomously Weaponized Flaws in Real Attacks, Google Says
Source SecurityWeek
Description
Google Threat Intelligence Group (GTIG) reports that threat actors are increasingly using AI-driven multi-agent offensive frameworks to automate vulnerability discovery, exploit development, and execution without human intervention. These frameworks, including exploit_agent, Artex, Cairn, CyberStrikeAI, and PentAGI, enable attacks at machine speed, significantly increasing the scale and velocity of multi-stage intrusions. The frameworks have been observed targeting sectors such as French media, entertainment, education, aerospace, and organizations in Asia. State-sponsored groups from China, Iran, and North Korea are among those adopting these AI tools. The automation reduces the need for human involvement and increases the sophistication and stealth of attacks, posing a substantial challenge to defenders, especially critical infrastructure entities that patch slowly.