OpenAI has revealed that a China-linked propaganda group attempted to use the artificial intelligence (AI) chatbot ChatGPT to generate opposition against former U.S. President Donald Trump’s tariff policies and the construction of AI data centers.
In a report released on Wednesday, OpenAI stated that a group believed to originate from China used ChatGPT from late last year to early this year to create content targeting U.S. trade and technology policies, as well as the AI industry.
While OpenAI assessed that these activities had little to no actual impact on public opinion, they noted that this case demonstrates how generative AI is being widely exploited for propaganda efforts, even to criticize AI companies and the industry itself.
According to OpenAI, a group of Chinese-speaking accounts used ChatGPT to create anti-Trump slogans and political cartoons, which were then posted on X (formerly Twitter).
These cartoons depicted Trump destroying a wall labeled Global Future with a hammer or sawing off the ladder he was standing on. The group also used ChatGPT to generate comments for Chinese articles and create content in Japanese and Italian.
OpenAI also identified another group of accounts linked to a tech company involved in Chinese government operations, though they did not disclose the company’s name. This group allegedly attempted to interfere in U.S. debates surrounding AI and data center construction.
In the U.S., over ten states have either implemented or are considering restrictions on data center construction, citing concerns over power consumption, rising electricity costs, and environmental impacts.
The propaganda cartoons portrayed the AI industry as a greedy interest group that consumes vast amounts of electricity at the expense of ordinary citizens.
Ben Nimmo, OpenAI’s chief investigator, described these cases as attempts to manipulate legitimate debates about U.S. AI and technology policy, highlighting the irony of using American-made AI for such efforts.
Reuters reported that while OpenAI concluded these propaganda efforts had limited impact, they illustrate how AI-generated images and content are becoming commonplace in disinformation campaigns.
According to Reuters, the Chinese embassy in the U.S. claimed to be unaware of OpenAI’s investigation but firmly opposed baseless attacks or slander targeting China, asserting that China is working to ensure AI becomes a beneficial force for everyone.