Over 800 Artex AI penetration-testing deployments found exposed online
Security firm Censys reports hundreds of open-source Artex instances accessible without authentication, raising misuse risks despite tool’s shift to closed-source.
1 source · cross-referenced
- Security researchers identified more than 800 publicly accessible Artex deployments, an open-source AI penetration-testing tool.
- Censys attributes the exposure to misconfigurations, not required internet access by design.
- Artex has been linked to multiple cyberattacks, including incidents targeting Korean finance and Microsoft cloud customers.
- Artex’s maintainers announced a shift to closed-source in response to misuse concerns.
Security firm Censys reported identifying more than 800 distinct deployments of Artex, an open-source autonomous penetration-testing tool, that were accessible from the internet without authentication. The tool, which probes systems for vulnerabilities with minimal human oversight, was designed for local use and does not require internet exposure to function.
Principal security researcher Silas Cutler noted that while some deployments may be used by authorized testers, the public accessibility of hundreds of instances increases the risk of misuse. The exposure raises questions about why so many systems are configured for internet access despite not requiring it.
Artex has been linked to multiple cyberattacks, including a September incident targeting the Korean finance industry and an October attack on IT support provider ConnectWise. In the ConnectWise incident, Microsoft reported that attackers used Artex to infiltrate cloud environments and exfiltrate data rapidly, leaving defenders with little time to respond.
Many of the attacks associated with Artex exploited basic vulnerabilities, such as weak or known passwords and the absence of multi-factor authentication. Censys found that one exposed Artex instance, linked to a Chinese-speaking hacker, had been publicly accessible for months and contained attack data extending beyond South Korean banks to include payment and gaming companies.
In response to misuse concerns, Artex’s maintainers announced a shift to closed-source, restricting further public releases to paid customers. However, dozens of copies of the tool’s code remain available on code-sharing platforms, enabling continued use in attacks.
- Nov 11, 2026 · Hugging Face
Hugging Face proposes token pooling-aware training to improve compressed multi-vector retrieval
Trust79 - Nov 11, 2026 · TechCrunch — AI
Fractile raises $720M Series C to develop AI chips and compiler stack
Trust79 - Nov 11, 2026 · The Verge — AI
Google’s Tabletop AI experiment lets users build custom collaborative apps via waitlist
Trust79