Athena Tong & Yun-Ting Cai
In mid-November, the artificial intelligence (AI) firm Anthropic disclosed the first publicly documented case of a near end-to-end espionage operation orchestrated through a commercial AI coding assistant. The company claimed that a state-linked actor in the People’s Republic of China (PRC) had weaponized its Claude Code tool to automate elements of a cyber-espionage campaign targeting dozens of governments, defense firms, and technology companies worldwide (Anthropic, November 14). The attackers used the large language model to script malware, generate spear-phishing lures, and optimize software infrastructure. U.S. lawmakers responded by summoning Anthropic’s CEO to testify on AI-enabled cyber threats (Cyberscoop, November 26).
The model of state-linked commercial actors supporting the PRC’s hitherto unique approach to Internet governance is now moving overseas. A tranche of documents leaked in September 2025 show that Beijing is systematically equipping foreign governments to replicate and innovate its authoritarian Internet governance model. The materials, consisting of source code, field-test reports, project management software tickets, and internal briefings, reveal how Beijing’s Great Firewall is being packaged and exported as turnkey infrastructure to One Belt One Road (OBOR) initiative and Digital Silk Road partner countries. [1]
No comments:
Post a Comment