Skip to content
NewA calmer, faster ZeroHour

Google says Chinese state groups are building an Artex-style tool

Original title: Chinese state group developing automated attack tool for cyber espionage, Google says

The Record ()4 min read5 · 3 reads
Part of a story covered by 15 sourcesGoogle says unattended AI agent built and ran exploitsMerged summary & timeline
AI summary · x-ai/grok-4.7Summaries by Vani AI

Google’s AI Threat Tracker also ties CAIRN to U.S. aerospace intrusions and a North Korean Claude Code backdoor.

Google’s AI Threat Tracker says Chinese state-backed and criminal actors use AI for reconnaissance, scanning, exploit research, and influence work, and that government groups are developing their own version of Artex for likely espionage. The FBI has warned that Artex, made by a Chinese researcher, can turn a large language model into a tool that automates intrusion stages through data exfiltration. Google says similar frameworks, including CAIRN, PentAGI, CyberStrikeAI, and HexStrike AI, were used in real attacks or private tests, including CAIRN against U.S. aerospace companies. The report also describes APT42 using Gemini, North Korean ClickFix activity that planted the SOMBERMEME backdoor via Claude Code, and faster patch-to-exploit work, citing PaperCut bugs.

Unconfirmed: Google’s claim that Chinese government groups are building their own Artex for espionage is the company’s assessment.

Key facts
  • Google says zero-days rose from eight a month in 2025 to 22 in August.
  • Researchers produced working PaperCut exploits within a day of disclosure using LLMs.
  • PentAGI was used for reconnaissance and scanning against organizations in Asia.
  • North Korean actors used AI for fake recruiter profiles and IT-worker resumes.
  • Google says the tools leave forensic traffic and recognizable code traces.
Security assessment · AI, from the article

Impact: Automates intrusions from reconnaissance through data exfiltration

Analyze this story with Vani AIImpact, who is affected and what defenders should do, in a new chat

From the original article

4 min read at therecord.media

Google released its latest AI Threat Tracker report on Monday, writing that threat actors are rapidly evolving their use of artificial intelligence in attacks — now leaning more heavily on autonomous multi-agent offensive frameworks like Artex, CAIRN, CyberStrikeAI and PentAGI to execute multi-stage intrusions at greater operational velocity.

The Record

therecord.media

Read the full article at The Record

A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.