Artex AI hacking tool exposed in hundreds of public deployments
Over 800 deployments of the open-source Artex AI penetration testing tool were found publicly accessible without passwords, creating a major security risk.
Hundreds of copies of Artex, a Chinese-made open source AI hacking tool, were found publicly exposed to the web with no authentication required. The exposure required no authentication, allowing anyone with a web browser access to the full suite of Artex's penetration testing features, which is designed for enterprise security testing. The misconfiguration stemmed from improper deployment practices where organizations failed to implement basic access controls.
Scale of exposure
Censys identified over 800 distinct deployments of Artex accessible without a password. The AI-powered penetration testing platform was publicly released in July and is used for autonomous security testing. Its AI component automates testing, making it more powerful than traditional tools but also easier for less sophisticated attackers to exploit.
Malicious use cases
Artex has been linked to cyberattacks against the Korean finance industry in September. It was also used in an attack on ConnectWise in October, enabling Chinese-linked hackers to spy on thousands of organizations with data in Microsoft’s cloud. Microsoft stated that hackers used Artex to break into customers’ cloud environments and steal data rapidly. The Artex instance used in these attacks was publicly exposed for months and contained data beyond South Korean banks, including payment and gaming companies. CrowdStrike attributed attacks using Artex to a Chinese-speaking hacker who left an open directory hosting the tool. These attacks exploited simple vulnerabilities, such as weak passwords and lack of multi-factor authentication.
Security researcher perspective
Silas Cutler, principal security researcher at Censys, wrote the blog post identifying the exposed systems. He noted that not all Artex instances are operated maliciously as many are used for authorized testing. "Not every instance of Artex is operated maliciously and many are used for authorized testers," Cutler said.
AI security implications
The incident reveals how AI security tools' valuable automation features become dangerous when misconfigured, creating critical blind spots amid enterprise AI adoption. It represents a critical blind spot as enterprises race to adopt AI security tools amid massive venture capital investment. Penetration testing tools have deep access to internal networks, enabling attackers to map corporate infrastructures if exposed. This public exposure of Artex highlights the fundamental tension where powerful features become dangerous vulnerabilities.
The venture capital market reflects this focus, with recent large funding rounds and acquisitions aimed at new AI security risks. Dozens of copies of Artex’s code remain available on code-sharing sites and can still be used in active attacks. Artex announced it would go closed-source, limiting further releases to paid customers.
