Analysis of AI-Generated Malware by APT36

Researchers analyzed AI-generated malware samples from the state-affiliated group APT36, labeling this category as "vibeware." The malware uses large language models (LLMs) to adapt basic attack logic into niche programming languages like Nim, Zig, and Crystal, while leveraging legitimate services for command-and-control (C2). Telemetry data shows a tenfold increase in such threats over six months. The trend highlights a shift toward scalable, automated attacks targeting organizations with weak security practices.