OpenAI has detailed a recently disrupted influence operation believed to have originated in China, which utilized ChatGPT to generate content for social media. The company's June 2026 threat report, titled "PRC-linked influence operations are targeting AI debates in the US," alleges that China-based actors employed OpenAI's tools to create material aimed at fostering anti-data centre, anti-tariff, and anti-US sentiment online.
According to the report, a cluster of ChatGPT accounts, likely originating from China, was banned for generating social media content as part of a covert influence campaign. The operation involved crafting political cartoons critical of AI data centers' electricity consumption and former President Donald Trump's trade policies. The users, writing in simplified Chinese, also directed ChatGPT to produce antisemitic memes referencing "Jewish capital" and to create disparaging content about Chinese dissidents. OpenAI noted that the prompts used were consistent with terminology associated with China's public security system.
This generated content, including cartoons, short phrases, and unverified rumors of data leaks, was then disseminated across social media platforms such as X and Facebook through networks of fake accounts. OpenAI observed the irony of an operation using American AI to generate content critical of American AI, though it could not determine the specific reasoning behind this choice.
Despite the sophisticated nature of the alleged operation, OpenAI's assessment indicates that its impact was minimal. The report categorizes the activity as "Category One: activity spanning one platform, with no evidence of breakout." OpenAI also acknowledged that a significant portion of the material fueling these discussions stemmed from legitimate reporting on the environmental impact of data centers and the repercussions of US tariffs.
OpenAI maintains that such operations are significant because they attempt to link US technology policies with economic anxieties and geopolitical instability. The company suggests that Chinese influence operations could exploit legitimate public debates by subtly directing audiences towards distrust of US institutions, technology companies, and democratic policy decisions, thereby advancing Beijing's strategic advantage in AI development.
The findings have the potential to fuel narratives among pro-data center proponents in the US who already attribute anti-AI sentiment to foreign interference. The Chinese Embassy in Washington, responding to inquiries about the report, stated that it had not reviewed OpenAI's findings but firmly opposed any baseless accusations against China, emphasizing a desire for AI to be a force for good.