In an advisory, the FBI said some attacks took less than 3 hours from selecting a target to gaining high-level access to its systems.[1]

Artex's developer removed the original GitHub repository in October after reports of abuse, but copies are still available, the FBI said, so pulling the code did not take the capability back.[1] The automated attacks could set off security alerts, CNET reported, but they moved fast enough to leave defenders little time to respond.[1]

An operator gives Artex a target and an objective, and a planning agent splits the work into tasks that other AI agents carry out with security tools, according to Malwarebytes.[2] The tool has no AI model of its own; it connects to external large language models and needs relatively little human input once a task begins.[2]

CNET reported that Microsoft warns attackers breached over 2,250 organizations' Microsoft 365 systems in a separate campaign using stolen ConnectWise backup credentials alongside the Artex tool, though Microsoft cautioned that a successful sign-in doesn't necessarily mean data was accessed or stolen.[1] Amazon Threat Intelligence linked a credential-theft scheme to Artex with moderate confidence but could not verify results the agents reported on systems outside its monitoring, according to CNET.[1]

An FBI official told ABC News that Artex does not appear to be connected to a major Microsoft 365 outage or to other recent disruptions, including banking outages involving JPMorgan Chase, CNET reported.[1] Palmer Luckey, the Anduril Industries co-founder, said "the advantages are going to accrue to the side that is most effectively using AI for defense," TechCrunch reported.[3]