Hundreds of Artex AI Pen-Testing Deployments Exposed Online Without Password Protection
Hundreds of Artex AI hacking tools found publicly exposed to the web
Cybersecurity firm Censys discovered that more than 800 deployment instances of Artex, an open-source autonomous AI penetration testing tool, are exposed on the public internet without password protection, leaving them accessible to anyone. Microsoft has characterized Artex as an AI agent system capable of autonomously executing entire kill chains from reconnaissance to data exfiltration; the tool has already been linked to several cyberattacks targeting South Korea's financial sector and ConnectWise. Although Artex previously announced a shift to a closed-source model restricted to paying customers, copies of its code continue to circulate across online platforms.
The report highlights the risks of publicly exposed, misconfigured open-source autonomous penetration testing tools, providing defenders with reference points to identify unauthorized, AI-automated attack vectors.
Source:TechCrunch AI · techcrunch.com