AEGLIX News
According to Google Threat Intelligence Group, threat actors are deploying autonomous multi-agent offensive frameworks such as ARTEX, CAIRN, CyberStrikeAI, and PentAGI to carry out multi-stage intrusions with increased speed. The firm described this adoption of AI agentic frameworks as a paradigm shift where attacks unfold at machine velocity, enabling operations to scale by linking diverse attack paths even when core techniques stay the same. Google also documented the use of ZeroHollow, an inference service for exploit development and LLM routing, alongside Agent Zero, an open-source agent framework capable of autonomous code execution and dynamic tool creation.
Various threat actors, among them financially driven groups and suspected espionage operators linked to China, have been seen testing these frameworks and toolsets. Open-source tooling reduces the barrier to entry, letting adversaries readily incorporate fresh exploits, scale attacks autonomously, and craft persuasive phishing campaigns that shrink attack timelines. Adversaries employ LLMs and agentic frameworks to automate patch reverse engineering through differential analysis, which accelerates weaponization of newly disclosed flaws, as demonstrated with PaperCut NG/MF vulnerabilities earlier this September. LLMs can also be used to quickly weaponize patched n-day vulnerabilities by comparing source code versions, reviewing release notes, security research, and proof-of-concept code tied to specific weaknesses.
Adversaries automate vulnerability assessment and proof-of-concept weaponization via LLM-driven pipelines, narrowing the gap between patch release and active exploitation. The Chinese-language phishing platform Jieyou added an LLM chatbot that lets customers configure phishing panels and alter templates through natural language commands without deep technical skill. Google noted that Jieyou's evolution illustrates a trend where Phishing-as-a-Service providers treat LLMs as modular force multipliers, allowing low-skilled affiliates to dynamically tailor pretext lures and phishing interfaces on demand. Threat actors also target AI infrastructure for LLMjacking by stealing LLM credentials and cloud resources from compromised enterprises and reselling access via underground proxy services such as APIMart and JetNight at reduced rates.
Google Threat Intelligence Group stated that misuse of open-source agentic penetration-testing frameworks automates distinct phases and specific tasks during live intrusions. The company disclosed that threat actors employ autonomous multi-agent offensive frameworks including ARTEX, CAIRN, CyberStrikeAI, and PentAGI to conduct multi-stage intrusions at higher operational speed. Google reported observing ZeroHollow, an inference service for exploit development and LLM routing, and Agent Zero, an open-source agent framework with autonomous code execution and dynamic tool generation capabilities. The firm characterized the adoption of AI agentic frameworks as a paradigm shift where cyber attacks occur at machine speed, permitting operations to scale significantly by chaining varied attack paths even when underlying mechanics remain unchanged.
Report prepared 2026-11-16 · Based on reporting from The Hacker News. Source publication date above; subsequent developments may not be reflected.