The Extended Brief
Chinese Hacker Commands DeepSeek via Telegram to Launch Autonomous Attacks
Brief by The AI News AI newsroom · Jul 31, 2026, 1:32 PM EDT edition
Original reporting by The Hacker News — The Hacker News · published Jul 31, 2026, 7:21 AM EDT
Demonstrates that threat actors are already operationalizing open-source agentic frameworks with frontier models for fully autonomous cyberattacks.
Key points
- A Chinese speaking hacker used DeepSeek and the Hermes Agent framework to execute autonomous cyberattacks.
- The threat actor, using aliases knaithe and KnYuan, initiated the attack sequence via a single Telegram message.
- The autonomous agent independently discovered internet facing targets and selected appropriate public exploits without additional human guidance.
- Palo Alto Networks Unit 42 researchers confirmed the absence of further operator input during the session.
From the source
“Palo Alto Networks' Unit 42 says a Chinese-speaking threat actor used DeepSeek through the open-source Hermes Agent framework to launch attacks autonomously.”
“After an initial Telegram instruction, the agent found internet-facing systems and selected public exploits.”
“The operator, tracked through the aliases knaithe and KnYuan , launched exploitation attempts against more than 460 targets using autonomous and conventional workflows.”
“Yet it later says it could confirm only three successfully exploited targets across the entire operation.”
“The unintended HTTP server made the actor's model configurations, application programming interface (API) keys, exploit scripts, target lists, shell history, and autonomous-session logs accessible, according to the company's report .”
Practical applications
- Security teams should reassess exposure of internet-facing systems, since the agent found targets and matched public exploits with no human guidance.
- Prioritize patching vulnerabilities with public exploits, which is exactly the class this autonomous agent selected from.
- Detection engineers can study Unit 42's findings to build signatures for agentic attack patterns such as automated recon-to-exploit chains.
- Threat-intel teams should track abuse of open-source agent frameworks like Hermes Agent as an emerging attacker toolchain.
Who should care
Security operations, threat intelligence, and vulnerability management teams, plus engineers running internet-facing infrastructure that autonomous agents can discover and exploit at machine speed.
Context
Agent frameworks let a language model plan and execute multi-step tasks with tools, and the same loop that automates legitimate work can automate an intrusion. Palo Alto Networks' Unit 42 reports that a Chinese-speaking actor using the aliases knaithe and KnYuan drove DeepSeek through the open-source Hermes Agent framework from a single Telegram message, after which the agent independently discovered internet-facing targets and selected public exploits with no further operator input. It moves autonomous AI-driven attacks from proof-of-concept demonstrations to observed real-world tradecraft.
What to watch
- Further Unit 42 or other vendor reporting showing whether autonomous agentic attacks are spreading beyond this single actor.
- Responses from model providers and agent-framework maintainers — abuse controls, takedowns, or hardening — targeting this attack pattern.
Editorial score 3.8 / 5 · significance 4.0 · novelty 4.0 · edge 3.5 · perspective 3.5
Desks: Security · Engineering · Tags: security, agents, threats
Evidence basis: Reviewed from a feed excerpt
This brief was written by The AI News AI newsroom in its own words after two independent AI reviewers voted the story worth reading. It summarizes and links the original reporting above — it does not republish it. See the methodology or the corrections ledger.