Security researchers have located more than 800 instances of Artex, a Chinese-made, open-source autonomous penetration testing tool, that were accessible from the web without requiring a login. The exposed deployments raise the risk that outsiders can reuse the tool, harvest implants, or extract attack data from those systems.
Artex is an autonomous penetration-testing framework that automatically probes systems for vulnerabilities. Its code was released publicly in July and, as open-source software, anyone can download and run it on their own servers. That accessibility has made it attractive for operators who lack access to commercial AI models from providers such as Anthropic or OpenAI.
Microsoft has previously described Artex as a system of AI agents capable of conducting attacks "from reconnaissance through data exfiltration with little needed human interaction."
Censys analysis and key findings
Censys, a cybersecurity firm that scans the internet for exposed or vulnerable systems, published a blog post reporting that it found over 800 distinct Artex deployments reachable on the public web without a password. Silas Cutler, principal security researcher at Censys and author of the post, noted that not every Artex instance is used maliciously and that many are operated by authorized testers. Still, Censys said it is unclear why so many instances were internet-facing, as Artex itself does not require public exposure.
Publicly reachable instances also reveal attacker tooling and implants — hidden malicious software used to compromise and maintain persistent access in victim networks — potentially allowing others to repurpose those tools.
Incidents linked to Artex in September and October
Security firms have tied Artex use to several large-scale attacks. In September, CrowdStrike attributed attacks against South Korea's finance sector to a Chinese-speaking actor who used an Artex instance; investigators cited an "open directory" that hosted the system. Censys reported that the implicated Artex instance had been publicly exposed for months and appeared to contain attack data extending beyond South Korean banks to payment and gaming companies.
In early October, Artex was used in an attack on IT-support provider ConnectWise. That incident enabled Chinese-linked hackers to spy on thousands of organizations whose data and files were hosted in Microsoft's cloud. Microsoft said the attackers used Artex to break into customers' cloud environments and steal data at such speed that defenders had "little time to detect and respond."
Simple vulnerabilities and misconfigurations
Many attacks associated with Artex exploited straightforward weaknesses, including guessable or known passwords and a lack of multi-factor authentication. Censys's findings illustrate that basic security failures and misconfigurations can expose both targets and the attackers' own infrastructure.
Project status and remaining code availability
After public scrutiny and misuse, the Artex project announced it would move to a closed-source model, restricting future releases to paying customers. Nevertheless, dozens of copies of Artex's code remain available on code-sharing sites and can still be used in active attacks.
Why this matters
The discovery of hundreds of internet-exposed Artex instances shows how easily powerful AI-enabled cyber tools can be misconfigured or left accessible, increasing the chances of reuse and abuse. The incidents linked to Artex also underscore the continued importance of basic defensive measures — strong authentication, proper access controls, and avoiding unnecessary internet exposure of systems — to reduce both direct compromises and secondary misuse.
