AI Is Going Just Great
← Timeline
·3d agoScaryMajor

AI Hacking Framework Artex Linked to Attacks on 200+ Organizations Globally

Published · curated by AI Is Going Just Great

Source: wired.com ↗ · archived copy ↗

"I would say my grandmother can do it, except that she doesn't have a laptop, but the bar is that low."

Artex, an open-source AI attack framework published on GitHub in July by a pseudonymous developer, has been used by cybercriminals to breach financial institutions, government agencies, defense contractors, and technology firms across multiple countries. The tool is not itself an AI model. It functions as a guide that connects external LLMs via API to existing penetration testing tools, then orchestrates multiple agents to carry out attack steps autonomously based on a plain-language prompt. Attacks that previously took days are completing in hours. Dosu, an agentic coding platform, logged 80,000 requests to its systems in 90 minutes from a single Artex-powered campaign.

In October, China-linked group Vapor Panda used Artex to target more than 200 organizations, including US semiconductor and defense firms and European government ministries. Microsoft tracked a separate Chinese group, Storm-0940, using it to compromise Microsoft 365 environments after a supply chain attack, gaining privileged access in under three hours. Security firm GitGuardian has documented 45 distinct AI attack frameworks in the same vein. The Artex developer pulled the repository from GitHub on October 8, citing educational intent; multiple forks remain active. Censys and the Shadowserver Foundation found thousands of Artex instances exposed to the open internet, some unsecured enough for researchers to inspect live operator sessions.