ARTEX Agentic AI Framework Adopted Globally by Adversaries
Score: 10/10

ARTEX Agentic AI Framework Adopted Globally by Adversaries

State-sponsored and criminal threat actors are using the ARTEX agentic AI framework to automate high-velocity reconnaissance, cloud exploitation, and data exfiltration.

Executive Summary

ARTEX, an open-source autonomous AI penetration testing framework released in July 2026, has been rapidly adopted by multiple threat clusters including PRC-state-aligned actors (Storm-0940, Vapor Panda), DPRK-linked groups (Onyx Sleet, Pearl Sleet), and financially motivated actors (Storm-3358). The framework uses Large Language Models (LLMs) to decompose natural language objectives into autonomous attack steps, enabling a single operator to sustain a volume of concurrent intrusions previously requiring a full team.

Technically, ARTEX excels at high-speed cloud reconnaissance and exploitation, documented to complete full attack chains from GitHub secret discovery to Azure Key Vault exfiltration in under five minutes. While currently 'loud'—generating over 1,500 commands per hour—its speed often outpaces human-driven detection and response cycles. Adversaries have successfully leveraged ARTEX in supply chain compromises (ConnectWise), zero-day exploitations (Zammad), and large-scale data breaches at financial institutions.

The widespread availability of ARTEX and its emerging stealth-focused forks represent a significant shift in adversary tradecraft. Organizations must transition from reactive patching to proactive identity hygiene, secret rotation, and behavioral monitoring of high-velocity API activity to counter machine-speed autonomous threats.

Key Details

Threat Name

ARTEX Agentic AI Framework

Affects

Zammad helpdesk platform

Adversary

Storm-0940 Other Adversaries and Aliases: Vapor Panda; Storm-3358; Onyx Sleet; Pearl Sleet

Malware/Tools

BLOODALCHEMY, ARTEX, WISP, vshell, AdaptixC2, SmallTiger

Report Score

10out of 10
Quality Score
Excellent
IOC Quality10
TTP Details9
Detection Guidance10
Enterprise Relevance10
Clarity & Structure10
Technical Depth8

Sources