Intel Name: Now you see me, now you don’t: using llms to obfuscate malicious javascript
Date of Scan: December 23, 2024
Impact: High
Summary: Recent advancements in the code understanding capabilities of LLMs have raised concerns about their misuse to generate novel malware. While LLMs struggle to create malware from scratch, criminals can leverage it to rewrite or obfuscate existing malware, complicating detection efforts. Traditional obfuscation tools are well-known to defenders, making their output easier to detect. However, prompts to LLMs can produce transformations that appear more natural, significantly increasing the challenge of identifying such malware.
