A new artificial intelligence (AI) tool meant to help companies spot and fix security flaws has been turned into a weapon by cybercriminals. The tool, called Hexstrike-AI, was created to let security teams think like hackers and find vulnerabilities before bad actors do. But its creators may not have fully seen how it could be misused, leading to serious security concerns.

The Original Goal of Hexstrike-AI

Hexstrike-AI was built as an AI “brain” that manages a network of over 150 specialized AI agents and security tools. Its job was to test a company’s defenses, find weaknesses like zero-day vulnerabilities, and report back with details. The idea was to give security teams a powerful way to proactively defend their systems.

What made Hexstrike-AI so effective was its ability to automate complex security tasks. It could quickly identify weak spots that would normally take days or weeks for human teams to find. However, this same capability also made it attractive to hackers looking for a way to automate their attacks.

The Dangerous Side of Automation

The timing of Hexstrike-AI’s release was unfortunate, as it coincided with the discovery of three major zero-day flaws in Citrix’s NetScaler products. These flaws were so new that no patches had been made yet. Traditionally, exploiting such vulnerabilities requires skilled hackers and lots of time. But with Hexstrike-AI, attackers can do it in under 10 minutes.

The AI system automates the hacking process, choosing the best tools and steps needed to exploit the vulnerabilities. Cybercriminals are already using Hexstrike-AI to target systems at a rapid pace. They can simply instruct the tool to exploit a specific vulnerability, and it takes care of the rest, making hacking feel almost effortless.