Hackers are using an artificial intelligence tool to break into organizations’ computer systems, with some intrusions reaching privileged access in less than three hours, the FBI warns.
The November 13 bulletin identifies ARTEX, a security-testing framework that attackers have weaponized to automate operations from searching for weaknesses to stealing data. Privileged access gives intruders elevated permissions within a compromised system.
Hackers have targeted financial services, government, defense, energy and healthcare organizations worldwide since at least August, according to the bureau. The framework allows attackers with limited technical expertise to conduct complex operations against multiple targets simultaneously.
A security tool turned against businesses
ARTEX was intended for authorized security testing. Its automated agents can carry out computer tasks after receiving an overall objective, “requiring minimal human interaction beyond initial objective setting,” the FBI bulletin states.
The bureau cited attacks against South Korean financial institutions that resulted in the theft of personal information belonging to approximately 67,000 people.
The original developer removed the public code repository in October after reports of criminal misuse. Copies and modified versions remain available, according to the FBI.
Speed is a central concern. The attacks can generate enough suspicious activity to trigger security alerts, but leave defenders little time to respond before attackers achieve their objectives, the bureau said.
China-linked hackers targeted thousands of cloud environments
A group aligned with the Chinese state, tracked as Storm-0940, used ARTEX in attacks against hundreds of organizations, according to Microsoft research cited by the FBI.
A campaign involving compromised ConnectWise Cloud Backup applications led to attempted sign-ins across more than 2,250 customer cloud environments, Microsoft said.
That figure does not represent 2,250 confirmed data breaches. Microsoft said most targeting was unsuccessful, although the group established persistent, privileged access at numerous organizations. Even a successful sign-in does not by itself establish that data was stolen, the company cautioned.
The expanding use of AI in hacking follows a separate China-linked espionage campaign disclosed by Anthropic in 2025, as previously reported by The Dallas Express.
What the FBI recommends
Organizations should inventory internet-facing systems, including supplier portals, employee applications and older services that may receive less attention than their main websites, the bureau advised.
Its recommendations include restricting administrative systems to trusted connections, limiting application permissions, monitoring unusual access and data transfers, and preserving evidence when responding to an intrusion.
Organizations that identify suspicious activity should contact their local FBI cyber squad or submit a report to the Internet Crime Complaint Center.