
Machine-Speed Breaches: FBI Warns Hackers Use Artex AI Tool to Compromise Networks in Under Three Hours
Federal law enforcement officials are sounding the alarm over cybercriminals weaponizing autonomous artificial intelligence penetration testing tools to infiltrate corporate and government networks worldwide. In a flash cybersecurity advisory, the Federal Bureau of Investigation warned that attackers are using an open-source framework called Artex to carry out automated intrusions across financial services, government agencies, technology vendors, defense contractors, telecom providers, and law firms.
Artex belongs to a growing category of AI-powered penetration testing platforms built to probe network defenses automatically. Attackers can connect the software to hosted frontier large language models from providers like Anthropic and OpenAI, as well as open-source and non-U.S. AI models. While the FBI noted that Artex intrusions generate heavy network traffic that can trigger security alerts, the software executes commands at machine speed, making it extremely difficult for human security teams to respond in time. In several documented incidents, Artex-driven attacks moved from initial network targeting to active system compromise in under three hours.
Instead of developing complex zero-day exploits from scratch, Artex succeeds by rapidly scanning external networks for basic security weaknesses. The automated agents look for unpatched software bugs, weak access controls, and exposed or stolen passwords to gain an initial foothold. Once inside, the software spreads laterally across internal infrastructure to harvest emails, database records, and sensitive corporate files.
Since Artex first debuted publicly in July, attackers have deployed it in multiple high-profile campaigns. The FBI highlighted a wave of intrusions targeting South Korean financial institutions, where weak perimeter defenses and missing multi-factor authentication allowed hackers to steal personal data belonging to at least 58,000 individuals. Federal investigators also confirmed earlier findings from Microsoft showing that suspected Chinese threat actors used Artex to breach IT management firm ConnectWise, gaining unauthorized access to Microsoft 365 cloud environments across at least 2,250 downstream customers. ConnectWise did not respond to requests for comment regarding the federal advisory.
Cybersecurity firm Barracuda also confirmed it suffered an Artex-linked breach carried out by financially motivated criminals. Attackers discovered an exposed credential belonging to a third-party vendor that Barracuda had stopped working with but failed to revoke. Using that leftover access, the intruders reached internal credential stores Barracuda used to collect telemetry from customer systems. Because internal audit logs lacked sufficient detail to rule out customer exposure, Barracuda chose to assume the worst and rotated affected customer credentials. Barracuda Chief Information Security Officer Arve Kjoelen warned that security teams relying on reactive alert monitoring often arrive after automated attackers have already finished stealing data. A company spokesperson added that Barracuda notified a small group of XDR customers directly to recreate their integration keys.
Following widespread abuse of the software, Artex’s creator converted the project into a closed-source tool to curb malicious attacks. However, the FBI warned that archived copies and forks of the original open-source code remain widely available online. Even more concerning, sloppy operational security among hackers has exposed the attack infrastructure itself. Security researchers recently discovered hundreds of publicly accessible Artex consoles sitting online without password protection, including the exact server used to attack South Korean banks, allowing outsiders to view or seize powerful offensive hacking tools.







