OpenAI disclosed that Chinese propaganda actors have exploited its flagship AI tool to influence American public opinion on contentious topics like U.S. tariffs and the regulation of data centers. These efforts, dating from late 2025 to early 2026, aimed to generate anti-tariff slogans, critical cartoons, and multilingual online commentary, although OpenAI stated their impact appeared minimal.
The company uncovered coordinated groups using ChatGPT to create content criticizing former U.S. trade policies, notably by producing cartoons depicting a disruptive depiction of Donald Trump affecting global affairs. This material circulated on platforms such as X, alongside Chinese-language comments and posts in Italian and Japanese, suggesting a broad multilingual campaign. Another cluster of users traced to a Chinese technology firm with government contracts sought to sway debates around AI development and the expansion of data centers in the U.S., a particularly sensitive issue where many states have proposed or implemented restrictions.
OpenAI shared examples of the propaganda, including cartoons that portrayed the data center and AI industries as exploitative entities damaging communities due to their energy consumption. According to Ben Nimmo, OpenAI’s principal investigator, these operations appeared to manipulate a genuine public discourse on American technology policy, ironically using an American AI platform itself to do so.
The Chinese Embassy in Washington responded by denying knowledge of OpenAI’s findings and rejecting what it termed baseless accusations against China. It emphasized China’s commitment to ensuring AI serves as a positive force globally. This episode underscores how generative AI has become a new frontline in state-sponsored digital influence efforts, even targeting the platforms that underpin it.