Key Points:
- Banned ChatGPT accounts linked to China attempted to influence domestic American political debates.
- The campaigns generated targeted cartoons and slogans to fuel opposition to Trump’s trade tariffs.
- Operatives posed as Americans to falsely claim AI data centers were raising household electricity bills.
- The disinformation was traced directly to a private Chinese tech firm working for government clients.
State-backed Chinese propagandists have targeted the United States by utilizing America’s own artificial intelligence tools to interfere in domestic political debates. A recently published security threat report reveals that operatives utilized the popular ChatGPT platform to generate highly targeted propaganda, focusing on Donald Trump’s trade tariffs and the domestic expansion of massive AI data centers. The discovery marks a significant escalation in the ongoing digital information war between Washington and Beijing, demonstrating how adversarial actors are now turning domestic technology against national interests.
The developer of the generative AI platform discovered and banned two major clusters of accounts likely originating from China following a comprehensive security audit of its networks. The first campaign, which analysts dubbed the “Tech and Tariffs” operation, leveraged the chatbot to generate English-language slogans, social media posts, and political cartoons criticizing President Donald Trump’s economic policies. The AI-generated cartoons depicted Trump as a disruptive global force, showing him swinging a hammer at a wall labeled “Global Future” or sawing apart a ladder while standing on it, in an effort to turn public opinion against trade tariffs.
To protect their domestic interests, the campaign operators gave the AI system strict instructions to shield their leaders from criticism. Public investigative findings showed that the propagandists programmed the accounts to explicitly exclude China’s leader, Xi Jinping, and other senior Beijing officials from any satirical or critical output. This selective targeting confirms that the operation was not a grassroots expression of public frustration, but a highly coordinated, state-guided campaign designed specifically to damage U.S. political cohesion while sheltering Chinese leadership from scrutiny.
The second banned cluster, nicknamed the “Data Center Bandwagon” campaign, focused on exploiting highly sensitive local energy and infrastructure debates. The operatives used the chatbot to pose as ordinary American citizens on social media, posting AI-generated memes and articles falsely claiming that massive artificial intelligence data centers were driving up electricity bills for everyday families. The cartoons generated by the tool portrayed the tech sector as greedy profiteers whose high energy consumption was harming ordinary citizens, tapping into legitimate concerns about grid stability and power costs.
The investigation traced the second campaign directly to a private technology firm operating in China. Security analysts revealed that this private technology firm held active service contracts with Chinese provincial-level government clients, directly linking the disinformation network to state-sponsored actors. Rather than deploying native Chinese large language models, the state-backed operatives chose to utilize the leading American platform, likely believing that its natural language capabilities would produce more authentic, localized American English and help the accounts blend seamlessly into domestic social media forums.
Despite the highly organized nature of the campaigns, the operations ultimately failed to generate any meaningful organic audience engagement. Ben Nimmo, a principal investigator at the AI development firm, told journalists that the state-sponsored operations tried to manipulate what remains a highly legitimate debate about American AI and broader technology policies. Nimmo pointed out the supreme irony of the situation, noting that the adversaries sought to use American AI tools to disrupt the very debate over those same technologies.
While the Chinese-led campaign flopped, the domestic controversy surrounding data center construction is very real and continues to expand across the United States. In 2025, local community opposition and environmental concerns successfully blocked or delayed dozens of major data center projects, representing more than $150 billion in potential capital investment. High-profile politicians have called for immediate moratoriums on new facilities, as ordinary citizens push back against strained power grids, rising household utility bills, and immense water consumption.
The unmasking of these state-sponsored campaigns highlights the double-edged sword of the generative AI era. While these automated tools can democratize productivity, they also provide foreign adversaries with a highly scalable, low-cost platform to generate customized disinformation at a national scale. As the United States prepares for upcoming elections, the threat of AI-driven foreign interference will likely continue to intensify. To protect national sovereignty, both private technology platforms and government intelligence agencies must remain highly vigilant, ensuring that the open tools designed for innovation do not become weapons of geopolitical subversion.
