Particle.news

OpenAI Details Chinese State-Linked AI Operation Targeting Dissidents and Foreign Officials

The disclosure spurs limited platform removals after investigators link ChatGPT usage to forged documents, mass reporting, cross-platform deception.

Overview

  • OpenAI’s report describes a large, sustained effort involving hundreds of personnel, thousands of fake accounts across many platforms, and the use of locally deployed AI models.
  • The company says it removed a ChatGPT account linked to Chinese authorities that uploaded periodic status reports and requested editing.
  • Documented tactics include forged U.S. county court papers to force post removals, coordinated abusive reporting to trigger bans, and AI-generated images posed as screenshots.
  • Targets included people inside China, global dissidents, and foreign officials up to Japan’s prime minister, with the prominent X account Teacher Li reporting coordinated pressure.
  • Platforms reported early actions and limits: Bluesky removed a small number of inauthentic accounts, and OpenAI notes ChatGPT declined a 2025 request tied to a smear plan as operators appear to have used local models for similar campaigns.