FBI: AI hacking tool Artex cracks networks in under three hours
The FBI says hackers are using Artex, an open-source AI penetration testing tool, to smash into networks at machine speed, hitting victims in under three hours. Its flash advisory ties the tool to a South Korean breach affecting at least 68,000 people, a ConnectWise hack touching 2,250 customers, and a Barracuda intrusion.
- 2 sources report it: techcrunch.com, crypto.news.
- The source published it on Nov 16, 2026 at 5:36 PM.
Nothing open. Every check passed.
The FBI dropped a flash cybersecurity advisory warning that hackers are leaning on autonomous AI hacking tools to break into networks in the U.S. and abroad. The list of targets is broad: financial services, government, tech, defense, telecom and internet providers, and law firms. The tool at the center of it is Artex, an open-source AI-powered penetration testing platform that hackers can wire up to hosted frontier models from Anthropic and OpenAI, as well as open-source and non-U.S. models.
Speed is the whole problem. Artex intrusions can be noisy, throwing off enough traffic and data to trip a victim's defensive alerts, but the tool moves faster than human defenders can react. In some cases, attacks built on it go from targeting a system to full compromise in under three hours. The playbook is unglamorous: scan the victim's network fast, spot weak protections or exposed and stolen passwords, slip in, then spread toward databases and mailboxes.
South Korea's financial sector took a brutal hit. Artex was used against a number of the country's financial institutions, an outcome the FBI largely blames on weak cybersecurity and missing protections like multi-factor authentication. Hackers walked off with the personal data of at least 68,000 people.
ConnectWise got hit too. The FBI confirmed Microsoft's earlier finding that hackers used Artex to break into the IT management giant and reach the Microsoft 365 cloud environments of at least 2,250 customers, stealing data from those affected. Microsoft says the intruders were likely Chinese. A ConnectWise spokesperson did not respond to a request for comment on the FBI's claim.
Barracuda also disclosed a breach tied to an Artex-linked attack, this one by financially motivated criminals. According to Barracuda, the tool found an exposed credential from a third-party service the company hadn't revoked after ending work with the vendor, giving hackers a path to the credential stores Barracuda uses to pull data from customer systems. Barracuda said its logs weren't detailed enough to rule out access, so it chose to assume the worst and rotated all customer credentials. The company didn't answer questions about how many customers were affected.
Barracuda's chief information security officer, Arve Kjoelen, framed the stakes bluntly: defenders who wait to spot malicious activity and then respond will often arrive after the attacker has finished. Artex made its public debut in July, and its developer switched it to closed-source once the attacks landed. The FBI says the old open-source version is still circulating and remains widely accessible. Last week researchers found hundreds of Artex systems exposed online without a password, including the one used against South Korea.
TopicsFBI · Artex · ConnectWise · Microsoft · Barracuda · Anthropic · OpenAI · Arve Kjoelen
Written by AI from verified reporting.

