The FBI is warning that threat actors have weaponized Artex, an open-source autonomous hacking framework driven by large language models, to target organizations in healthcare and nine other industries.
In a Nov. 13 FLASH alert, the FBI said a variety of malicious actors have used Artex since at least August to automate cyber intrusions from reconnaissance through data exfiltration. Other targeted sectors include financial services, government, technology, defense, energy and telecommunications.
A developer known as “Autumn-27” released Artex on GitHub in July. The platform converts natural-language objectives into commands carried out by AI agents with minimal human involvement. It connects to outside AI models, including Anthropic’s Claude, OpenAI’s GPT and Chinese-origin models such as DeepSeek.
Some intrusions moved from initial targeting to privileged access in under three hours, the FBI said. The tool lets attackers with little technical expertise run multiple campaigns at once, “fundamentally accelerating attack velocity and expanding operational scale,” according to the alert.
The FBI identified peripheral systems as a key weak point. These include partner and supplier portals, mobile backends, employee apps, legacy services and third-party integrations. In one campaign, a supply chain compromise of ConnectWise Cloud Backup customers gave attackers access to more than 2,250 downstream tenants.
The developer pulled the GitHub repository Oct. 8, but Artex continues to circulate through forks, archived copies and a commercial version. Researchers found more than 1,600 publicly exposed Artex consoles as of early November.
Corrected: The FBI urged organizations to inventory internet-facing applications, restrict administrative interfaces to trusted networks and rate-limit high-impact APIs. It also flagged unexpected outbound calls to AI model providers as a sign of possible Artex activity.