All stories
Cybersecurity

FBI: Open-source AI tool Artex breaches networks within hours

FBI: Open-source AI tool Artex breaches networks within hours

The FBI has issued a security advisory warning that hackers are using Artex, an open-source AI-powered penetration testing tool, to carry out cyberattacks. According to the report, the campaigns include a widespread breach affecting customers of U.S. IT company ConnectWise and targets across South Korea's financial sector. The advisory says financial services, government, tech, defense, phone and internet companies, and law firms have been targeted in recent months.

Artex is one of several open-source AI platforms built to probe a network's defenses. Hackers can connect it to hosted frontier large language models from providers such as Anthropic and OpenAI. The article notes that the FBI describes Artex intrusions as potentially "noisy," generating enough traffic to trigger defensive alerts, but the tool lets attackers move at a speed human defenders struggle to match.

According to the report, Artex-linked attacks can move from targeting a system to active compromise in under three hours in some cases. The tool rapidly scans a victim's network for signs of weakness, then exploits simple security bugs such as weak protections and exposed or stolen passwords before spreading toward databases and mailboxes.

The FBI said Artex has been used in several attacks since its first public debut in July. The breaches at South Korean institutions exposed the personal data of at least 68,000 people. The FBI also confirmed Microsoft's findings that hackers used the tool to break into ConnectWise and reach the Microsoft 365 environments of at least 2,250 customers. After the attacks, the tool's developer made the software closed-source, but the FBI says the earlier open-source version continues to circulate and remains widely accessible.