The first ransom based on artificial intelligence was identified as promptlock

Security Company ESET From identifying the first ransom based on artificial intelligence Promptlock He has announced. According to the company, Promptlock can contaminate Windows, Mac and Linux systems using large language models and to ineffective conventional malware identification methods.

According to reports, the malware, called Promptlock, uses an open source language model developed by Openai to produce scripts that can perform various functions on Windows, Mac and Linux operating systems. According to ESET, this constant change makes security tools unable to easily identify the behavior of this ransom.

How to operate ransom based on artificial intelligence

Promptlock ransom using LUA scripts produced through predetermined perpetrators, list the victim’s system files, extracts selective data and can further encrypt them. ESET explained on Mastodon’s social network that the malware can stole information, encrypted or even destroy the information depending on the type of user files. However, the data demolition capability has not yet been fully implemented.

According to ESET explanations, Promptlock uses the Gpt-site: 20b Openai company locally and via the OLLAMA platform API to produce the instant production of malicious Lua scripts. Implementation of the model on the local device also makes Openai fail to track the activity of this ransom.

Choosing the Lua programming language for a ransomware at first glance can seem strange; Because it is more commonly known as the game development language in Roblox or NeoVIM editor extensions. But Lua, as an all -purpose language that supports a few platforms, has a considerable simplicity that has made it a suitable option for cyber invaders.

Promptlock seems to be a new example of artificial intelligence techniques in cyber attacks, indicating that hackers have also turned to the use of large language models. This could be a danger to security companies that must prepare themselves to counter a new generation of artificial malware.