ARTEX, PentAGI and Agent Zero consoles found open on the internet
Original title: AI Hacking Frameworks Like ARTEX, PentAGI, and Agent Zero Found Running on Internet-Accessible Servers
A November 7 Shodan snapshot matched 1,428 ARTEX, 419 PentAGI and 173 Agent Zero services; 39 Agent Zero consoles had no login.
GitGuardian reported on November 9 that dual-use AI penetration-testing frameworks are reachable from the internet, some without authentication. The team fingerprinted 41 of about 45 open-source projects and cataloged 32 frameworks. Public matches do not prove criminal use, and differing queries should not be added into a unique-server count. ARTEX’s creator removed the original project on October 8 after abuse reports, but 102 forks appeared that week.
Unconfirmed: GitGuardian says a fingerprint match alone does not prove criminal activity or a unique exposed server.
- Of 173 Agent Zero matches, 134 showed a login page and 39 were open.
- PentAGI had 419 matches, including 233 with the default certificate.
- Eleven of 41 assessed frameworks served a web interface by default.
- Only Pojun, Cybermes, Xalgorix and AIRecon shipped an agent-connected secret scanner.
- GitGuardian published detection descriptions, not a target list.
Impact: Open agent consoles can run commands and security tools
| Product | Affected |
|---|---|
| ARTEX | — |
| PentAGI | — |
| Agent Zero | — |
- 1.Use published fingerprints to find your own internet-facing deployments.
- 2.Validate query hits before treating a host as exposed or compromised.
- Organizations
- GitGuardian
From the original article
3 min read at cybersecuritynews.com
AI-powered hacking and penetration-testing frameworks are running on internet-accessible servers, with some management consoles left open without authentication. A November 7, 2026, Shodan snapshot recorded 1,428 services matching the ARTEX family, 419 matching PentAGI, and 173 matching Agent Zero, showing how widely these tools have spread. These platforms connect language models to security tools, allowing […]
By Tushar Subhra Dutta · cybersecuritynews.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.