Security researchers have found that hundreds of copies of AI-powered hacking tools made in China were publicly exposed on the internet, accessible to anyone, potentially enabling outsiders to launch cyberattacks on unsuspecting targets.
The hacking tool named Artex is an open-source autonomous penetration testing software capable of automatically detecting vulnerabilities in computer systems, and has been linked to multiple cyberattacks since its public release in July this year. As open-source software, it is freely available for anyone to download and run on their own servers, making it a preferred tool for those unable to access the most advanced cybersecurity AI models from Anthropic or OpenAI to identify security flaws. Microsoft previously described Artex as an AI agent system capable of launching cyberattacks “from reconnaissance to data exfiltration” with minimal human intervention.
Cybersecurity firm Censys stated in a blog post that, during scans of internet-exposed or vulnerable systems, it identified over 800 distinct Artex deployments accessible via web browsers without requiring a password.
Silas Cutler, Chief Security Researcher at Censys, stated that not every Artex instance is being used maliciously—many are intended for authorized testers. However, it remains unclear why hundreds of Artex systems are exposed on the internet, as the software does not require them to be accessible from the internet.
These findings highlight that even powerful hacking tools can become publicly visible due to misconfiguration, increasing the risk of discovery and misuse by others. Publicly accessible systems can also expose attack tools, such as implants—hidden malware used to infiltrate and maintain persistent access to victims' networks.
A security firm has linked multiple large-scale cyberattacks to the use of Artex, including an attack in September targeting South Korea’s financial sector. Cybersecurity giant CrowdStrike attributed these attacks to a Chinese-speaking hacker, noting that the attacker used an Artex instance, partly because the hacker left an “open directory” hosting the system. Censys stated that its data shows the implicated Artex instance had been publicly exposed for months and appeared to contain additional attack data extending “far beyond South Korean banks,” including payment and gaming companies.
Artex was also used in an attack against IT support giant ConnectWise in early October this year, enabling hackers linked to China to monitor thousands of organizations whose data and files are hosted on Microsoft’s cloud. Microsoft had visibility into the attack and stated in a blog post that the hackers used Artex to infiltrate customers’ cloud environments and steal data so quickly that defenders had “little time to detect and respond.”
Many attacks related to Artex exploit simple vulnerabilities, such as guessing weak or known passwords and the lack of multi-factor authentication.
These findings indicate that simple vulnerabilities, including misconfigurations, can simultaneously expose both attackers and their targets.
Artex subsequently announced it would become proprietary, effectively limiting further distribution to paying customers. However, dozens of copies of the Artex code remain available on code-sharing websites and can still be used in ongoing attacks.
