OpenAI: China ran large scale social media operation to silence dissidents
Mar 3rd 2026
OpenAI released a report detailing what it calls China’s coordinated "cyber special operations" to intimidate and remove critics online, describing extensive use of fake accounts, AI tools and legal forgeries across major social platforms.
- OpenAI published a report titled "Disrupting Malicious Uses of AI" on March 2, 2026, with about a third devoted to China's activities.
- The campaign is described as large scale, resource intensive and sustained, with at least hundreds of staff and thousands of fake accounts.
- Targets include dissidents inside and outside China and representatives of foreign countries, including the prime minister of Japan.
- Tactics documented include forged US county court documents, mass abusive reporting to trigger platform bans, AI generated fake screenshots, and locally deployed AI models.
- OpenAI found and removed a ChatGPT account it says belonged to the Chinese government that uploaded status reports for polishing.
- Platforms named as targets include X, YouTube and Bluesky, and Bluesky confirmed removing a small number of inauthentic coordinated accounts.