LIVE

FBI Warns Hackers Are Using Autonomous AI Tool Artex to Breach Networks in Hours

🧠 AI Analysis 2 min read 👁 6

The FBI reports that threat actors are deploying the open-source AI penetration tool Artex to compromise corporate and government networks globally in under three hours.

FBI Warns Hackers Are Using Autonomous AI Tool Artex to Breach Networks in Hours🧠AI Analysis

Photo: TechCrunch AI

Key points

  • The FBI warns that autonomous AI tools like Artex are enabling rapid network breaches in under three hours.
  • High-profile targets include ConnectWise, South Korean financial institutions, and security firm Barracuda.
  • Despite the developer restricting the software, older open-source versions remain widely accessible online.

Law enforcement and cybersecurity agencies are raising alarms over the rapid adoption of autonomous AI hacking tools by malicious actors. According to a flash advisory issued by the FBI, open-source penetration testing software such as Artex is being leveraged to orchestrate sophisticated cyberattacks against organizations in the United States and internationally, impacting sectors ranging from financial services and defense to telecommunications and government.

Artex operates by linking open-source platforms or frontier large language models—including systems from major providers like OpenAI and Anthropic—to rapidly scan target infrastructure. The software hunts for vulnerabilities such as exposed credentials, weak security protocols, and absent multi-factor authentication. Once an initial foothold is secured, the tool helps attackers navigate corporate networks to harvest sensitive databases and communications. Although these automated intrusions often generate heavy network traffic, their execution speed routinely outpaces human defenders.

Documented incidents linked to the software include a major campaign affecting South Korean financial institutions, which compromised the personal data of at least 68,000 individuals. Furthermore, Microsoft investigations revealed that actors, suspected to be based in China, utilized Artex to breach IT management company ConnectWise, infiltrating the Microsoft 365 cloud environments of over 2,250 clients. Security firm Barracuda also reported a breach via an unrevoked third-party credential exploited during an Artex-driven operation.

Although the original creator restricted the software after its initial July release, older open-source versions continue to circulate widely. Recent security research has even uncovered hundreds of misconfigured, password-less Artex instances exposed online. What to watch next: How enterprise defenders adapt automated response systems to counter high-speed generative AI threats.

Source TechCrunch AI — techcrunch.com/2026/11/16/fbi-says-artex-ai-open-source-hacking-tool-us… ↗

Contact the newsroom

Got a tip, a proposal or a partnership question? Write to us and we will reply shortly.