Chinese State Hackers Target AI Policy Experts, Taiwan and Multiple Asian Governments; Credential Theft and Backdoor Access Documented
The timing exposes a coordination problem inside the US government. OpenAI's disruption of the distillation campaign happened weeks AFTER Proofpoint's July phishing wave—meaning Chinese operatives were likely extracting reasoning data from frontier models while simultaneously harvesting the credentials of the officials who would design the export controls to stop them. The backdoor persistence across Taiwan, India, and the Philippines suggests Beijing is building a persistent-access posture in allied intelligence networks specifically to monitor how Washington and Tokyo coordinate Taiwan contingency planning.
OpenAI just disrupted a Moonshot AI distillation campaign (late September), and now Proofpoint and Cisco document a parallel effort: Chinese state actors are systematically harvesting credentials from the exact people who design AI export controls and sit on allied defense policy bodies.
The phishing targets—White House OSTP alumni, Heritage Foundation researchers, Japan's Defense Ministry contacts—are the humans who write the rules that constrain Chinese AI access. Compromising them means Beijing gets advance warning of policy shifts, can map the networks that design sanctions, and knows which allied officials are moving against Chinese tech.
The Antino backdoor across eight Asian governments compounds this: 350 compromised endpoints in Taiwan, India, and the Philippines means persistent access to the intelligence and decision-making apparatus of the Quad's Asian tier. Watch whether the White House names the Chinese agency behind these campaigns in the next 30 days—public attribution is the only lever that forces allied governments to harden their networks faster than Beijing can exploit them.
How many of the compromised endpoints in Taiwan were inside the Defense Ministry or National Security Council? Cisco's count of 350 across eight countries is aggregate; the granularity matters for assessing whether Beijing has readable intelligence on Taiwan contingency planning.
Fault Lines
Strategic intelligence, synthesized daily — with a public track record. Every call graded against what actually happened.