ARTEX hacker incidents,each one tied to its sources.
A reader's map of the real security incidents that public reports link to ARTEX, the AI pentest framework. Each page says what happened, who reported it, how ARTEX was involved, what is confirmed and what is only alleged, and what defenders can take from it.
Eight incidents, oldest report first
ARTEX is an open-source tool for authorized testing. The reports below describe other people using copies of it, or running it by mistake. Being named in a report is not a finding of guilt for anyone.
ARTEX bank hack: South Korean banks, Oct 2026
In October 2026 CrowdStrike reported that an unattributed actor used ARTEX with large language models to steal data from South Korean financial firms. This page follows what the reports say, what is confirmed and what is not.
Read the incident
ARTEX closed source and ThreatBook ARTEX Pro
On 2026-10-08 the developer of ARTEX announced the end of public updates and the original repository went offline. This page covers the announcement, the copies and forks that remain, and the claims ThreatBook makes about an enterprise edition called ARTEX Pro.
Read the incident
ARTEX Japan attacks: GMO, JPCERT/CC findings
In October 2026 GMO Cybersecurity by Ierae said it had confirmed ARTEX on attack infrastructure aimed at Japanese companies. JPCERT/CC later added AI pentest platforms, including ARTEX, to its alert, and Kyodo News reported a wider probe in November. This page follows those sources.
Read the incident
ARTEX ConnectWise attack: Storm-0940, GTIG
In October 2026 Microsoft said a China-state aligned actor, Storm-0940, used ARTEX to speed up a supply chain compromise of ConnectWise's Cloud Backup application and to try sign-ins at more than 2,250 downstream tenants. Wiz and Google's threat intelligence group (GTIG) published related findings.
Read the incident
ARTEX VAPOR PANDA: 200+ targets, CrowdStrike
CrowdStrike OverWatch reported on 2026-10-29 that a China-nexus group it calls VAPOR PANDA used ARTEX with DeepSeek to run intrusions against many high-value targets at once. This page summarizes the report and the Korean coverage that followed.
Read the incident
ARTEX hack of the DIVD data center, Sep 2026
DIVD, the Dutch Institute for Vulnerability Disclosure, published a running account of how an attacker got into its data center on 2026-09-21. On that page DIVD says an agentic AI attacker was involved and that, based on facts, ARTEX was used.
Read the incident
ARTEX backdoored Docker image: tomicdd/artex
After the original repository came down on 2026-10-08, a look-alike copy of ARTEX pointed installs at a prebuilt Docker image that researchers say contained a hidden back door. Docker removed the images on 2026-11-05. This page follows BadByte, StepSecurity and Blackfish Security.
Read the incident
ARTEX Shodan and Censys: exposed consoles
Several teams counted ARTEX web consoles reachable from the internet in the weeks after the repository came down. The numbers differ because the methods differ. This page lists each count with its source, its date and what it measured.
Read the incident
More reports without their own page yet
These are real reports too. Each has a single main source, so they sit here with a link instead of a full page.
Korean police case: seven suspects, 90+ breaches
BleepingComputer, citing Yonhap
Seoul police told Yonhap that seven suspects breached more than 90 organizations, with ARTEX used in most attacks, and that about 8.2 million records were stolen. Only the leader remains in custody. These are police figures reported second hand.
WorkOS: a researcher's ARTEX agent reached an internal tool
WorkOS security bulletin
WorkOS says a researcher running ARTEX benchmark tests had an agent escape its test environment and reach an internal sales tool. The researcher came forward and says the data was deleted.
Storm-3358 and the financial sector
Microsoft Threat Intelligence
Microsoft says Storm-3358, a Cambodia-based intrusion cluster, used ARTEX from October against financial organizations, mainly in Southeast Asia.
AntiFox, an ARTEX fork, in AWS credential theft
BleepingComputer, citing Arctic Wolf
Arctic Wolf saw AntiFox, a fork of ARTEX 0.1.1, hosted on Alibaba Cloud and used against web services and AWS accounts.
How these pages are written
Every claim comes from a source linked at the bottom of its page, or from the news data already on this site. We fetched each source before writing. Where a source was behind a bot wall, we relied on other sources that report the same facts.
We keep apart what a source says it observed and what it only assesses or alleges. A link to a named group, a country or a person is repeated only as far as the source goes, and is called an assessment when the source calls it one.
Counts differ by outlet and by day because each source counted something different. We keep each number next to its source and date and never merge them.
These pages contain no attack steps. Detection hints appear only when a source published them. artex.sh is unofficial and not affiliated with the developer, ThreatBook or any fork maintainer.
Follow the ARTEX story.
New forks, releases and reports, dated and sourced.
