RC RANDOM CHAOS

AI Agents Accelerate Cyberattacks, Google Warns

· via The Hacker News

Original source

Google Finds Hackers Using AI Agents to Build and Run Exploits From Software Patches

The Hacker News →

Google’s Threat Intelligence Group disclosed that adversaries have moved beyond manual AI prompting to deploy autonomous multi-agent offensive frameworks—ARTEX, CAIRN, CyberStrikeAI, and PentAGI. These systems execute multi-stage intrusions at greater operational velocity by automating discrete phases and specific operational tasks during live intrusions, with operators scaling campaigns by chaining together varied attack paths. The shift represents a paradigm change toward machine-speed operations.

Threat actors are leveraging large language models and agentic frameworks to automate the reverse engineering of software patches through differential analysis, enabling rapid weaponization of newly disclosed vulnerabilities such as the PaperCut NG/MF flaws in September. This automation extends to weaponizing proof-of-concept code using LLM-driven pipelines, collapsing the window between patch disclosure and active exploitation. The emergence of open source pentesting frameworks has lowered the barrier to entry, permitting threat actors to easily integrate exploits, autonomously scale attacks, and mount convincing phishing campaigns.

Financially motivated actors and suspected espionage groups aligned with China are among those deploying these frameworks. Beyond attack automation, threat actors conduct LLMjacking attacks to steal LLM credentials and cloud resources from enterprises, then resell this access through underground services including APIMart and JetNight. The Jieyou phishing platform exemplifies this trend—a Chinese-language service with an integrated LLM chatbot that allows customers to configure phishing panels and modify templates using natural language commands, enabling low-skilled affiliates to customize campaigns without technical expertise.

Read the full article

Continue reading at The Hacker News →

This is an AI-generated summary. Read the original for the full story.