Hundreds of Artex AI hacking tools found publicly exposed to the web

Security researchers reported finding hundreds of servers running Artex, an AI-powered penetration testing tool, exposed on the open internet, according to TechCrunch. The exposure reportedly allowed anyone who located one of these servers to access the hacking tool without authorization.

TechCrunch’s report, published November 6, characterizes Artex as a tool built for penetration testing, the practice of simulating attacks on computer systems to find weaknesses before malicious actors do. The researchers’ discovery concerned not the tool’s intended use but its unsecured deployment, with hundreds of instances left reachable by anyone on the web rather than restricted to authorized users.

The scale cited, hundreds of servers, indicates a pattern across multiple deployments rather than an isolated misconfiguration. TechCrunch’s account does not specify which organizations or individuals operated the exposed servers, nor does it detail what if any unauthorized access occurred before the exposure was found.

The report situates the finding within a broader concern about AI-powered security tools: software designed to assist in identifying vulnerabilities can itself become a vulnerability if improperly secured. A penetration testing tool left open to the public web inverts its intended function, offering the same capabilities to any uninvited visitor that it offers to the security teams meant to use it.

No further detail on remediation steps or affected parties was included in the available reporting.