The FBI has warned that attackers have been exploiting ARTEX, a tool developed for security testing, to automate attacks against organizations and companies worldwide.
ARTEX is open-source software used for penetration testing—an authorized and simulated cyberattack on a network, computer system, or application for security-testing purposes—according to a description on a GitHub page and the FBI.
GitHub is a cloud-based developer platform where software can be created and worked on.
In a bulletin released to the public on Nov. 13, the law enforcement bureau said that in some reported attacks, cybercriminals have weaponized ARTEX to “automate multiple phases of cyber intrusions at machine speed, from reconnaissance through data exfiltration” since at least August.
“A variety of malicious cyber actors have used ARTEX to conduct campaigns targeting organizations globally across multiple sectors, including financial services, government, technology, defense, semiconductor, energy, healthcare, manufacturing, telecommunications, and legal industries,” the FBI warned.
Because ARTEX can “significantly” lower the time required to carry out chain attack techniques, some intrusions have progressed “from initial targeting to privileged access in under three hours,” the agency said.
That can allow cybercriminals and other threat actors with little technical knowledge to carry out “complex, multi-target operations simultaneously” in a way that accelerates and expands operational scale.
In a campaign of ARTEX-related breaches, an estimated 67,000 people saw their personal information stolen via South Korean financial institutions, the FBI said.
The FBI noted that the original GitHub page for ARTEX was removed, but community forks of the software have continued to be uploaded and updated, with localized variations, archived copies, and a commercial version.
In November, researchers were able to identify more than 1,600 ARTEX publicly exposed consoles “with total deployments likely significantly higher when including systems behind firewalls or in private networks.”
With that in mind, the FBI said that organizations should implement recommended fixes to reduce the impact of possible ARTEX intrusions.
Groups that operate internet-exposed services, portals, cloud infrastructure, and partner-supply systems were specifically flagged by the bureau.
The original variant of ARTEX was released in July 2026 on GitHub before it was rapidly adopted, with the developer writing that it was intended for authorized security testing of systems. The repository was made unavailable in October, according to the FBI.
ARTEX is not a standalone large language model (LLM)—an artificial intelligence system developed on exceptionally wide-ranging data—but connects to external LLM providers such as OpenAI GPT, Claude, and DeepSeek to help organizations test for vulnerabilities in their networks.
“Given the misuse of the tool, the ARTEX project will no longer be updated and will be converted to closed source. No further versions will be released to the public, nor will maintenance support be provided,” the developer, who uses the GitHub handle “Autumn-27,” said on the code-hosting platform last month.
U.S. cybersecurity firm CrowdStrike said in October that the suspect behind the cyberattacks at South Korean banks aimed at stealing customers’ personal data was likely a China-based 26-year-old who used the ARTEX AI agent and Anthropic’s Claude Code. CrowdStrike also said that ARTEX was also “developed in China.”
“While this activity has not been attributed to a named adversary, the threat actor is likely a Chinese speaker and financially motivated. This assessment is made with moderate confidence based on the use of the Chinese-developed tool ARTEX and observed Chinese-language prompts,” the company wrote.
Chinese foreign ministry spokesperson Mao Ning told a regular press briefing in October that the ministry was not familiar with the case.
Reuters contributed to this report.

