The FBI has issued a warning about the increasing use of autonomous AI hacking tools, notably the open-source platform Artex, in cyberattacks targeting industries across the United States and globally. Recent attacks have affected a leading American IT company, ConnectWise, and several financial institutions in South Korea. These attacks are characterized by their speed and efficiency, often breaching security defenses in less than three hours.

According to the FBI's latest cybersecurity advisory, hackers have been employing Artex to target sectors such as finance, government, technology, defense, telecommunications, and law firms. Artex, which is designed for AI-powered penetration testing, can connect to various large language models from companies like Anthropic and OpenAI. While the tool generates significant data traffic that can alert companies to its presence, its rapid operation poses a challenge for human defenders.

The FBI highlights that Artex is adept at scanning networks for vulnerabilities, exploiting weak security measures, and using compromised passwords to infiltrate systems. This capability was demonstrated in attacks on South Korean financial firms, which resulted in the theft of personal data from at least 68,000 individuals. Similarly, the tool was used to breach ConnectWise and access Microsoft 365 environments of over 2,250 customers, with Microsoft attributing the attacks to Chinese hackers.

In another incident, cybersecurity firm Barracuda disclosed a breach involving Artex, where the tool exploited an exposed credential from a third-party service. This breach prompted Barracuda to update all customer credentials as a precaution.

Although Artex has been made closed-source following these attacks, its previous open-source version remains accessible and continues to be used in cyberattacks. Researchers recently discovered hundreds of publicly exposed Artex systems online, highlighting the ongoing risk of these powerful hacking tools.