ARTEX LLM-Driven Offensive Framework Exploration and Detection
Score: 10/10

ARTEX LLM-Driven Offensive Framework Exploration and Detection

The ARTEX LLM-driven framework is being utilized by multiple threat actors for automated reconnaissance, authentication probing, and multi-vulnerability exploitation targeting web applications and cloud environments.

Executive Summary

Previdian and MalwareMustDie have identified widespread exploitation attempts associated with ARTEX, an LLM-driven agentic offensive security framework. Between October and November 2026, global sensors recorded thousands of HTTP records involving application discovery, remote code execution (RCE), and sensitive data retrieval attempts. The framework demonstrates the ability to adapt payloads based on application feedback, specifically targeting Oracle WebLogic, Langflow, and Flowise.

Technical analysis reveals that ARTEX and its forks (AntiFox, ScopeWeaver, Ayven) automate complex attack chains, including authentication bypass, path traversal, and credential enumeration. Observed activity is linked to campaigns by suspected actors Storm-0940 and Storm-3358, targeting IT vendors and financial sectors. This highlights a shift toward machine-speed offensive operations where LLMs compress vulnerability research and exploitation timelines into minutes.

This threat matters because ARTEX lowers the barrier for sophisticated, adaptive attacks against internet-facing infrastructure. While successful compromise was not confirmed in all honeypot telemetry, the high volume of targeted exploits against known vulnerabilities (e.g., Log4Shell, Struts OGNL) poses a significant risk to organizations with unpatched legacy or AI-related application interfaces.

Key Details

Threat Name

ARTEX

Affects

Langflow, Flowise, Oracle WebLogic, Apache Log4j, Apache Struts, PHPUnit, Apache Tomcat

Adversary

Storm-0940 Other Adversaries and Aliases: Storm-3358

Malware/Tools

ARTEX, AntiFox, ScopeWeaver, Ayven, BLOODALCHEMY, Chisel

Report Score

10out of 10
Quality Score
Excellent
IOC Quality10
TTP Details9
Detection Guidance10
Enterprise Relevance10
Clarity & Structure10
Technical Depth9

Sources