- FBI confirms hackers used open-source AI pentesting tools like Artex connected to frontier models from Anthropic and OpenAI to launch autonomous attacks
- Targets span finance, government, tech, defense, telecom and law firms, including ConnectWise customers and South Korea's financial sector
- AI-driven attacks move faster and noisier than human defenders can handle, and some Artex systems were exposed online without passwords
AI CURATED HOT STORY · 11/17/2026
FBI Warns Open-Source AI Hacking Tool Artex Breaches Networks in Three Hours
The FBI has issued an urgent cybersecurity advisory warning that hackers are using Artex, an open-source AI penetration testing platform, to connect with frontier large language models and launch autonomous cyberattacks across multiple sectors, with some breaches completed within three hours. This marks the first official confirmation that AI-driven attacks have reached operational, automated capability, signaling a new era in cybersecurity threats.
According to multiple sources, targets include finance, government, technology, defense, telecommunications, and law firms, with U.S. ConnectWise customers and South Korea's financial sector already affected. Artex integrates with models from Anthropic and OpenAI to automate attack workflows, enabling high-speed, high-volume operations that outpace human defenders. However, the tool generates significant traffic noise that can trigger alerts, and some Artex systems have been exposed online due to misconfiguration, creating a double-edged sword.
The warning highlights the dual-use risks of AI: open-source penetration tools lower the barrier to entry, while frontier models are being misused for automated attacks. Defenders must accelerate AI-driven security responses. The incident may prompt regulators to tighten scrutiny of open-source AI tools and push enterprises to reassess AI supply chain security.
Hourly heat curve
4 fully observed hoursThe line compares only sources observed throughout; gap hours are interpolated to keep the trend continuous.
Coverage timeline
Total 2 reports · Latest first- Hacker News AIT2·78 pts
- TechCrunch AIT2·78 pts
- The FBI confirmed attackers are using open-source AI pentesting platforms like Artex connected to frontier models from Anthropic and OpenAI
- Targets span finance, government, tech, defense, telecom and law firms, with ConnectWise customers and South Korea's financial sector affected
- Some intrusions completed in under three hours, outpacing human defenders, though the noisy traffic is more likely to trigger alerts