Windows malware uses AI to autonomously select post-compromise targets

A Windows implant named CLOSEDQUORUM uses language models to autonomously decide its actions post-compromise.

Published 2j2 sources✓ ConfirmedImportantupdated 1j

≈ 19sSpeed

The fact

It is the first publicly documented malware to integrate LLMs for command and control.

This method could make attacks harder to detect and counter.

2 sources — click a link to read an article on the topic:

Explore this topic

Auto-synthesis from 2 media sources · identified on September 22, 2026