CybersecurityAI Research4 min reading time

Chinese Hacker Commands DeepSeek via Telegram to Launch Autonomous Attacks

The Hacker News
Read full post
Palo Alto Networks' Unit 42 uncovered a Chinese hacker using DeepSeek via the Hermes Agent framework to autonomously attack over 460 targets after initial Telegram commands, exploiting multiple CVEs but with limited confirmed successful breaches.

More in Cybersecurity

Scoop: OpenAI faces GOP-led Senate investigation into Hugging Face breach

Covered by 2 sources
Cybersecurity6 min read

Anthropic Discloses Fourth AI Hacking Incident Involving Claude Opus 4.6

Covered by 2 sources
Cybersecurity5 min read

Every AI Incident Has Two Timelines. We Default To One

Forbes