OpenAI Flags Chinese-Linked Misinformation Campaigns Using ChatGPT

OpenAI has revealed that groups linked to the Chinese government have been misusing its AI tools, particularly ChatGPT, to conduct covert influence operations, spread misinformation, and gather intelligence globally. In a report released Thursday, OpenAI detailed how it disrupted multiple campaigns over the past three months, banning accounts involved in these efforts.

Ben Nimmo, lead investigator on OpenAI’s intelligence team, stated, “We’re seeing a wider variety of covert operations coming out of China, using increasingly diverse tactics.” The report identifies four distinct campaigns using ChatGPT in ways that violated the platform’s usage policies.

One operation, dubbed “Sneer Review,” used AI-generated content to simulate organic conversations across platforms like TikTok, X (formerly Twitter), Reddit, and Facebook. The operation targeted a range of topics, including criticism of Taiwan-based content and opinions on U.S. policy matters.

In another disturbing case, ChatGPT was used to generate internal performance reviews of influence campaigns, indicating systematic planning. Other campaigns saw the AI used for cyber operations—such as modifying scripts and configuring hacking tools—or posing as journalists to gather intelligence through fake accounts.

The report also highlighted ChatGPT-generated messages related to U.S. political issues, including content about former President Donald Trump’s trade policies.

OpenAI reaffirmed its commitment to fighting abuse of its technology and said it continues to invest in threat monitoring and safety systems. The company stressed the importance of responsible AI use as global reliance on generative models increases.