Mark RadarMARK RADAR
About
EN
Sign in
Event File AI Agentic AI

China-Linked Hackers Deploy AI Agents in Taiwan Cyber Campaign

1 reports · First detected 2026-08-12 · Last active 2026-08-12

Taiwan’s semiconductor, manufacturing and supply-chain sectors have long been targets of China-linked cyber espionage. The latest campaign marks a shift from using generative AI as a coding assistant to embedding it in live attack workflows. Claude Code handled command execution, persistent sessions and parallel tasks, while DeepSeek-v4-pro supplied attack reasoning, adapted exploits and helped build phishing pages, potentially allowing intrusions to advance faster than conventional human-led operations.

Hunt.io disclosed the campaign on July 14, 2026, after tracing TencShell infrastructure in June to 13 Hong Kong-based servers across four autonomous systems. The network scanned more than 5,890 government hosts in 10 countries. In Taiwan, eight supply-chain and manufacturing organizations were reconnoitered and two had been compromised as recently as June 11, including a chemical company and a telecom and edge-device manufacturer. An exposed directory held 2,431 files across 80 subdirectories; no financial loss was disclosed.

All Coverage

1 original reports

The Backstory

The history behind this event
June 2 Cybersecurity Report: Chinese Cyberespionage and AI Misuse Threats2026-06-02 · 1 reports · similarity 0.80

Chinese hackers have long targeted governments and diplomatic networks to gather intelligence. The latest “Dragon Weave” cyberespionage campaign involved both Taiwan and the Czech Republic and took place shortly before a Czech Senate delegation visited Taiwan, underscoring the close link between geopolitical activity and state-level cybersecurity risks. The rapid adoption of AI tools is also making traditional account fraud and system infiltration easier to conduct at scale.

The June 2 cybersecurity report said Chinese hackers launched the “Dragon Weave” espionage campaign against Taiwan and the Czech Republic ahead of the Czech Senate delegation’s visit to Taiwan. Other key developments included the abuse of Meta AI customer support that led to user accounts being stolen, as well as hackers using AI agents to infiltrate databases. The incidents show that AI has evolved from a supporting tool into a new threat in cyber offense and defense.

Mark Radar|MARK RADAR
All times are in Taipei time (GMT+8)