FBI Warns AI Hacking Tool ARTEX is Powering Global Cyberattacks
· GlobeWire Daily Editorial Desk · Source: www.bleepingcomputer.com

Image: BleepingComputer · Photo source · Resized and letterboxed to 16:9; no AI generation
The Federal Bureau of Investigation (FBI) has issued a warning regarding an artificial intelligence-powered penetration testing framework known as ARTEX, which is reportedly being used by threat actors to automate cyberattacks globally. According to a FLASH advisory released by the bureau, multiple groups have weaponized the open-source tool since August 2026 to target sectors including finance, government, technology, defense, and healthcare.
Understanding the ARTEX Framework
ARTEX is described as an autonomous penetration testing framework written in Go. Rather than utilizing an internal proprietary AI model, it integrates with external large language models (LLMs) such as Anthropic Claude, OpenAI GPT, DeepSeek, GLM, Grok, and Kimi. Operators supply goals and tasks that specialized AI agents translate into executed commands. This setup allows individuals with minimal technical expertise to perform reconnaissance, discover vulnerabilities, exploit systems, escalate privileges, and exfiltrate data simultaneously across multiple targets with limited human intervention.
The FBI advisory notes that the framework significantly reduces the timeline for chaining attack techniques, with certain intrusions progressing from initial targeting to privileged access in under three hours. The tool is reportedly effective at finding vulnerabilities in overlooked systems, including legacy services, partner portals, and third-party integrations.
Documented Incidents and Attribution
The advisory highlights several campaigns linked to the framework:
- ConnectWise Supply-Chain Breach: In October 2026, a campaign attributed by Microsoft to a China-state-aligned group tracked as Storm-0940 compromised ConnectWise cloud backup service application identities, allowing attackers to attempt authentication against thousands of downstream Microsoft 365 customer tenants.
- South Korean Financial Institutions: Threat actors combined ARTEX with Claude Code agents to target banks such as Shinhan Bank, KB Kookmin Bank, and Hana Bank, exposing personal data belonging to approximately 67,000 individuals. Separately, South Korean police arrested seven suspects linked to attacks using ARTEX against more than 90 organizations.
- Vapor Panda Cyberespionage: CrowdStrike reported that a Chinese group tracked as Vapor Panda used ARTEX to target more than 200 organizations across 14 countries, successfully compromising at least nine.
Mitigation and Verification
The original developer removed the GitHub repository for ARTEX in October 2026, but the framework remains accessible via community forks, archived copies, localized variants, and a commercial edition named ARTEX PRO. Cybersecurity researcher MalwareMustDie released an open-source detection signature pack to help defenders identify related activity.
Organizations are advised to secure internet-facing applications, restrict access to cloud administration interfaces, review authentication logs, and implement rate limits on sensitive APIs. The FBI encourages entities identifying suspicious activity to report findings to the bureau or the Internet Crime Complaint Center (IC3). Indicators of compromise, including IP addresses, should be cross-referenced carefully as infrastructure details may change over time.
AI-assisted reporting; source claims are attributed and have not been independently verified.



