AI-Powered Malware Leverages Multiple Models for Autonomous Attacks

A new Windows malware called ClosedQuorum has been discovered by Cisco Talos researchers. This malware uses artificial intelligence (AI) to make decisions during post-compromise stages of an attack, acting autonomously with no human commands.

ClosedQuorum uses four AI models - Google Gemini, DeepSeek, Qwen, and Mistral - to decide its next steps on infected hosts. When votes are tied, DeepSeek has priority in making the final call, followed by Qwen, Mistral, and Gemini.

The malware's actions are restricted to a predefined set of decisions, including stealing credentials, injecting shellcode, persisting on the system, and moving laterally. The stolen details are then passed to the operators via a Discord webhook, allowing for fully automated attacks.

Cisco Talos describes ClosedQuorum as the first publicly documented Windows implant to delegate tactical command-and-control decisions to AI models, adding speed and scaling potential to malicious operations. However, this may also pose challenges in certain cases where rate limits are hit or commercial APIs are temporarily unavailable.