Report
China-Linked Hackers Target AI Policy Experts; OpenAI Accuses Chinese Labs of Model Theft
TA419 impersonated US AI insiders and Anthropic employees in phishing campaigns targeting policy experts. OpenAI accused Chinese labs of large-scale adversarial distillation—stealing reasoning via API queries. Reports reference US-China Super Intelligence Dialogue and AI incident hotline.
TLDR
Escalates AI cold war narrative with documented security incidents and IP theft accusations. Real espionage and alleged model theft tie into policy debates around export controls, self-policing accords, and geopolitical competition. The incidents underscore tensions between cooperation frameworks and adversarial intelligence operations.
Combined views
—
2 Sources, first seen 8h ago