AI-Powered Hacking Threat: FBI Warns Cybercriminals Exploiting ARTEX Framework To Target Organisations

The agency said the tool has been used in campaigns across multiple sectors worldwide since at least August 2026.

Advertisement
Read Time: 3 mins
Representative Image
Unsplash

The US Federal Bureau of Investigation has warned that malicious cyber actors are using ARTEX, an open-source, AI-driven penetration testing framework, to automate cyber intrusions at machine speed.

In a FLASH alert dated 13 November, the agency said the tool has been used in campaigns across multiple sectors worldwide since at least August 2026.

Advertisement

Intrusions In Under Three Hours

ARTEX is a large language model-driven platform that turns plain-language objectives into machine commands run by autonomous agents, requiring minimal human input. The FBI said some intrusions went from initial targeting to privileged access in under three hours.

The agency warned that the framework lets actors with little technical skill run complex, multi-target operations at the same time. It said ARTEX is particularly effective at exploiting overlooked systems such as partner and supplier portals, mobile backends and legacy services.

Advertisement

Targets Span Key Sectors

Targeted sectors include financial services, government, technology, defence, semiconductors, energy, healthcare, manufacturing, telecommunications and legal services.

The FBI cited attacks on South Korean financial institutions that led to the theft of personal data of about 67,000 individuals. It also cited a supply chain compromise affecting ConnectWise Cloud Backup customers, which resulted in attempted sign-ins to more than 2,250 downstream tenants.

ALSO READ | FBI Points Finger At North Korea For Record $6 Billion Binance Crypto Heist

Frontier And Chinese Models Used

ARTEX has no AI model of its own. It connects to external providers, including Anthropic's Claude and OpenAI's GPT, as well as DeepSeek, GLM, Grok and Kimi.

Advertisement

According to the FBI, analysis of attacker-controlled instances shows Chinese models are the most widely used. DeepSeek's Flash family dominates because of its capability, low cost and lack of observable guardrails when requests are framed as penetration testing.

Takedown Has Not Stopped Spread

The tool was released on GitHub in July by a developer using the name "Autumn-27". The developer pulled the repository on 8 October after reports of abuse, saying it was meant for authorised security testing.

The framework still circulates through forks, archived copies and a commercial version called ARTEX PRO. Researchers identified over 1,600 publicly exposed ARTEX console instances in early November, with actual deployments likely far higher.

ALSO READ | Hackers Turn To Chinese AI Agent To Target South Korea's Biggest Banks, Raising Security Concerns

What Organisations Should Do

The FBI urged organisations to inventory all internet-facing applications, including peripheral systems, and to restrict cloud administration interfaces to trusted networks. It also advised applying least privilege to application permissions, putting rate limits on business APIs and scanning for exposed secrets.

Advertisement

Organisations hit by an ARTEX-linked intrusion were asked to report it to their local FBI field office or the Internet Crime Complaint Center. The alert was published with the TLP:CLEAR marking, meaning it can be shared without restriction.

Essential Business Intelligence, Sharp Market Insights, Practical Personal Finance Advice, Daily Fuel, Gold and Silver Prices and Latest Stories — On NDTV Profit.


Loading...