More than 1,400 consoles running ARTEX, an AI-powered hacking framework already linked to attacks on banks and a cloud backup supplier, are sitting exposed on the open internet. Internet-scanning firm Censys counted 1,465 hosts running an exposed ARTEX console as of 4 November 2026, and published its findings on 5 November.

ARTEX is an open-source, LLM-driven autonomous penetration testing framework, released on GitHub on 26 July 2026. It chains AI agents together to run attacks from reconnaissance through data theft with minimal human input. About 650 of the exposed hosts sit in two blocks of identically configured hosting, which leaves roughly 815 distinct-looking deployments. Censys notes the console does not need to be public at all, so the true total is likely higher.

Censys also found 32 open directories on ARTEX hosts leaking post-exploitation kits, stagers and implants, including a Windows lateral-movement kit, a public sudo exploit and, on one host, an operator's entire home directory with `.ssh` keys and shell history. Another 155 hosts, 10.6 per cent of the total, also run at least one other threat Censys detects, from Cobalt Strike to other AI attack frameworks.

Already tied to two major campaigns

CrowdStrike reported on 7 October that an unattributed actor used ARTEX against South Korean financial institutions, breaching a bank's loan inquiry service and an employee mobile system. Microsoft attributed the ConnectWise Cloud Backup supply-chain compromise to Storm-0940, a China state-aligned group, which used ARTEX to automate scanning, password spraying and OAuth abuse across more than 2,250 tenants, peaking at 2,900 sign-in attempts per minute.

Not every exposed console is malicious — authorised testers use the tool too — but an exposed console plus an open directory hands attackers' tools to anyone who looks. The most common console port is 8787, ARTEX's default.