The cybersecurity world is facing a long-awaited new type of attack. Expert warnings have come true, and the first ransomware using artificial intelligence has emerged. According to reports, a cybersecurity company has identified this dangerous software.

Artificial Intelligence Ransomware Has Arrived

This ransomware, called PromptLock, uses a modified version of the open-source large language model developed by OpenAI. The software runs on Windows, macOS, and Linux systems.

The AI ​​generates Lua scripts on these systems, which allow it to find and lock target files. According to an ESET report, the software not only locks files but can also export or completely delete some data.

The key feature that makes this new type of attack dangerous is its ability to evade security software. Because large language models follow a different path for each command, the patterns detected by existing cybersecurity programs are not evident.

This makes PromptLock difficult for antivirus software to detect. Furthermore, because the ransomware runs AI locally on the device it infiltrated, OpenAI has no chance of detecting these suspicious activities.

ESET’s findings signal the dawn of a new era in cybersecurity. This means a new “arms war” has begun between cybercriminals and security experts. Both cybersecurity firms and AI developers are being forced to develop defenses against these new threats.