OpenAI banned an account linked to an individual associated with Chinese law enforcement, using AI to plan influence activity, harassment, and online operations.

This case study was originally published in OpenAI’s February 2026⁠(opens in a new window) report on disrupting malicious uses of AI.

We banned a ChatGPT account linked to an individual associated with Chinese law enforcement. They tried to use our model to plan a covert influence operation (IO) targeting the Japanese prime minister. Our model refused to assist in such planning. They also asked our model to edit and polish periodic status reports on the conduct of what they termed “cyber special operations” (网络特战) - covert influence operations against domestic and foreign adversaries. The broad range of activities they described illustrates the scope and the scale of these operations, including the range of countries, issues and people they target. Using open-source investigative techniques, we were able to identify activity across the internet that matched some of the activity referenced in the user’s engagement with ChatGPT.

The available evidence suggests that Chinese law enforcement is implementing a strategy of “cyber special operations” to suppress dissent and silence critics both online and offline, at home and abroad. This effort appears to be large-scale, resource-intensive and sustained, counting at least hundreds of staff, thousands of fake accounts across scores of platforms, the use of locally deployed AI models, and a playbook of dozens of tactics. These range from abusive reporting of dissidents’ social media accounts, through mass online posting, to forging documents and impersonating US officials.

Source link