# CNCERT
Latest news and articles about CNCERT
Total: 4 articles found
China’s Cybersecurity Watchdog Sounds Alarm Over Rogue AI Agent ‘Skills’
China’s CNCERT has warned of malicious AI agent plugins being used for model jailbreaking and illegal crypto-mining, highlighting new security risks in the nation’s rapidly expanding AI ecosystem. The regulator emphasized that such 'Skills' could lead to account bans, device performance loss, and legal liabilities for users.
The Hijacked Gateway: China’s Cybersecurity Watchdog Warns of Massive Home Router DNS Attacks
China's CNCERT has issued an emergency alert after discovering a massive DNS hijacking campaign affecting over 700,000 unique IP addresses daily through compromised home routers. The attack redirects users to illegal gambling and pornographic sites, highlighting critical security vulnerabilities in residential networking hardware.
The Price of Autonomy: China Issues Urgent Red Alert Over 'OpenClaw' AI Agents
Chinese authorities have issued an urgent security guide for the popular AI agent tool OpenClaw after discovering over 220,000 instances exposed on the public internet. The government warns that the tool's high privilege requirements and weak default settings pose a severe risk to corporate and personal data security.
China’s Finance Industry on Alert as OpenClaw AI Agent Sparks Security and Fraud Fears
China’s Internet Finance Association has warned that the open‑source AI agent OpenClaw poses serious risks to online finance, citing high default privileges, known vulnerabilities, malicious plugins and persistent memory that can expose sensitive data. The body urged consumers and firms to restrict installation and permissions, and to treat such agents as part of enterprise security governance to prevent fund theft, regulatory breaches and AI‑enabled fraud.