OpenAI has announced that it disrupted two covert influence operations that allegedly used its ChatGPT platform to create content aimed at manipulating public debate in the United States. The company stated that it banned "two clusters of ChatGPT accounts likely originating from China" after they used the AI models to promote narratives in an attempt to influence American AI and technology policies.
OpenAI's Commitment to Democratic AI
In a report released on June 10, OpenAI emphasized its mission to ensure that artificial general intelligence benefits all of humanity. By publishing its findings, the company aims to assist governments, technology firms, and the public in identifying and preventing future influence operations. OpenAI also highlighted its role in "identifying and disrupting attempts by authoritarian regimes and their proxies to use AI systems to coerce critics, surveil communities or covertly interfere in democratic societies."
Details of the Two Campaigns
According to OpenAI, one campaign—dubbed "Data Center Bandwagon"—focused on criticizing AI data center projects in the US. It generated social media comments and images claiming that these data centers were driving up electricity costs for American families. The second campaign, called "Tech and Tariffs," produced content criticizing US tariffs and technology policies.
The company revealed that the prompts used by the operators specifically instructed ChatGPT to avoid mentioning Chinese President Xi Jinping and instead focus on US President Donald Trump. Additionally, the second campaign was linked to a network of likely inauthentic social media accounts that spread false claims about ChatGPT user data being compromised. OpenAI clarified that these allegations were entirely false.
Limited Impact and Banned Accounts
OpenAI did not disclose the exact number of accounts removed but confirmed that the campaigns involved hundreds of ChatGPT accounts that have now been banned. Despite the scale, OpenAI found no evidence that the operations significantly influenced public opinion. "The targeting of OpenAI and US data center buildouts is significant not because the operation appears to have shifted public opinion, but because it shows PRC-origin influence operators testing narratives against AI infrastructure," the company stated.
The operators attempted to exploit existing concerns about energy prices and the impact of data center development, but their efforts did not achieve meaningful reach or engagement.
