gcdRadar
archive/Breaches & attacks/AIR-2026-0373
closedclosed 15 Jul 2026Breaches & attacks4 Confirmed real-world lossAI

Suspected Hackers Used AI Agents to Compromise Taiwanese Systems

Hermes, OpenClaw

Threat actors deployed a multi-agent framework to compromise Taiwanese government systems and exfiltrate personnel records. Taiwan confirmed the foreign AI-assisted campaign.

first report
1 Jul 2026
last update
1 Jul 2026

Timeline · 1

  1. 1 Jul 2026Suspected China-Linked Hackers Reportedly Used AI Agents to Compromise Taiwanese Government Systems and Exfiltrate Personnel Records

Sources · 1

Research and catalogues

  1. 2026-07-01
    Suspected China-Linked Hackers Reportedly Used AI Agents to Compromise Taiwanese Government Systems and Exfiltrate Personnel Records
    AI Incident Database · incidentdatabase.ai/cite/1646

Titles and links belong to their publishers. The summary above is written by gcdRadar in its own words; data is CC BY-SA 4.0.

Classification

lane
Breaches & attacks
AI role
AI-enabled attack
categories
AI-assisted intrusionAgent hijack
severity
4 · Confirmed real-world loss
targets
Open-weight model, Framework / tooling, Enterprise deployment
owasp llm top 10
LLM06:2025 Excessive Agency
mitre atlas
AML.T0053 AI Agent Tool Invocation
tags
agentic-threatscredential-compromiseexfiltration
dates
happened 1 Jul 2026 · first reported 1 Jul 2026
origin
AI Incident Database
Derived from the AI Incident Database (CC BY-SA 4.0).
record
confidence 0.95 · created 2026-10-18 06:33Z · updated 2026-10-18 06:59Z

Lane, AI role, categories and severity come from the classifier; the OWASP and ATLAS rows are derived from the categories. How these are assigned

Copy a brief

This record as a short Markdown brief, for a ticket, a chat or a coding agent.

the brief as text
# Suspected Hackers Used AI Agents to Compromise Taiwanese Systems

https://gcdradar.vercel.app/incidents/suspected-hackers-used-ai-agents-to-compromise-taiwanese-sys · AIR-2026-0373 · closed 2026-07-15

- Lane: Breaches & attacks
- Severity: 4 (Confirmed real-world loss)
- Fix status: No fix needed
- First reported: 2026-07-01
- AI role: AI-enabled attack
- Products: Hermes, OpenClaw

## Summary

Threat actors deployed a multi-agent framework to compromise Taiwanese government systems and exfiltrate personnel records. Taiwan confirmed the foreign AI-assisted campaign.

## What to do

No action recorded; read the sources below and the vendor's advisory.

## Latest updates

- 2026-07-01: Suspected China-Linked Hackers Reportedly Used AI Agents to Compromise Taiwanese Government Systems and Exfiltrate Personnel Records (https://incidentdatabase.ai/cite/1646)

## Sources

- 2026-07-01 AI Incident Database: Suspected China-Linked Hackers Reportedly Used AI Agents to Compromise Taiwanese Government Systems and Exfiltrate Personnel Records https://incidentdatabase.ai/cite/1646

Verify against the vendor's advisory before changing production. Data CC BY-SA 4.0, https://gcdradar.vercel.app.

Related

  1. closed
    Hermes AI agent Threat Actor Used Hermes AI Agent for Post-Compromise Activity in Thailand

    Breaches & attacks · Exploited · since 9 Jul 2026

    AI
  2. closed
    Hermes Multi-Agent Framework Compromises Taiwanese Government Systems

    Breaches & attacks · Exploited · since 1 Jul 2026

    AI
  3. closed
    marimo Threat Actor Used LLM Agent to Exfiltrate Database via marimo Notebook

    Breaches & attacks · Exploited · since 10 May 2026

    AI
  4. closed
    Claude Opus 4.7 Claude Opus 4.7 Compromises Real Production Infrastructure in Evaluation

    Research · since 30 Jul 2026

    AI
  5. live
    Cursor AI Aur0ra ransomware operators used AI coding agents for intrusion workflows

    Breaches & attacks · Real loss · since Aug 2026 (approx.)

    AI
  6. closed
    OpenClaw Authorization-Grounded Workflow Attacks on Tool-Using Agents

    Research · since 19 Oct 2026

    AI